# Pinecone

> ## Documentation Index

---

# Source: https://docs.pinecone.io/release-notes/2022.md

# Source: https://docs.pinecone.io/assistant-release-notes/2022.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# 2022 releases

## December 22, 2022

#### Pinecone is now available in Google Cloud Marketplace

You can now [sign up for Pinecone billing through Google Cloud Marketplace](/guides/organizations/manage-billing/upgrade-billing-plan).

## December 6, 2022

#### Organizations are generally available

Pinecone now features [organizations](/guides/organizations/understanding-organizations), which allow one or more users to control billing and project settings across multiple projects owned by the same organization.

#### p2 pod type is generally available

The [p2 pod type](/guides/index-data/indexing-overview#p2-pods) is now generally available and ready for production workloads. p2 pods are now available in the Starter plan and support the [dotproduct distance metric](/guides/index-data/create-an-index#dotproduct).

#### Performance improvements

* [Bulk vector\_deletes](/guides/index-data/upsert-data/#deleting-vectors) are now up to 10x faster in many circumstances.
* [Creating collections](/guides/manage-data/back-up-an-index) is now faster.

## October 31, 2022

#### Hybrid search (Early access)

Pinecone now supports keyword-aware semantic search with the new hybrid search indexes and endpoints. Hybrid search enables improved relevance for semantic search results by combining them with keyword search.

This is an **early access** feature and is available only by [signing up](https://www.pinecone.io/hybrid-search-early-access/).

## October 17, 2022

#### Status page

The new [Pinecone Status Page](https://status.pinecone.io/) displays information about the status of the Pinecone service, including the status of individual cloud regions and a log of recent incidents.

## September 16, 2022

#### Public collections

You can now create indexes from public collections, which are collections containing public data from real-world data sources. Currently, public collections include the Glue - SSTB collection, the TREC Question classification collection, and the SQuAD collection.

## August 16, 2022

#### Collections (Public preview)("Beta")

You can now \[make static copies of your index]\(/guides/manage-data/back-up-an-index using collections]\(/guides/manage-data/back-up-an-index#pod-based-index-backups-using-collections). After you create a collection from an index, you can create a new index from that collection. The new index can use any pod type and any number of pods. Collections only consume storage.

This is a **public preview** feature and is not appropriate for production workloads.

#### Vertical scaling

You can now [change the size of the pods](/guides/indexes/pods/scale-pod-based-indexes#increase-pod-size) for a live index to accommodate more vectors or queries without interrupting reads or writes. The p1 and s1 pod types are now available in [4 different sizes](/guides/index-data/indexing-overview/#pods-pod-types-and-pod-sizes): `1x`, `2x`, `4x`, and `8x`. Capacity and compute per pod double with each size increment.

#### p2 pod type (Public preview)("Beta")

The new [p2 pod type](/guides/index-data/indexing-overview/#p2-pods) provides search speeds of around 5ms and throughput of 200 queries per second per replica, or approximately 10x faster speeds and higher throughput than the p1 pod type, depending on your data and network conditions.

This is a **public preview** feature and is not appropriate for production workloads.

#### Improved p1 and s1 performance

The [s1](/guides/index-data/indexing-overview/#s1-pods) and [p1](/guides/index-data/indexing-overview/#p1-pods) pod types now offer approximately 50% higher query throughput and 50% lower latency, depending on your workload.

## July 26, 2022

You can now specify a [metadata filter](/guides/index-data/indexing-overview#metadata/) to get results for a subset of the vectors in your index by calling [describe\_index\_stats](/reference/api/2024-07/control-plane/describe_index) with a [filter](/reference/api/2024-07/control-plane/describe_index#!path=filter\&t=request) object.

The `describe_index_stats` operation now uses the `POST` HTTP request type. The `filter` parameter is only accepted by `describe_index_stats` calls using the `POST` request type. Calls to `describe_index_stats` using the `GET` request type are now deprecated.

## July 12, 2022

#### Pinecone Console Guided Tour

You can now choose to follow a guided tour in the [Pinecone console](https://app.pinecone.io). This interactive tutorial walks you through creating your first index, upserting vectors, and querying your data. The purpose of the tour is to show you all the steps you need to start your first project in Pinecone.

## June 24, 2022

#### Updated response codes

The [create\_index](/reference/api/2024-07/control-plane/create_index), [delete\_index](/reference/api/2024-07/control-plane/delete_index), and `scale_index` operations now use more specific HTTP response codes that describe the type of operation that succeeded.

## June 7, 2022

#### Selective metadata indexing

You can now store more metadata and more unique metadata values! [Select which metadata fields you want to index for filtering](/guides/indexes/pods/manage-pod-based-indexes#selective-metadata-indexing) and which fields you only wish to store and retrieve. When you index metadata fields, you can filter vector search queries using those fields. When you store metadata fields without indexing them, you keep memory utilization low, especially when you have many unique metadata values, and therefore can fit more vectors per pod.

#### Single-vector queries

You can now [specify a single query vector using the vector input](/reference/api/2024-07/data-plane/query/#!path=vector\&t=request). We now encourage all users to query using a single vector rather than a batch of vectors, because batching queries can lead to long response messages and query times, and single queries execute just as fast on the server side.

#### Query by ID

You can now [query your Pinecone index using only the ID for another vector](/reference/api/2024-07/data-plane/query/#!path=id\&t=request). This is useful when you want to search for the nearest neighbors of a vector that is already stored in Pinecone.

#### Improved index fullness accuracy

The index fullness metric in [describe\_index\_stats()](/reference/api/2024-07/control-plane/describe_index#!c=200\&path=indexFullness\&t=response) results is now more accurate.

## April 25, 2022

#### Partial updates (Public preview)

You can now perform a partial update by ID and individual value pairs. This allows you to update individual metadata fields without having to upsert a matching vector or update all metadata fields at once.

#### New metrics

Users on all plans can now see metrics for the past one (1) week in the Pinecone console. Users on the Enterprise plan now have access to the following metrics via the [Prometheus metrics endpoint](/guides/production/monitoring/):

* `pinecone_vector_count`
* `pinecone_request_count_total`
* `pinecone_request_error_count_total`
* `pinecone_request_latency_seconds`
* `pinecone_index_fullness` (Public preview)

**Note:** The accuracy of the `pinecone_index_fullness` metric is improved. This may result in changes from historic reported values. This metric is in public preview.

#### Spark Connector

Spark users who want to manage parallel upserts into Pinecone can now use the [official Spark connector for Pinecone](https://github.com/pinecone-io/spark-pinecone#readme) to upsert their data from a Spark dataframe.

#### Support for Boolean and float metadata in Pinecone indexes

You can now add `Boolean` and `float64` values to [metadata JSON objects associated with a Pinecone index.](/guides/index-data/indexing-overview#metadata)

#### New state field in describe\_index results

The [describe\_index](/reference/api/2024-07/control-plane/describe_index/) operation results now contain a value for `state`, which describes the state of the index. The possible values for `state` are `Initializing`, `ScalingUp`, `ScalingDown`, `Terminating`, and `Ready`.

##### Delete by metadata filter

The [Delete](/reference/api/2024-07/data-plane/delete/) operation now supports filtering my metadata.

---

# Source: https://docs.pinecone.io/release-notes/2023.md

# Source: https://docs.pinecone.io/assistant-release-notes/2023.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# 2023 releases

## December 2023

### Features

* The free Starter plan now supports up to 100 namespaces. [Namespaces](/guides/index-data/indexing-overview#namespaces) let you partition vectors within an index to speed up queries or comply with [multitenancy](/guides/index-data/implement-multitenancy) requirements.

## November 2023

### Features

* The new [Pinecone AWS Reference Architecture](https://github.com/pinecone-io/aws-reference-architecture-pulumi/tree/main) is an open-source, distributed system that performs vector-database-enabled semantic search over Postgres records. You can use it as a learning resource or as a starting point for high-scale use cases.

### SDKs

* [Canopy](https://github.com/pinecone-io/canopy/blob/main/README.md) is a new open-source Retrieval Augmented Generation (RAG) framework and context engine built on top of Pinecone. It enables you to start chatting with your documents or text data with a few simple commands.\
  The latest version of the Canopy SDK (v0.2.0) adds support for OpenAI SDK v1.2.3. See the [release notes](https://github.com/pinecone-io/canopy/releases/tag/V0.2.0) in GitHub for more details.

### Billing

* Pinecone is now registered to collect Value Added Tax (VAT) or Goods and Services Tax (GST) for accounts based in various global regions. If applicable, add your VAT or GST number to your account under **Settings > Billing**.

### October 2023

### Features

* [Collections](/guides/manage-data/back-up-an-index#pod-based-index-backups-using-collections) are now generally available (GA).

### Regions

* Pinecone Azure support via the [‘eastus-azure\` region](/guides/projects/understanding-projects#project-environments) is now generally available (GA).

### SDKs

* The latest version of our Node SDK is v1.1.2. See the [release notes](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/v1.1.2) in GitHub for more details.

### Console

* The Index Browser is now available in the console. This allows you to preview, query, and filter by metadata directly from the console. The Index Browser can be found within the index detail page.
* We’re improved the design of our metrics page to include new charts for record and error count plus additional latencies (p90, p99) to help triage and understand issues.

### Integrations

* Knowledge Base for Amazon Bedrock is now available in private preview. Integrate your enterprise data via retrieval augmented generation (RAG) when building search and GenAI applications. [Learn more](https://www.pinecone.io/blog/amazon-bedrock-integration/).
* Pinecone Sink Connector for Confluent is now available in public preview. Gain access to data streams from across your business to build a real-time knowledge base for your AI applications. [Learn more](https://www.pinecone.io/confluent-integration).

### Billing

* You can now [sign up for Pinecone billing through Microsoft Marketplace](/guides/organizations/manage-billing/upgrade-billing-plan).

### Privacy

* Pinecone is now HIPAA compliant across all of our cloud providers (AWS, Azure, and GCP).

## September 11, 2023

Pinecone Azure support via the [eastus-azure region](/guides/projects/understanding-projects#project-environments) is now generally available (GA).

## August 14, 2023

Pinecone now supports deploying projects to Azure using the new [eastus-azure region](/guides/projects/understanding-projects#project-environments). This is a public preview environment, so test thoroughly before deploying to production.

## June 21, 2023

The new `gcp-starter` region is now in public preview. This region has distinct limitations from other Starter Plan regions. `gcp-starter` is the default region for some new users.

## April 26, 2023

[Indexes in the starter plan](/guides/index-data/indexing-overview#starter-plan) now support approximately 100,000 1536-dimensional embeddings with metadata. Capacity is proportional for other dimensionalities.

## April 3, 2023

Pinecone now supports [new US and EU cloud regions](/guides/projects/understanding-projects#project-environments).

## March 21, 2023

Pinecone now supports SSO on some Enterprise plans. [Contact Support](https://app.pinecone.io/organizations/-/settings/support) to set up your integration.

## March 1, 2023

Pinecone now supports [40kb of metadata per vector](/guides/index-data/indexing-overview#metadata#supported-metadata-size).

## February 22, 2023

#### Sparse-dense embeddings are now in public preview.

Pinecone now supports [vectors with sparse and dense values](/guides/search/hybrid-search#use-a-single-hybrid-index). To use sparse-dense embeddings in Python, upgrade to Python SDK version 2.2.0.

#### Pinecone Python SDK version 2.2.0 is available

Python SDK version 2.2.0 with support for sparse-dense embeddings is now available on [GitHub](https://github.com/pinecone-io/pinecone-python-client) and [PYPI](https://pypi.org/project/pinecone-client/2.2.0/).

## February 15, 2023

#### New Node.js SDK is now available in public preview

You can now try out our new [Node.js SDK for Pinecone](https://sdk.pinecone.io/typescript/).

## February 14, 2023

#### New usage reports in the Pinecone console

You can now monitor your current and projected Pinecone usage with the [**Usage** dashboard](/guides/manage-cost/monitor-usage-and-costs).

## January 31, 2023

#### Pinecone is now available in AWS Marketplace

You can now [sign up for Pinecone billing through Amazon Web Services Marketplace](/guides/organizations/manage-billing/upgrade-billing-plan).

## January 3, 2023

#### Pinecone Python SDK version 2.1.0 is now available on GitHub.

The [latest release of the Python SDK](https://github.com/pinecone-io/pinecone-python-client/releases/tag/2.1.0) makes the following changes:

* Fixes "Connection Reset by peer" error after long idle periods
* Adds typing and explicit names for arguments in all client operations
* Adds docstrings to all client operations
* Adds Support for batch upserts by passing `batch_size` to the upsert method
* Improves gRPC query results parsing performance

---

# Source: https://docs.pinecone.io/release-notes/2024.md

# Source: https://docs.pinecone.io/assistant-release-notes/2024.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# 2024 releases

## December 2024

<Update label="2024-12-23" tags={["Database"]}>
  ### Increased namespaces limit

  Customers on the [Standard plan](https://www.pinecone.io/pricing/) can now have up to 25,000 namespaces per index.
</Update>

<Update label="2024-12-19" tags={["Assistant"]}>
  ### Pinecone Assistant JSON mode and EU region deployment

  Pinecone Assistant can now [return a JSON response](/guides/assistant/chat-with-assistant#json-response).

  ***

  You can now [create an assistant](/reference/api/2025-01/assistant/create_assistant) in the `eu` region.
</Update>

<Update label="2024-12-17" tags={["Database"]}>
  ### Released Spark-Pinecone connector v1.2.0

  Released [`v1.2.0`](https://github.com/pinecone-io/spark-pinecone/releases/tag/v1.2.0) of the [Spark-Pinecone connector](/reference/tools/pinecone-spark-connector). This version introduces support for stream upserts with structured streaming. This enhancement allows users to seamlessly stream data into Pinecone for upsert operations.
</Update>

<Update label="2024-12-10" tags={["Docs"]}>
  ### New integration with HoneyHive

  Added the [HoneyHive](/integrations/honeyhive) integration page.
</Update>

<Update label="2024-12-09" tags={["SDK"]}>
  ### Released Python SDK v5.4.2

  Released [`v5.4.2`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v5.4.2) of the [Pinecone Python SDK](/reference/sdks/python/overview). This release adds a required keyword argument, `metric`, to the `query_namespaces` method. This change enables the SDK to merge results no matter how many results are returned.
</Update>

<Update label="2024-12-06" tags={["General"]}>
  <img className="block max-w-full" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=3155e298e3d3636631238d3b3ac05a9f" data-og-width="2038" width="2038" data-og-height="1050" height="1050" data-path="images/release-notes/launch-week.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=36944215c7e424e0c03bdbffc71be67a 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=020d85488acfadd9554e2ae19e975fdf 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=6b7e05cd22f72c3afb3e01c8ab37d11f 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=e195d61caa0cd69d12339738d532225d 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=32e430d71f4226fc3045bd2dff063ac9 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=6267312f97e819865dfa7db6ce869cdd 2500w" />

  ### Launch week: Pinecone Local

  Pinecone now offers Pinecone Local, an in-memory database emulator available as a Docker image. You can use Pinecone Local to [develop your applications locally](/guides/operations/local-development), or to [test your applications in CI/CD](/guides/production/automated-testing), without connecting to your Pinecone account, affecting production data, or incurring any usage or storage fees. Pinecone Local is in [public preview](/release-notes/feature-availability).
</Update>

<Update label="2024-12-05" tags={["General"]}>
  ### Launch week: Enhanced security and access controls

  Support for [customer-managed encryption keys (CMEK)](/guides/production/configure-cmek) is now in [public preview](/release-notes/feature-availability).

  ***

  You can now [change API key permissions](/guides/projects/manage-api-keys#update-an-api-key).

  ***

  Private Endpoints are now in [general availability](/release-notes/feature-availability). Use Private Endpoints to [connect AWS PrivateLink](/guides/production/connect-to-aws-privatelink) to Pinecone while keeping your VPC private from the public internet.

  ***

  [Audit logs](/guides/production/security-overview#audit-logs), now in early access, provide a detailed record of user and API actions that occur within the Pinecone platform.
</Update>

<Update label="2024-12-04" tags={["Database"]}>
  ### Launch week: `pinecone-rerank-v0` and `cohere-rerank-3.5` on Pinecone Inference

  Released [`pinecone-rerank-v0`](/guides/search/rerank-results#pinecone-rerank-v0), Pinecone's state of the art reranking model that out-performs competitors on widely accepted benchmarks. This model is in [public preview](/release-notes/feature-availability).

  ***

  Pinecone Inference now hosts [`cohere-rerank-3.5`](/guides/search/rerank-results#cohere-rerank-3.5), Cohere's leading reranking model.
</Update>

<Update label="2024-12-03" tags={["Database"]}>
  ### Launch week: Integrated Inference

  You can now use [embedding models](/guides/index-data/create-an-index#embedding-models) and [reranking models](/guides/search/rerank-results#reranking-models) hosted on Pinecone as an integrated part of upserting and searching.
</Update>

<Update label="2024-12-05" tags={["SDK"]}>
  ### Released .NET SDK v2.1.0

  Released [`v2.1.0`](https://github.com/pinecone-io/pinecone-dotnet-client/releases/tag/2.1.0) of the [Pinecone .NET SDK](/reference/sdks/dotnet/overview). This version adds support for [index tags](/guides/manage-data/manage-indexes#configure-index-tags) and introduces the `ClientOptions.IsTlsEnabled` property, which must be set to `false` for non-secure client connections.
</Update>

<Update label="2024-12-05" tags={["Docs"]}>
  ### Improved batch deletion guidance

  Improved the guidance and example code for [deleting records in batches](/guides/manage-data/delete-data#delete-records-in-batches).
</Update>

<Update label="2024-12-02" tags={["Database"]}>
  ### Launch week: Released `pinecone-sparse-english-v0`

  Pinecone Inference now supports [`pinecone-sparse-english-v0`](/guides/search/rerank-results#pinecone-sparse-english-v0), Pinecone's sparse embedding model, which estimates the lexical importance of tokens by leveraging their context, unlike traditional retrieval models like BM25, which rely solely on term frequency. This model is in [public preview](/release-notes/feature-availability).
</Update>

## November 2024

<Update label="2024-11-30" tags={["Docs"]}>
  ### Pinecone docs: New workflows and best practices

  Added typical [Pinecone Database and Pinecone Assistant workflows](/guides/get-started/overview) to the Docs landing page.

  ***

  Updated various examples to use the production best practice of [targeting an index by host](/guides/manage-data/target-an-index) instead of name.

  ***

  Updated the [Amazon Bedrock integration setup guide](/integrations/amazon-bedrock#setup-guide). It now utilizes Bedrock Agents.
</Update>

<Update label="2024-11-27" tags={["SDK"]}>
  ### Released Java SDK v3.1.0

  Released [`v3.1.0`](https://github.com/pinecone-io/pinecone-java-client/releases/tag/v3.1.0) of the [Pinecone Java SDK](/reference/sdks/java/overview). This version introduces support for specifying a base URL for control and data plane operations.
</Update>

<Update label="2024-11-25" tags={["Assistant"]}>
  ### Pinecone Assistant: Context snippets and structured data files

  You can now [retrieve the context snippets](/guides/assistant/retrieve-context-snippets) that Pinecone Assistant uses to generate its responses. This data includes relevant chunks, relevancy scores, and references.

  ***

  You can now [upload JSON (.json) and Markdown (.md) files](/guides/assistant/manage-files#upload-a-local-file) to an assistant.
</Update>

<Update label="2024-11-14" tags={["General"]}>
  ### Monthly spend alerts

  You can now set an organization-wide [monthly spend alert](/guides/manage-cost/manage-cost#set-a-monthly-spend-alert). When your organization's spending reaches the specified limit, you will receive an email notification.
</Update>

<Update label="2024-11-14" tags={["SDK"]}>
  ### Released .NET SDK v2.0.0

  Released [`v2.0.0`](https://github.com/pinecone-io/pinecone-dotnet-client/releases/tag/2.0.0) of the [Pinecone .NET SDK](/reference/sdks/dotnet/overview). This version uses the latest stable API version, `2024-10`, and adds support for [embedding](/reference/api/2025-01/inference/generate-embeddings), [reranking](https://docs.pinecone.io/reference/api/2025-01/inference/rerank), and [import](/guides/index-data/import-data). It also adds support for using the .NET SDK with [proxies](/reference/sdks/dotnet/overview#proxy-configuration).
</Update>

<Update label="2024-11-13" tags={["SDK"]}>
  ### Released Python SDK v5.4.0 and v5.4.1

  Released [`v5.4.0`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v5.4.0) and [`v5.4.1`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v5.4.1) of the [Pinecone Python SDK](/reference/sdks/python/overview). `v5.4.0` adds a `query_namespaces` utility method to [run a query in parallel across multiple namespaces](/reference/sdks/python/overview#query-across-namespaces) in an index and then merge the result sets into a single ranked result set with the `top_k` most relevant results. `v5.4.1` adds support for the `pinecone-plugin-inference` package required for some [integrated inference](/reference/api/introduction#inference) operations.
</Update>

<Update label="2024-11-08" tags={["General"]}>
  ### Enabled CSV export of usage and costs

  You can now download a CSV export of your organization's usage and costs from the [Pinecone console](https://app.pinecone.io/organizations/-/settings/usage).
</Update>

<Update label="2024-11-07" tags={["General"]}>
  ### Added Support chat in the console

  You can now chat with the Pinecone support bot and submit support requests directly from the [Pinecone console](https://app.pinecone.io/organizations/-/settings/support).
</Update>

<Update label="2024-11-05" tags={["Docs"]}>
  ### Published Assistant quickstart guide

  Added an [Assistant quickstart](/guides/assistant/quickstart/sdk-quickstart).
</Update>

## October 2024

<Update label="2024-10-31" tags={["SDK"]}>
  ### Cequence released updated Scala SDK

  [Cequence](https://github.com/cequence-io) released a new version of their community-supported [Scala SDK](https://github.com/cequence-io/pinecone-scala) for Pinecone. See their [blog post](https://cequence.io/blog/industry-know-how/introducing-the-pinecone-scala-client-async-intuitive-and-ready-for-action) for details.
</Update>

<Update label="2024-10-28" tags={["Database"]}>
  ### Added index tagging for categorization

  You can now [add index tags](/guides/manage-data/manage-indexes#configure-index-tags) to categorize and identify indexes.
</Update>

<Update label="2024-10-24" tags={["SDK"]}>
  ### Released major SDK updates: Node.js, Go, Java, and Python

  Released [`v4.0.0`](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/v4.0.0) of the [Pinecone Node.js SDK](/reference/sdks/node/overview). This version uses the latest stable API version, `2024-10`, and adds support for [reranking](/guides/search/rerank-results) and [import](/guides/index-data/import-data).

  ***

  Released [`v2.0.0`](https://github.com/pinecone-io/go-pinecone/releases/tag/v2.0.0) of the [Pinecone Go SDK](/reference/sdks/go/overview). This version uses the latest stable API version, `2024-10`, and adds support for [reranking](/guides/search/rerank-results) and [import](/guides/index-data/import-data).

  ***

  Released [`v3.0.0`](https://github.com/pinecone-io/pinecone-java-client/releases/tag/v3.0.0) of the [Pinecone Java SDK](/reference/sdks/java/overview). This version uses the latest stable API version, `2024-10`, and adds support for [embedding](/reference/api/2025-01/inference/generate-embeddings), [reranking](/reference/api/2025-01/inference/rerank), and [import](/guides/index-data/import-data).

  <Warning>
    `v3.0.0` also includes the following [breaking change](/reference/api/versioning#breaking-changes): The `control` class has been renamed `db_control`. Before upgrading to this version, be sure to update all relevant `import` statements to account for this change.

    For example, you would change `import org.openapitools.control.client.model.*;` to `import org.openapitools.db_control.client.model.*;`.
  </Warning>

  ***

  `v5.3.0` and `v5.3.1` of the [Pinecone Python SDK](/reference/sdks/python/overview) use the latest stable API version, `2024-10`. These versions were release previously.
</Update>

<Update label="2024-10-24" tags={["API"]}>
  ### Pinecone API version `2024-10` is now the latest stable version

  `2024-10` is now the latest [stable version](/reference/api/versioning#release-schedule) of the [Database API](/reference/api/2024-10/data-plane/) and [Inference API](/reference/api/2024-10/inference/). For highlights, see [SDKs](#sdks) below.
</Update>

<Update label="2024-10-17" tags={["General"]}>
  ### Pinecone Inference now available on the free Starter plan

  The free [Starter plan](https://www.pinecone.io/pricing/) now supports [reranking documents with Pinecone Inference](/guides/search/rerank-results).
</Update>

<Update label="2024-10-17" tags={["Database"]}>
  ### Customer-managed encryption keys (CMEK) in early access

  You can now use [customer-managed encryption keys (CMEK)](/guides/production/configure-cmek) to secure indexes within a Pinecone project. This feature is in [early access](/release-notes/feature-availability).
</Update>

<Update label="2024-10-07" tags={["Database"]}>
  ### Serverless index monitoring generally available

  Monitoring serverless indexes with [Prometheus](/guides/production/monitoring#monitor-with-prometheus) or [Datadog](/integrations/datadog) is now in [general availability](/release-notes/feature-availability).
</Update>

<Update label="2024-10-03" tags={["Database"]}>
  ### Data import from Amazon S3 in public preview

  You can now [import data](/guides/index-data/import-data) into an index from [Amazon S3](/guides/operations/integrations/integrate-with-amazon-s3). This feature is in [public preview](/release-notes/feature-availability).
</Update>

<Update label="2024-10-02" tags={["Assistant"]}>
  ### Chat and update features added to Assistant

  Added the [`chat_assistant`](/reference/api/2025-01/assistant/chat_assistant) endpoint to the Assistant API. It can be used to chat with your assistant, and get responses and citations back in a structured form.

  ***

  You can now add instructions when [creating](/guides/assistant/create-assistant) or [updating](/guides/assistant/manage-assistants#update-an-existing-assistant) an assistant. Instructions are a short description or directive for the assistant to apply to all of its responses. For example, you can update the instructions to reflect the assistant's role or purpose.

  ***

  You can now [update an existing assistant](/guides/assistant/manage-assistants#update-an-existing-assistant) with new instructions or metadata.
</Update>

## September 2024

<Update label="2024-09-25" tags={["Docs"]}>
  Added the [Matillion](/integrations/matillion) integration page.
</Update>

<Update label="2024-09-23" tags={["Docs"]}>
  Added guidance on using the Node.js SDK with [proxies](/reference/sdks/node/overview#proxy-configuration).
</Update>

<Update label="2024-09-19" tags={["SDK"]}>
  Released [`v5.3.1`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v5.3.1) of the [Pinecone Python SDK](/reference/sdks/python/overview). This version adds a missing `python-dateutil` dependency.

  ***

  Released [`v1.1.1`](https://github.com/pinecone-io/go-pinecone/releases/tag/v1.1.1) of the [Pinecone Go SDK](/reference/sdks/go/overview). This version adds support for non-secure client connections.

  ***

  Released [`v2.1.0`](https://github.com/pinecone-io/pinecone-java-client/releases/tag/v2.1.0) of the [Pinecone Java SDK](/reference/sdks/java/overview). This version adds support for non-secure client connections.
</Update>

<Update label="2024-09-18" tags={["SDK"]}>
  Released [`v5.3.0`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v5.3.0) of the [Pinecone Python SDK](/reference/sdks/python/overview). This version adds support for [import](/guides/index-data/import-data) operations. This feature is in [public preview](/release-notes/feature-availability).
</Update>

<Update label="2024-09-18" tags={["Assistant"]}>
  Added the `metrics_alignment` operation, which provides a way to [evaluate the correctness and completeness of a response](/guides/assistant/evaluate-answers) from a RAG system. This feature is in [public preview](/release-notes/feature-availability).

  ***

  When using Pinecone Assistant, you can now [choose an LLM](/guides/assistant/chat-with-assistant#choose-a-model-for-your-assistant) for the assistant to use and [filter the assistant's responses by metadata](/guides/assistant/chat-with-assistant#filter-chat-with-metadata).
</Update>

<Update label="2024-09-18" tags={["Docs"]}>
  Added the [Datavolo](/integrations/datavolo) integration pages.
</Update>

<Update label="2024-09-17" tags={["SDK"]}>
  Released [`v5.2.0`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v5.2.0) of the [Pinecone Python SDK](/reference/sdks/python/overview). This version adds support for [reranking documents with Pinecone Inference](/guides/search/rerank-results); it is no longer necessary to install the `pinecone-plugin-inference` package separately. This feature is in [public preview](/release-notes/feature-availability).
</Update>

<Update label="2024-09-16" tags={["Database"]}>
  [Prometheus monitoring for serverless indexes](/guides/production/monitoring#monitor-with-prometheus) is now in [public preview](/release-notes/feature-availability).
</Update>

<Update label="2024-09-12" tags={["SDK"]}>
  Released [`v3.0.3`](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/3.0.3) of the [Pinecone Node.js SDK](/reference/sdks/node/overview). This version removes extra logging and makes general internal enhancements.
</Update>

<Update label="2024-09-11" tags={["General"]}>
  If you are upgrading from the [Starter plan](https://www.pinecone.io/pricing/), you can now connect your Pinecone organization to the [AWS Marketplace](/guides/organizations/manage-billing/upgrade-billing-plan), [Google Cloud Marketplace](/guides/organizations/manage-billing/upgrade-billing-plan), or [Microsoft Marketplace](/guides/organizations/manage-billing/upgrade-billing-plan) for billing purposes.
</Update>

<Update label="2024-09-11" tags={["General"]}>
  Refreshed the navigation and overall visual interface of the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/).
</Update>

<Update label="2024-09-11" tags={["Docs"]}>
  Added Go examples for [batch upserts](/guides/index-data/upsert-data#upsert-in-batches), [parallel upserts](/guides/index-data/upsert-data#send-upserts-in-parallel), and [deleting all records for a parent document](/guides/index-data/data-modeling#delete-chunks).
</Update>

## August 2024

<Update label="2024-08-28" tags={["Docs"]}>
  Added the [Aryn](/integrations/aryn) integration page.
</Update>

<Update label="2024-08-28" tags={["SDK"]}>
  Released [`v3.0.2`](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/3.0.2) of the [Pinecone Node.js SDK](/reference/sdks/node/overview). This version removes a native Node utility function that was causing issues for users running in `Edge`. There are no downstream affects of its removal; existing code should not be impacted.
</Update>

<Update label="2024-08-27" tags={["SDK"]}>
  Released [`v5.1.0`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v5.1.0) of the [Pinecone Python SDK](/reference/sdks/python/overview). With this version, the SDK can now be installed with `pip install pinecone` / `pip install "pinecone[grpc]"`. This version also includes a `has_index()` helper function to check if an index exists.

  ***

  Released [`v0.1.0`](https://github.com/pinecone-io/pinecone-rust-client/releases/tag/v0.1.0) and [`v0.1.1`](https://github.com/pinecone-io/pinecone-rust-client/releases/tag/v0.1.1) of the [Pinecone Rust SDK](/reference/sdks/rust/overview). The Rust SDK is in "alpha" and is under active development. The SDK should be considered unstable and should not be used in production. Before a 1.0 release, there are no guarantees of backward compatibility between minor versions. See the [Rust SDK README](https://github.com/pinecone-io/pinecone-rust-client) for full installation instructions and usage examples.
</Update>

<Update label="2024-08-27" tags={["SDK"]}>
  Released [`v1.0.0`](https://github.com/pinecone-io/pinecone-dotnet-client/releases/tag/1.0.0) of the [Pinecone .NET SDK](/reference/sdks/dotnet/overview). For usage examples, see [our guides](/guides/get-started/quickstart) or the [GitHub README](https://github.com/pinecone-io/pinecone-dotnet-client).
</Update>

<Update label="2024-08-26" tags={["Database"]}>
  You can now [back up](/guides/manage-data/back-up-an-index) and [restore](/guides/manage-data/restore-an-index) serverless indexes. This feature is in public preview.

  ***

  Serverless indexes are now in [general availability on GCP and Azure](/guides/index-data/create-an-index#cloud-regions) for Standard and Enterprise plans.

  ***

  You can now deploy [serverless indexes](/guides/index-data/indexing-overview) in the `europe-west1` (Netherlands) region of GCP.
</Update>

<Update label="2024-08-26" tags={["SDK"]}>
  Released [`v1.1.0`](https://github.com/pinecone-io/go-pinecone/releases/tag/v1.1.0) of the [Pinecone Go SDK](/reference/sdks/go/overview). This version adds support for generating embeddings via [Pinecone Inference](/reference/api/introduction#inference).
</Update>

<Update label="2024-08-21" tags={["Docs"]}>
  Added the [Nexla](/integrations/nexla) integration page.
</Update>

<Update label="2024-08-15" tags={["Assistant"]}>
  [Pinecone Assistant](/guides/assistant/overview) is now in [public preview](/release-notes/feature-availability).
</Update>

<Update label="2024-08-13" tags={["API"]}>
  The Pinecone Inference API now supports [reranking](https://docs.pinecone.io/guides/search/rerank-results). This feature is in [public preview](/release-notes/feature-availability).
</Update>

<Update label="2024-08-06" tags={["SDK"]}>
  Released [`v1.0.0`](https://github.com/pinecone-io/go-pinecone/releases/tag/v1.0.0) of the [Pinecone Go SDK](/reference/sdks/go/overview). This version depends on Pinecone API version `2024-07` and includes the ability to [prevent accidental index deletion](/guides/manage-data/manage-indexes#configure-deletion-protection). With this version, the Go SDK is [officially supported](/troubleshooting/pinecone-support-slas) by Pinecone.
</Update>

<Update label="2024-08-05" tags={["Docs"]}>
  Added the [Nuclia](/integrations/nuclia) integration page
</Update>

## July 2024

<Update label="2024-07-30" tags={["Docs"]}>
  Added the [Redpanda](/integrations/redpanda) integration page.
</Update>

<Update label="2024-07-26" tags={["Docs"]}>
  Updated the [Build a RAG chatbot](/guides/get-started/build-a-rag-chatbot) guide to use Pinecone Inference for generating embeddings.
</Update>

<Update label="2024-07-19" tags={["Database"]}>
  Added the ability to [prevent accidental index deletion](/guides/manage-data/manage-indexes#configure-deletion-protection).
</Update>

<Update label="2024-07-19" tags={["SDK"]}>
  Released [`v5.0.0`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v5.0.0) of the [Pinecone Python SDK](/reference/sdks/python/overview). This version depends on Pinecone API version `2024-07` and includes the ability to [prevent accidental index deletion](/guides/manage-data/manage-indexes#configure-deletion-protection). Additionally, the `pinecone-plugin-inference` package required to [generate embeddings with Pinecone Inference](/reference/api/2025-01/inference/generate-embeddings) is now included by default; it is no longer necessary to install the plugin separately.

  ***

  Released [`v3.0.0`](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/v3.0.0) of the [Pinecone Node.js SDK](/reference/sdks/node/overview). This version depends on Pinecone API version `2024-07` and includes the ability to [prevent accidental index deletion](/guides/manage-data/manage-indexes#configure-deletion-protection). Additionally, this version supports generating embeddings via [Pinecone Inference](/reference/api/2025-01/inference/generate-embeddings).

  ***

  Released [`v2.0.0`](https://github.com/pinecone-io/pinecone-java-client/releases/tag/v2.0.0) of the [Pinecone Java SDK](/reference/sdks/java/overview). This version depends on Pinecone API version `2024-07` and includes the ability to [prevent accidental index deletion](/guides/manage-data/manage-indexes#configure-deletion-protection). Additionally, this version includes the following **breaking changes**:

  * `createServerlessIndex()` now requires a new argument: `DeletionProtection.ENABLED` or `DeletionProtection.DISABLED`.
  * `configureIndex()` has been renamed `configurePodsIndex()`.

  For more details, see the [Java SDK v2.0.0 migration guide](https://github.com/pinecone-io/pinecone-java-client/blob/main/v2-migration.md).
</Update>

<Update label="2024-07-19" tags={["API"]}>
  Released version `2024-07` of the [Database API](/reference/api/2024-07/data-plane/) and Inference API. This version includes the following highlights:

  * The [`create_index`](/reference/api/2024-07/control-plane/create_index) and [`configure_index`](/reference/api/2024-07/control-plane/configure_index) endpoints now support the `deletion_protection` parameter. Setting this parameter to `"enabled"` prevents an index from accidental deletion. For more details, see [Prevent index deletion](/guides/manage-data/manage-indexes#configure-deletion-protection).

  * The [`describe_index`](/reference/api/2024-07/control-plane/describe_index) and [`list_index`](/reference/api/2024-07/control-plane/list_indexes) responses now include the `deletion_protection` field. This field indicates whether deletion protection is enabled for an index.

  * The `spec.serverless.cloud` and `spec.serverless.region` parameters of [`create_index`](/reference/api/2024-07/control-plane/create_index) now support `gcp` / `us-central` and `azure` / `eastus2` as part of the serverless public preview on GCP and Azure.
</Update>

<Update label="2024-07-17" tags={["Database"]}>
  Serverless indexes are now in [public preview on Azure](/guides/index-data/create-an-index#cloud-regions) for Standard and Enterprise plans.
</Update>

<Update label="2024-07-10" tags={["Database"]}>
  Released [version 1.1.0](https://github.com/pinecone-io/spark-pinecone/releases/tag/v1.1.0) of the official Spark connector for Pinecone. In this release, you can now set a [source tag](/integrations/build-integration/attribute-usage-to-your-integration). Additionally, you can now [upsert records](/guides/index-data/upsert-data) with 40KB of metadata, increased from 5KB.
</Update>

<Update label="2024-07-01" tags={["Database"]}>
  Serverless indexes are now in [public preview on GCP](/guides/index-data/create-an-index#cloud-regions) for Standard and Enterprise plans.
</Update>

<Update label="2024-07-01" tags={["Docs"]}>
  Added an introduction to [key concepts in Pinecone](/guides/get-started/concepts) and how they relate to each other.

  ***

  Added the [Twelve Labs](/integrations/twelve-labs) integration page.
</Update>

## June 2024

<Update label="2024-06-30" tags={["Docs"]}>
  Added a [model gallery](/models/overview) with details and guidance on popular embedding and reranking models, including models hosted on Pinecone's infrastructure.
</Update>

<Update label="2024-06-28" tags={["SDK"]}>
  Released [version 1.2.2](https://github.com/pinecone-io/pinecone-java-client/releases/tag/v1.2.2) of the Pinecone Java SDK. This release simplifies the proxy configuration process. It also fixes an issue where the user agent string was not correctly setup for gRPC calls. Now, if the source tag is set by the user, it is appended to the custom user agent string.
</Update>

<Update label="2024-06-27" tags={["Database"]}>
  You can now load a [sample dataset](/guides/data/use-sample-datasets) into a new project.

  ***

  Simplified the process for [migrating paid pod indexes to serverless](/guides/indexes/pods/migrate-a-pod-based-index-to-serverless).
</Update>

<Update label="2024-06-25" tags={["Assistant"]}>
  The [Assistant API](/guides/assistant/overview) is now in beta release.
</Update>

<Update label="2024-06-20" tags={["Database"]}>
  The [Inference API](/reference/api/introduction#inference) is now in public preview.
</Update>

<Update label="2024-06-18" tags={["Docs"]}>
  Added a new [legal semantic search](https://docs.pinecone.io/examples/sample-apps/legal-semantic-search) sample app that demonstrates low-latency natural language search over a knowledge base of legal documents.
</Update>

<Update label="2024-06-17" tags={["Docs"]}>
  Added the [Instill](/integrations/instill) integration page.
</Update>

<Update label="2024-06-07" tags={["Docs"]}>
  Added the [Langtrace](/integrations/langtrace) integration page.
</Update>

<Update label="2024-06-06" tags={["Docs"]}>
  Updated Python code samples to use the gRPC version of the [Python SDK](/reference/sdks/python/overview), which is more performant than the Python SDK that interacts with Pinecone via HTTP requests.
</Update>

<Update label="2024-06-06" tags={["SDK"]}>
  Released [version 4.1.1](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v4.1.1) of the Pinecone Python SDK. In this release, you can now use colons inside source tags. Additionally, the gRPC version of the Python SDK now allows retries of up to `MAX_MSG_SIZE`.
</Update>

<Update label="2024-06-05" tags={["Database"]}>
  The Enterprise [quota for namespaces per serverless index](/reference/api/database-limits#namespaces-per-serverless-index) has increased from 50,000 to 100,000.
</Update>

<Update label="2024-06-04" tags={["Docs"]}>
  Added the [Fleak](/integrations/fleak) integration page.
</Update>

## May 2024

<Update label="2024-05-31" tags={["SDK"]}>
  Released [version 1.2.1](https://github.com/pinecone-io/pinecone-java-client/releases/tag/v1.2.1) of the [Pinecone Java SDK](/reference/sdks/java/overview). This version fixes the error `Could Not Find NameResolverProvider` using uber jar.
</Update>

<Update label="2024-05-31" tags={["Docs"]}>
  Added the [Gathr](/integrations/gathr) integration page.
</Update>

<Update label="2024-05-30" tags={["SDK"]}>
  Released [version 1.1.0](https://github.com/pinecone-io/pinecone-java-client/releases/tag/v1.1.0) of the [Pinecone Java SDK](/reference/sdks/java/overview). This version adds the ability to [list record IDs with a common prefix](/guides/manage-data/list-record-ids#list-the-ids-of-records-with-a-common-prefix).
</Update>

<Update label="2024-05-29" tags={["SDK"]}>
  Released version [1.2.0](https://github.com/pinecone-io/pinecone-java-client/releases/tag/v1.2.0) of the [Pinecone Java SDK](/reference/sdks/java/overview). This version adds the ability to [list all record IDs in a namespace](/guides/manage-data/list-record-ids#list-the-ids-of-all-records-in-a-namespace).
</Update>

<Update label="2024-05-29" tags={["Docs"]}>
  Added the following integration pages:

  * [Apify](/integrations/apify)
  * [Context Data](/integrations/context-data)
  * [Estuary](/integrations/estuary)
  * [GitHub Copilot](/integrations/github-copilot)
  * [Jina](/integrations/jina)
  * [FlowiseAI](/integrations/flowise)
  * [OctoAI](/integrations/octoai)
  * [Streamnative](/integrations/streamnative)
  * [Traceloop](/integrations/traceloop)
  * [Unstructured](/integrations/unstructured)
  * [VoyageAI](/integrations/voyage)
</Update>

<Update label="2024-05-21" tags={["General"]}>
  You can now use the `ConnectPopup` function to bypass the [**Connect** widget](/integrations/build-integration/connect-your-users-to-pinecone) and open the "Connect to Pinecone" flow in a popup. This can be used in an app or website for a seamless Pinecone signup and login process.
</Update>

<Update label="2024-05-16" tags={["Database"]}>
  Released [version 1.0.0](https://github.com/pinecone-io/spark-pinecone/releases/tag/v1.0.0) of the official Spark connector for Pinecone. In this release, you can now upsert records into [serverless indexes](/guides/index-data/indexing-overview).
</Update>

<Update label="2024-05-08" tags={["Database"]}>
  Pinecone now supports [AWS PrivateLink](/guides/production/connect-to-aws-privatelink). Create and use [Private Endpoints](/guides/production/connect-to-aws-privatelink#manage-private-endpoints) to connect AWS PrivateLink to Pinecone while keeping your VPC private from the public internet.
</Update>

<Update label="2024-05-01" tags={["SDK"]}>
  Released [version 4.0.0](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v4.0.0) of the Pinecone Python SDK. In this release, we are upgrading the `protobuf` dependency in our optional `grpc` extras from `3.20.3` to `4.25.3`. Significant performance improvements have been made with this update. This is a breaking change for users of the optional GRPC addon ([installed with `pinecone[grpc]`](https://github.com/pinecone-io/pinecone-python-client?tab=readme-ov-file#working-with-grpc-for-improved-performance)).
</Update>

## April 2024

<Update label="2024-04-30" tags={["Docs"]}>
  * The docs now have a new AI chatbot. Use the search bar at the top of our docs to find related content across all of our resources.
  * We've updated the look and feel of our [example notebooks](/examples/notebooks) and [sample apps](/examples/sample-apps). A new sample app, [Namespace Notes](/examples/sample-apps/namespace-notes), a simple multi-tenant RAG app that uploads documents, has also been added.
</Update>

<Update label="2024-04-30" tags={["Database"]}>
  The free [Starter plan](https://www.pinecone.io/pricing/) now includes 1 project, 5 serverless indexes in the `us-east-1` region of AWS, and up to 2 GB of storage. Although the Starter plan has stricter [limits](/reference/api/database-limits) than other plans, you can [upgrade](/guides/organizations/manage-billing/upgrade-billing-plan) whenever you're ready.
</Update>

<Update label="2024-04-26" tags={["General"]}>
  Pinecone now provides a [**Connect** widget](/integrations/build-integration/connect-your-users-to-pinecone) that can be embedded into an app, website, or Colab notebook for a seamless signup and login process.
</Update>

<Update label="2024-04-22" tags={["Docs"]}>
  Added the [lifecycle policy of the Pinecone API](/release-notes/feature-availability), which describes the availability phases applicable to APIs, features, and SDK versions.
</Update>

<Update label="2024-04-22" tags={["Database"]}>
  As announced in January 2024, [control plane](/reference/api/2024-07/control-plane) operations like `create_index`, `describe_index`, and `list_indexes` now use a single global URL, `https://api.pinecone.io`, regardless of the cloud environment where an index is hosted. This is now in general availability. As a result, the legacy version of the API, which required regional URLs for control plane operations, is deprecated as of April 15, 2024 and will be removed in a future, to be announced, release.
</Update>

<Update label="2024-04-15" tags={["Docs"]}>
  Added the [Terraform](/integrations/terraform) integration page.
</Update>

<Update label="2024-04-15" tags={["SDK"]}>
  Released version 0.9.0 of the [Canopy SDK](https://github.com/pinecone-io/canopy/blob/main/README.md). This version adds support for OctoAI LLM and embeddings, and Qdrant as a supported knowledge base. See the [v0.9.0 release notes](https://github.com/pinecone-io/canopy/releases/tag/v0.9.0) in GitHub for more details.
</Update>

<Update label="2024-04-12" tags={["Database"]}>
  You can now deploy [serverless indexes](/guides/index-data/indexing-overview) in the `eu-west-1` region of AWS.
</Update>

<Update label="2024-04-12" tags={["SDK"]}>
  Released version 1.0.0 of the [Pinecone Java SDK](/reference/sdks/java/overview). With this version, the Java SDK is [officially supported](/troubleshooting/pinecone-support-slas) by Pinecone. For full details on the release, see the [v1.0.0 release notes](https://github.com/pinecone-io/pinecone-java-client/releases/tag/v1.0.0) in GitHub. For usage examples, see [our guides](/guides/get-started/quickstart) or the [GitHub README](https://github.com/pinecone-io/pinecone-java-client). To migrate to v1.0.0 from version 0.8.x or below, see the [Java v1.0.0 migration guide](https://github.com/pinecone-io/pinecone-java-client/blob/main/v1-migration.md).
</Update>

## March 2024

<Update label="2024-03-31" tags={["Docs"]}>
  Added a [Troubleshooting](https://docs.pinecone.io/troubleshooting/) section, which includes content on best practices, troubleshooting, and how to address common errors.

  ***

  Added an explanation of the [Pinecone serverless architecture](/guides/get-started/database-architecture), including descriptions of the high-level components and explanations of the distinct paths for writes and reads.

  ***

  Added [considerations for querying serverless indexes with metadata filters](/guides/index-data/indexing-overview#metadata#considerations-for-serverless-indexes).
</Update>

<Update label="2024-03-30" tags={["SDK"]}>
  Released [version 3.2.2](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v3.2) of the [Pinecone Python SDK](/reference/sdks/python/overview). This version fixes a minor issue introduced in v3.2.0 that resulted in a `DeprecationWarning` being incorrectly shown to users who are not passing in the deprecated `openapi_config` property. This warning can safely be ignored by anyone who is not preparing to upgrade.
</Update>

<Update label="2024-03-29" tags={["SDK"]}>
  Released [version 3.2.0](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v3.2.0) of the [Pinecone Python SDK](/reference/sdks/python/overview). This version adds four optional configuration properties that enable the use of Pinecone [via proxy](/reference/sdks/python/overview#proxy-configuration).
</Update>

<Update label="2024-03-28" tags={["SDK"]}>
  Released [version 2.2.0](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/v2.2.0) of the [Pinecone Node.js SDK](/reference/sdks/node/overview). This releases adds an optional `sourceTag` that you can set when constructing a Pinecone client to help Pinecone associate API activity to the specified source.
</Update>

<Update label="2024-03-27" tags={["SDK"]}>
  Released version 0.4.1 of the [Pinecone Go SDK](/reference/sdks/go/overview). This version adds an optional `SourceTag` that you can set when constructing a Pinecone client to help Pinecone associate API activity to the specified source.

  ***

  Released version 2.2.0 of the [Pinecone Node.js SDK](/reference/sdks/node/overview).

  ***

  Released [version 0.4.1](https://github.com/pinecone-io/go-pinecone/releases/tag/v0.4.1) of the [Pinecone Go SDK](/reference/sdks/go/overview).
</Update>

<Update label="2024-03-25" tags={["SDK"]}>
  Released version 3.2.1 of the [Pinecone Python SDK](/reference/sdks/python/overview). This version adds an optional `source_tag` that you can set when constructing a Pinecone client to help Pinecone associate API activity to the specified source. See the [v3.2.1 release notes](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v3.2.1) in GitHub for more details.
</Update>

<Update label="2024-03-21" tags={["SDK"]}>
  Released version 0.8.1 of the [Canopy SDK](https://github.com/pinecone-io/canopy/blob/main/README.md). This version includes bug fixes, the removal of an unused field for Cohere chat calls, and added guidance on creating a knowledge base with a specified record encoder when using the core library. See the [v0.8.1 release notes](https://github.com/pinecone-io/canopy/releases/tag/v0.8.1) in GitHub for more details.
</Update>

<Update label="2024-03-20" tags={["Database"]}>
  The [Pinecone console](https://app.pinecone.io) has a new look and feel, with a brighter, minimalist design; reorganized menu items for quicker, more intuitive navigation; and easy access to recently viewed indexes in the sidebar.

  ***

  When viewing the list of indexes in a project, you can now search indexes by index name; sort indexes alphabetically, by how recently they were viewed or created, or by status; and filter indexes by index type (serverless, pod-based, or starter).
</Update>

<Update label="2024-03-15" tags={["SDK"]}>
  Released version 0.4.0 of the [Pinecone Go SDK](/reference/sdks/go/overview). This version is a comprehensive re-write and adds support for all current [Pinecone API operations](/reference/api/introduction).
</Update>

<Update label="2024-03-14" tags={["Database"]}>
  Fixed a bug that caused inaccurate index fullness reporting for some pod-based indexes on GCP.

  ***

  You can now deploy [serverless indexes](/guides/index-data/indexing-overview) in the `us-east-1` region of AWS.
</Update>

<Update label="2024-03-06" tags={["SDK"]}>
  Released version 2.1.0 of the [Pinecone Node.js SDK](/reference/sdks/node/overview). This version adds support for [listing the IDs of records in a serverless index](/guides/manage-data/list-record-ids). You can list all records or just those with a common ID prefix.
</Update>

<Update label="2024-03-05" tags={["General"]}>
  You can now [configure single single-on](/guides/production/configure-single-sign-on/okta) to manage your teams' access to Pinecone through any identity management solution with SAML 2.0 support, such as Okta. This feature is available on the [Enterprise plan](https://www.pinecone.io/pricing/) only.
</Update>

## February 2024

<Update label="2024-02-29" tags={["Docs"]}>
  Updated the [Langchain integration guide](/integrations/langchain) to avoid a [namespace collision issue](/troubleshooting/pinecone-attribute-errors-with-langchain).
</Update>

<Update label="2024-02-27" tags={["SDK"]}>
  The latest version of the [Canopy SDK](https://github.com/pinecone-io/canopy/blob/main/README.md) (v0.8.0) adds support for Pydantic v2. For applications depending on Pydantic v1, this is a breaking change; review the [Pydantic v1 to v2 migration guide](https://docs.pydantic.dev/latest/migration/) and make the necessary changes before upgrading. See the [Canopy SDK release notes](https://github.com/pinecone-io/canopy/releases/tag/v0.8.0) in GitHub for more details.
</Update>

<Update label="2024-02-23" description="SDK">
  The latest version of Pinecone's Python SDK (v3.1.0) adds support for [listing the IDs of records in a serverless index](/guides/manage-data/list-record-ids). You can list all records or just those with a [common ID prefix](/guides/index-data/data-modeling#use-structured-ids). See the [Python SDK release notes](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v3.1.0) in GitHub for more details.
</Update>

<Update label="2024-02-22" tags={["Docs"]}>
  Improved the docs for [setting up billing through the AWS Marketplace](/guides/organizations/manage-billing/upgrade-billing-plan) and [Google Cloud Marketplace](/guides/organizations/manage-billing/upgrade-billing-plan).
</Update>

<Update label="2024-02-16" tags={["Database"]}>
  It is now possible to convert a pod-based starter index to a serverless index. For organizations on the Starter plan, this requires upgrading to Standard or Enterprise; however, upgrading comes with \$100 in serverless credits, which will cover the cost of a converted index for some time.
</Update>

<Update label="2024-02-01" tags={["Docs"]}>
  Added a [Llamaindex integration guide](/integrations/llamaindex) on building a RAG pipeline with LlamaIndex and Pinecone.
</Update>

## January 2024

<Update label="2024-01-31" tags={["SDK"]}>
  The latest version of the [Canopy SDK](https://github.com/pinecone-io/canopy/blob/main/README.md) (v0.6.0) adds support for the new API mentioned above as well as namespaces, LLMs that do not have function calling functionality for query generation, and more. See the [release notes](https://github.com/pinecone-io/canopy/releases/tag/v0.6.0) in GitHub for more details.
</Update>

<Update label="2024-01-09" tags={["SDK"]}>
  The latest versions of Pinecone's Python SDK (v3.0.0) and Node.js SDK (v2.0.0) support the new API. To use the new API, existing users must upgrade to the new client versions and adapt some code. For guidance, see the [Python SDK v3 migration guide](https://canyon-quilt-082.notion.site/Pinecone-Python-SDK-v3-0-0-Migration-Guide-056d3897d7634bf7be399676a4757c7b) and [Node.js SDK v2 migration guide](https://github.com/pinecone-io/pinecone-ts-client/blob/main/v2-migration.md).
</Update>

<Update label="2024-01-09" tags={["Docs"]}>
  The Pinecone documentation is now versioned. The default "latest" version reflects the new Pinecone API. The "legacy" version reflects the previous API, which requires regional URLs for control plane operations and does not support serverless indexes.
</Update>

<Update label="2024-01-09" tags={["API"]}>
  The [new Pinecone API](/reference/api) gives you the same great vector database but with a drastically improved developer experience. The most significant improvements include:

  * [Serverless indexes](/guides/index-data/indexing-overview): With serverless indexes, you don't configure or manage compute and storage resources. You just load your data and your indexes scale automatically based on usage. Likewise, you don't pay for dedicated resources that may sometimes lay idle. Instead, the pricing model for serverless indexes is consumption-based: You pay only for the amount of data stored and operations performed, with no minimums.

  * [Multi-region projects](/guides/projects/understanding-projects): Instead of choosing a cloud region for an entire project, you now [choose a region for each index](/guides/index-data/create-an-index#create-a-serverless-index) in a project. This makes it possible to consolidate related indexes in the same project, even when they are hosted in different regions.

  * [Global URL for control plane operations](/reference): Control plane operations like `create_index`, `describe_index`, and `list_indexes` now use a single global URL, `https://api.pinecone.io`, regardless of the cloud environment where an index is hosted. This simplifies the experience compared to the legacy API, where each environment has a unique URL.
</Update>

---

# Source: https://docs.pinecone.io/release-notes/2025.md

# Source: https://docs.pinecone.io/assistant-release-notes/2025.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# 2025 releases

## December 2025

<Update label="2025-12-29" tags={["Assistant"]}>
  ### Increased metadata limit for assistants

  The metadata field limit for assistants has been increased from 1KB to 16KB.

  Assistant metadata is a JSON object that you can use to store custom organizational data, tags, and attributes for your assistants. You can specify metadata when creating an assistant by including a `metadata` field in your request, or update it later using the update assistant endpoint.

  For more information, see [Create an assistant](/reference/api/latest/assistant/create_assistant) and [Manage assistants](/guides/assistant/manage-assistants).
</Update>

<Update label="2025-12-23" tags={["Docs"]}>
  ### Test Pinecone at scale

  Added a new guide to help you [test Pinecone at production scale](/guides/get-started/test-at-scale). The guide describes how to create an index, import 10 million vectors, and then use Vector Search Bench (VSB) to capture performance metrics for 100,000 queries. To run this test, consider signing up for a [Standard plan trial](/guides/organizations/manage-billing/standard-trial).
</Update>

<Update label="2025-12-22" tags={["Assistant"]}>
  ### Pinecone Assistant now supports GPT-5

  Pinecone Assistant now supports the GPT-5 model. To use it, set `model` to `gpt-5` when [chatting with your assistant](/reference/api/latest/assistant/chat_assistant).

  For more information, see [Chat with Assistant](/guides/assistant/chat-with-assistant#choose-a-model) and [Chat through the OpenAI-compatible interface](/guides/assistant/chat-through-the-openai-compatible-interface#choose-a-model).
</Update>

<Update label="2025-12-12" tags={["General"]}>
  ### Upgrade from Starter to Standard trial

  Organizations on the Starter plan can now upgrade to a Standard plan trial at any time. The Standard trial provides 21 days and \$300 in credits to test Pinecone at scale, with access to Standard plan features such as higher limits. For more information, see [Standard plan trial](/guides/organizations/manage-billing/standard-trial).
</Update>

<Update label="2025-12-07" tags={["Assistant"]}>
  ### Delete Assistant files while they are still processing

  You can now delete a file uploaded to an assistant while it is still processing. Previously, you had to wait for file processing to complete before you could delete a file.

  For more information, see [Manage files](/guides/assistant/manage-files).
</Update>

<Update label="2025-12-02" tags={["General"]}>
  ### Annual commit discount

  In the Pinecone console, customers on Standard and Enterprise pay-as-you-go plans can now commit to an annual contract and receive a discount on usage. For more information, see [Understanding cost](/guides/manage-cost/understanding-cost#discounts).
</Update>

<Update label="2025-12-02" tags={["Assistant"]}>
  ### Increased instructions limit for assistants

  The maximum size for assistant instructions has been increased from 8 KB to 16 KB.

  Instructions are included in every chat API call. Longer instructions increase input token costs for each request and consume more of the LLM's context window, reducing available space for retrieved context and conversation history.

  For more information, see [Create an assistant](/reference/api/latest/assistant/create_assistant) and [Manage assistants](/guides/assistant/manage-assistants).
</Update>

<Update label="2025-12-01" tags={["API"]}>
  ### Dedicated Read Nodes: now in public preview

  [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) is now in [public preview](/release-notes/feature-availability). Users on Standard and Enterprise plans can now create Dedicated Read Nodes indexes using the console and API.

  With Dedicated Read Nodes, you can provision read hardware for large, high-throughput indexes that require predictable, low latency.

  For more information, see [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes).
</Update>

## November 2025

<Update label="2025-11-19" tags={["API"]}>
  ### Pinecone API version `2025-10` is now the latest stable version

  `2025-10` is now the latest [stable version](/reference/api/versioning#release-schedule) of the [Pinecone APIs](/reference/api/introduction). This version does not include any breaking changes, and provides the following new features:

  **[Dedicated read nodes](/guides/index-data/dedicated-read-nodes) (early access):**

  * The [Create an index](/reference/api/2025-10/control-plane/create_index) and [Configure index](/reference/api/2025-10/control-plane/configure_index) endpoints allow you to create and configure indexes that use dedicated read nodes. With dedicated read nodes, you can allocate dedicated hardware for read operations. This is useful for large, high-QPS indexes that require consistent, predictable low latency.
  * The [Describe an index](/reference/api/2025-10/control-plane/describe_index) endpoint now allows you to view information about a dedicated index: shards, replicas, and scaling status.

  **Namespace and metadata schema management:**

  * The [Create a namespace](/reference/api/2025-10/data-plane/createnamespace) endpoint allows you to create a namespace without upserting vectors. You can optionally configure a metadata schema when creating the namespace to pre-declare which metadata fields should be indexed for filtering.
  * The [List namespaces](/reference/api/2025-10/data-plane/listnamespaces) endpoint now supports filtering namespaces by prefix and returns the total count of matching namespaces.
  * The [Describe a namespace](/reference/api/2025-10/data-plane/describenamespace) endpoint allows you to view the metadata schema configuration for a namespace, including which fields are indexed for filtering.
  * The [Create an index](/reference/api/2025-10/control-plane/create_index) endpoint now allows you to specify a metadata schema at index creation time to pre-declare which metadata fields should be indexed for filtering across all namespaces.

  **Update and fetch by metadata:**

  * The [Update a vector](/reference/api/2025-10/data-plane/update) endpoint allows you to update metadata across multiple records in a namespace using a metadata filter expression, eliminating the need to update records individually by ID.
  * The [Fetch vectors by metadata](/reference/api/2025-10/data-plane/fetch_by_metadata) endpoint allows you to fetch vectors using metadata filters without knowing their vector IDs.

  **Enhanced sparse search:**

  * The [Search with text](/reference/api/2025-10/data-plane/search_records) endpoint now supports a `match_terms` parameter that allows you to specify terms that must be present in search results for sparse indexes.
</Update>

<Update label="2025-11-03" tags={["Assistant"]}>
  ### n8n quickstarts

  Added new quickstart options to create an n8n workflow that downloads files via HTTP and lets you chat with them using Pinecone and OpenAI.

  * [n8n quickstart for Pinecone Assistant](/guides/assistant/quickstart/n8n-quickstart)
  * [n8n quickstart for Pinecone Database](/guides/get-started/quickstart#n8n)
</Update>

<Update label="2025-11-03" tags={["SDK"]}>
  ### Released Node.js SDK v6.1.3

  Released [`v6.1.3`](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/v6.1.3) of the [Pinecone Node.js SDK](/reference/sdks/node/overview).

  This version of the SDK fixes a bug in `Assistant.listFiles()`. Previously, when passing a `filter` to `listFiles`, the top-level `metadata` object was not handled correctly. This caused the method to return all files, regardless of the filter.

  It's no longer necessary to provide a top-level `metadata` object. Instead, declare metadata fields directly in the `filter` object:

  * ✅ `const files = await assistant.listFiles({ filter: { document_type: 'manuscript' } });`
  * ❌ `const files = await assistant.listFiles({ filter: { metadata: { document_type: 'manuscript' } } });`

  If you're using the old syntax, update it so that your filter works correctly. For more information about listing the files associated with an assistant, see [Manage files](/guides/assistant/manage-files).
</Update>

## October 2025

<Update label="2025-10-30" tags={["Assistant"]}>
  ### Increased files per assistant on the Starter plan

  On the Starter plan, you can now upload up to 100 files to an assistant. Previously, the limit was 10 files.

  To learn more, see [Assistant limits](/guides/assistant/pricing-and-limits#assistant-limits).
</Update>

<Update label="2025-10-23" tags={["General"]}>
  ### Enhanced monthly spend alerts

  You can now set multiple spend alerts to monitor your organization's monthly spending. These alerts notify designated recipients when spending reaches specified thresholds. The alerts automatically reset at the start of each monthly billing cycle.

  Additionally, to protect from unexpected cost increases, Pinecone sends an alert when spending exceeds double your previous month's invoice amount. While the alert threshold is fixed, you can modify which email addresses receive the alert and enable or disable the alert notifications.

  To learn more, see [Manage cost](/guides/manage-cost/manage-cost).
</Update>

<Update label="2025-10-16" tags={["Docs"]}>
  ### Agentic quickstart

  Added new [agentic quickstart](/guides/get-started/quickstart#cursor) options to help you build Pinecone applications with AI coding agents like Claude Code and Cursor. Instead of copying code snippets, you work with an agent that understands Pinecone APIs and implements production-ready patterns automatically.
</Update>

<Update label="2025-10-16" tags={["Docs"]}>
  ### AI Engine integration

  Added the [AI Engine](/integrations/ai-engine) integration page.
</Update>

<Update label="2025-10-09" tags={["CLI"]}>
  ### Pinecone CLI v0.1.0

  We've released [v0.1.0](https://github.com/pinecone-io/cli/releases/tag/v0.1.0) of the [Pinecone CLI](https://github.com/pinecone-io/cli). The CLI lets you manage Pinecone infrastructure (organizations, projects, indexes, and API keys) directly from your terminal and in CI/CD.

  This feature is in [public preview](/release-notes/feature-availability). We'll be adding more features to the CLI over time, and we'd love your [feedback](https://community.pinecone.io/) on this early version.

  For more information, see the [CLI overview](/reference/cli/overview).
</Update>

## September 2025

<Update label="2025-09-30" tags={["Docs"]}>
  ### Production best practices

  Added a new [error handling guide](/guides/production/error-handling) to help you handle errors gracefully in production, including implementing retry logic with exponential backoff for rate limits and transient errors.

  Updated the [production checklist](/guides/production/production-checklist) with enhanced guidance on data modeling, database limits, and performance optimization.
</Update>

<Update label="2025-09-11" tags={["Docs"]}>
  ### Changing payment methods

  Added a new guide to help customers [change their payment method](/guides/organizations/manage-billing/change-payment-method) for Pinecone's Standard or Enterprise plan, including switching from credit card to marketplace billing and vice versa.
</Update>

<Update label="2025-09-04" tags={["Database"]}>
  ### Released Pinecone Terraform Provider v2.0.0

  Released [v2.0.0](https://github.com/pinecone-io/terraform-provider-pinecone/releases/tag/v2.0.0) of the [Terraform Provider for Pinecone](/integrations/terraform). This version adds support for managing API keys and projects.
</Update>

<Update label="2025-09-02" tags={["Assistant"]}>
  ### Multimodal context for assistants

  Assistants can now gather context from images in PDF files. To learn more, see [Multimodal context for assistants](/guides/assistant/multimodal). This feature is in [public preview](/release-notes/feature-availability).
</Update>

## August 2025

<Update label="2025-08-29" tags={["Database"]}>
  ### Filter lexical search by required terms

  You can now filter lexical search results to require specific terms. This is especially useful for filtering out results that don't contain essential keywords, requiring domain-specific terminology in results, and ensuring specific people, places, or things are mentioned. This feature is in [public preview](/release-notes/feature-availability).

  To learn more, see [Filter by required terms](/guides/search/lexical-search#filter-by-required-terms).
</Update>

<Update label="2025-08-26" tags={["Docs"]}>
  ### Zapier integration

  Added the [Zapier](/integrations/zapier) integration page.
</Update>

<Update label="2025-08-25" tags={["General"]}>
  ### SSO setup improvements

  We've streamlined the SSO setup process, eliminating the need to add placeholder URLs to your identity provider. To learn more, see [Configure SSO with Okta](/guides/production/configure-single-sign-on/okta).
</Update>

<Update label="2025-08-25" tags={["Database"]}>
  ### Update metadata across multiple records

  You can now [update metadata across multiple records](/guides/manage-data/update-data#update-metadata-across-multiple-records) in a namespace. This feature is in [early access](/release-notes/feature-availability).
</Update>

<Update label="2025-08-13" tags={["Database"]}>
  ### Data import from Azure Blob Storage

  Now, you can import data from an Azure Blob Storage container into a Pinecone index. This feature is in [public preview](/release-notes/feature-availability).

  To learn more, read:

  * [Integrate with Azure Blob Storage](/guides/operations/integrations/integrate-with-azure-blob-storage)
  * [Import records](/guides/index-data/import-data)
  * [Pinecone's pricing](https://www.pinecone.io/pricing/)
</Update>

<Update label="2025-08-12" tags={["Assistant"]}>
  ### Assistant MCP server endpoint update

  **Breaking Change**: After August 31, 2025 at 11:59:59 PM UTC, the SSE-based MCP endpoint for assistants (`/mcp/assistants/<YOUR_ASSISTANT_NAME>/sse`) will no longer work.

  Before then, update your applications to use the streamable HTTP transport MCP endpoint (`/mcp/assistants/<YOUR_ASSISTANT_NAME>`). This endpoint follows the current [MCP protocol specification](https://modelcontextprotocol.io/specification/2025-06-18/basic/transports#streamable-http) and provides improved flexibility and compatibility.

  Please note that Assistant MCP servers are in [early access](/release-notes/feature-availability) and are not intended for production usage.

  For more information, see [Use an Assistant MCP server](/guides/assistant/mcp-server).
</Update>

<Update label="2025-08-04" tags={["Docs"]}>
  ### VoltAgent integration

  Added the [VoltAgent](/integrations/voltagent) integration page.
</Update>

## July 2025

<Update label="2025-07-28" tags={["Database"]}>
  ### Increased context window for `pinecone-sparse-english-v0`

  You can now raise the context window for Pinecone's hosted [`pinecone-sparse-english-v0`](/guides/index-data/create-an-index#pinecone-sparse-english-v0) embedding model from `512` to `2048` using the `max_tokens_per_sequence` parameter.
</Update>

<Update label="2025-07-23" tags={["SDK"]}>
  ### Release Go SDK v4.1.0, v4.1.1, and v4.1.2

  Released [`v4.1.0`](https://github.com/pinecone-io/go-pinecone/releases/tag/v4.1.0), [`v4.1.1`](https://github.com/pinecone-io/go-pinecone/releases/tag/v4.1.1), and [`v4.1.2`](https://github.com/pinecone-io/go-pinecone/releases/tag/v4.1.2) of the [Pinecone Go SDK](/reference/sdks/go/overview).

  * `v4.1.0` adds support for admin API operations for working with API keys, projects, and service accounts.
  * `v4.1.1` adds `PercentComplete` and `RecordsImported` to the response when [describing an import](/guides/index-data/import-data#track-import-progress) and [listing imports](/guides/index-data/import-data#list-imports).
  * `v4.1.2` adds support for [migrating a pod-based index to serverless](/guides/indexes/pods/migrate-a-pod-based-index-to-serverless#3-start-migration).
</Update>

<Update label="2025-07-23" tags={["SDK"]}>
  ### Release Node.js SDK v6.1.2

  Released [`v6.1.2`](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/v6.1.2) of the [Pinecone Node.js SDK](/reference/sdks/node/overview). This version adds support for the following:

  * [Migrating a pod-based index to serverless](/guides/indexes/pods/migrate-a-pod-based-index-to-serverless#3-start-migration).
  * Controlling whether `signed_url` is included in the response when [describing a file](/guides/assistant/manage-files#get-the-status-of-a-file) for an assistant.
</Update>

## June 2025

<Update label="2025-06-26" tags={["Assistant"]}>
  ### Unlimited assistant file storage for paid plans

  Organizations on the [Standard and Enterprise plans](https://www.pinecone.io/pricing/) now have [unlimited file storage](/reference/api/assistant/assistant-limits) for their assistants. Previously, organizations on these plans were limited to 10 GB of file storage per project.
</Update>

<Update label="2025-06-23" tags={["Database"]}>
  ### Data import from Google Cloud Storage

  You can now [import data](/guides/index-data/import-data) into an index from [Google Cloud Storage](/guides/operations/integrations/integrate-with-google-cloud-storage). This feature is in [public preview](/release-notes/feature-availability).
</Update>

<Update label="2025-06-20" tags={["SDK"]}>
  ### Released Python SDK v7.1.0, v7.2.0, and v7.3.0

  Released [`v7.1.0`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v7.1.0), [`v7.2.0`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v7.2.0), and [`v7.3.0`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v7.3.0) of the [Pinecone Python SDK](/reference/sdks/python/overview).

  * `v7.1.0` fixes minor bugs.
  * `v7.2.0` adds support for [managing namespaces](/guides/manage-data/manage-namespaces).
  * `v7.3.0` adds support for admin API operations for working with API keys, projects, and service accounts.
</Update>

<Update label="2025-06-16" tags={["SDK"]}>
  ### Released Go SDK v4.0.0 and v4.0.1

  Released [`v4.0.0`](https://github.com/pinecone-io/go-pinecone/releases/tag/v4.0.0) and [`v4.0.1`](https://github.com/pinecone-io/go-pinecone/releases/tag/v4.0.1) of the [Pinecone Go SDK](/reference/sdks/go/overview).

  Go SDK `v4.0.0` uses the latest stable API version, `2025-04`, and includes support for the following:

  * [Managing namespaces](/guides/manage-data/manage-namespaces)
  * [Reusing an index connection with a new namespace](/guides/manage-data/target-an-index#target-by-index-host-recommended) (see the Go example)
  * [Creating and managing backups](/guides/manage-data/back-up-an-index)
  * [Restoring indexes from backups](/guides/manage-data/restore-an-index)
  * [Listing embedding and reranking models hosted by Pinecone](/reference/api/2025-04/inference/list_models)
  * [Getting details about a model hosted by Pinecone](/reference/api/2025-04/inference/describe_model)

  Go SDK `v4.0.1` expands the [`DescribeIndex`](/guides/production/connect-to-aws-privatelink#read-and-write-data) response to include the `private_host` value for connecting to indexes with a private endpoint.
</Update>

<Update label="2025-06-16" tags={["SDK"]}>
  ### Released Node.js SDK v6.1.1

  Released [`v6.1.1`](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/v6.1.1) of the [Pinecone Node.js SDK](/reference/sdks/node/overview). This version adds support for [setting the sampling temperature](/guides/assistant/chat-with-assistant#set-the-sampling-temperature) for an assistant, and expands the [`describeIndex`](/guides/production/connect-to-aws-privatelink#read-and-write-data) response to include the `private_host` value for connecting to indexes with a private endpoint.
</Update>

<Update label="2025-06-12" tags={["Docs"]}>
  ### Data modeling guide

  Added a new guide to help you [model your data](/guides/index-data/data-modeling) for efficient ingestion, retrieval, and management in Pinecone.
</Update>

<Update label="2025-06-05" tags={["SDK"]}>
  ### Released Java SDK v5.1.0

  Released [`v5.1.0`](https://github.com/pinecone-io/pinecone-java-client/releases/tag/v5.1.0) of the [Pinecone Java SDK](/reference/sdks/java/overview). This version adds support for [listing](/reference/api/2025-04/inference/list_models) and [describing](/reference/api/2025-04/inference/describe_model) embedding and reranking models hosted by Pinecone.
</Update>

<Update label="2025-06-05" tags={["SDK"]}>
  ### Released Node.js SDK v6.1.0

  Released [`v6.1.0`](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/v6.1.0) of the [Pinecone Node.js SDK](/reference/sdks/node/overview). This version adds support for [controlling the context snippets sent to the LLM](/guides/assistant/chat-with-assistant#control-the-context-snippets-sent-to-the-llm) by an assistant.
</Update>

## May 2025

<Update label="2025-05-29" tags={["SDK"]}>
  ### Released Python SDK v7.0.1 and v7.0.2

  Released [`v7.0.1`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v7.0.1) and [`v7.0.2`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v7.0.2) of the [Pinecone Python SDK](/reference/sdks/python/overview). These versions fix minor bugs discovered since the release of the `v7.0.0` major version.
</Update>

<Update label="2025-05-29" tags={["SDK"]}>
  ### Released Node.s SDK v6.0.1

  Released [`v6.0.1`](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/v6.0.1) of the [Pinecone Node.js SDK](/reference/sdks/node/overview). This version adds pagination to the [`listBackups`](/guides/manage-data/back-up-an-index#list-backups-in-a-project) operation.
</Update>

<Update label="2025-05-19" tags={["API"]}>
  ### Pinecone API version `2025-04` is now the latest stable version

  `2025-04` is now the latest [stable version](/reference/api/versioning#release-schedule) of the [Pinecone APIs](/reference/api/introduction). For highlights, see the SDK releases below.
</Update>

<Update label="2025-05-19" tags={["SDK"]}>
  ### Released Python SDK v7.0.0

  Released [`v7.0.0`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v7.0.0) of the [Pinecone Python SDK](/reference/sdks/python/overview). This version uses the latest stable API version, `2025-04`, and includes support for the following:

  * [Creating and managing backups](/guides/manage-data/back-up-an-index)
  * [Restoring indexes from backups](/guides/manage-data/restore-an-index)
  * [Listing embedding and reranking models hosted by Pinecone](/reference/api/2025-04/inference/list_models)
  * [Getting details about a model hosted by Pinecone](/reference/api/2025-04/inference/describe_model)
  * [Creating a BYOC index](/guides/production/bring-your-own-cloud#create-an-index)

  Additionally, the `pinecone-plugin-assistant` package required to work with [Pinecone Assistant](/guides/assistant/overview) is now included by default; it is no longer necessary to install the plugin separately.
</Update>

<Update label="2025-05-19" tags={["SDK"]}>
  ### Released Node.js SDK v6.0.0

  Released [`v6.0.0`](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/v6.0.0) of the [Pinecone Node.js SDK](/reference/sdks/node/overview). This version uses the latest stable API version, `2025-04`, and includes support for the following:

  * [Managing namespaces](/guides/manage-data/manage-namespaces)
  * [Creating and managing backups](/guides/manage-data/back-up-an-index)
  * [Restoring indexes from backups](/guides/manage-data/restore-an-index)
  * [Listing embedding and reranking models hosted by Pinecone](/reference/api/2025-04/inference/list_models)
  * [Getting details about a model hosted by Pinecone](/reference/api/2025-04/inference/describe_model)
</Update>

<Update label="2025-05-19" tags={["SDK"]}>
  ### Released Java SDK v5.0.0

  Released [`v5.0.0`](https://github.com/pinecone-io/pinecone-java-client/releases/tag/v5.0.0) of the [Pinecone Java SDK](/reference/sdks/java/overview). This version uses the latest stable API version, `2025-04`, and includes support for the following:

  * [Creating indexes with integrated embedding and reranking](/guides/index-data/indexing-overview#integrated-embedding)
  * [Upserting text to an integrated index](/guides/index-data/upsert-data)
  * [Searching an integrated index with text](/guides/search/semantic-search#search-with-text)
  * [Managing namespaces](/guides/manage-data/manage-namespaces)
  * [Creating and managing backups](/guides/manage-data/back-up-an-index)
  * [Restoring indexes from backups](/guides/manage-data/restore-an-index)
</Update>

<Update label="2025-05-19" tags={["SDK"]}>
  ### Released .NET SDK v4.0.0

  Released [`v4.0.0`](https://github.com/pinecone-io/pinecone-dotnet-client/releases/tag/4.0.0) of the [Pinecone .NET SDK](/reference/sdks/dotnet/overview). This version uses the latest stable API version, `2025-04`, and includes support for the following:

  * [Creating indexes with integrated embedding and reranking](/guides/index-data/indexing-overview#integrated-embedding)
  * [Upserting text to an integrated index](/guides/index-data/upsert-data)
  * [Searching an integrated index with text](/guides/search/semantic-search#search-with-text)
  * [Managing namespaces](/guides/manage-data/manage-namespaces)
  * [Creating and managing backups](/guides/manage-data/back-up-an-index)
  * [Restoring indexes from backups](/guides/manage-data/restore-an-index)
  * [Listing embedding and reranking models hosted by Pinecone](/reference/api/2025-04/inference/list_models)
  * [Getting details about a model hosted by Pinecone](/reference/api/2025-04/inference/describe_model)

  <Warning>
    Before upgrading to `v4.0.0`, update all relevant code to account for the following [breaking changes](/reference/api/versioning#breaking-changes). See the [`v4.0.0`](https://github.com/pinecone-io/pinecone-dotnet-client/releases/tag/4.0.0) release notes for full details.

    * The [`create_index`](/reference/api/2025-04/control-plane/create_index) and [`create_for_model`](/reference/api/2025-04/control-plane/create_for_model) operations:
      * `CreateIndexRequestMetric` has been renamed to `MetricType`.
    * The [`list_indexes`](/reference/api/2025-04/control-plane/list_indexes) operation:
      * `ModelIndexEmbedMetric` has been renamed to `MetricType`.
    * The [`embed`](/reference/api/2025-04/inference/generate-embeddings) operation:
      * `SparseEmbedding.SparseIndices` has changed from `IEnumerable<int>` to `IEnumerable<long>`.
  </Warning>
</Update>

<Update label="2025-05-06" tags={["Docs"]}>
  ### New Docs IA

  We've overhauled the information architecture of our guides to mirror the goals of users, from indexing to searching to optimizing to production.

  This change includes distinct pages for search types:

  * [Semantic search](https://docs.pinecone.io/guides/search/semantic-search)
  * [Lexical search](https://docs.pinecone.io/guides/search/lexical-search)
  * [Hybrid search](https://docs.pinecone.io/guides/search/hybrid-search)

  And optimization techniques:

  * [Increase relevance](https://docs.pinecone.io/guides/optimize/increase-relevance)
  * [Increase throughput](https://docs.pinecone.io/guides/optimize/increase-throughput)
  * [Decrease latency](https://docs.pinecone.io/guides/optimize/decrease-latency)
</Update>

## April 2025

<Update label="2025-04-25" tags={["Database"]}>
  ### Bring Your Own Cloud (BYOC) in GCP

  The [Bring Your Own Cloud (BYOC)](/guides/production/bring-your-own-cloud) offering is now available in GCP. Organizations with high security and compliance requirements can use BYOC to deploy Pinecone Database in their own GCP account. This feature is in [public preview](/release-notes/feature-availability).
</Update>

<Update label="2025-04-21" tags={["Database"]}>
  ### Integrate AI agents with Pinecone MCP

  [Pinecone's open-source MCP server](/guides/operations/mcp-server) enables AI agents to interact directly with Pinecone's functionality and documentation via the standardized [Model Context Protocol (MCP)](https://modelcontextprotocol.io/l). Using the MCP server, agents can search Pinecone documentation, manage indexes, upsert data, and query indexes for relevant information.
</Update>

<Update label="2025-04-21" tags={["Assistant"]}>
  ### Add context to AI agents with Assistant MCP

  Every Pinecone Assistant now has a [dedicated MCP server](/guides/assistant/mcp-server) that gives AI agents direct access to the assistant's knowledge through the standardized [Model Context Protocol (MCP)](https://modelcontextprotocol.io/).
</Update>

<Update label="2025-04-21" tags={["Assistant"]}>
  ### Upload a file from an in-memory binary stream

  You can [upload a file to an assistant directly from an in-memory binary stream](/guides/assistant/upload-files#upload-from-a-binary-stream) using the Python SDK and the BytesIO class.
</Update>

<Update label="2025-04-21" tags={["Database"]}>
  ### Released Pinecone Terraform Provider v1.0.0

  Released [v1.0.0](https://github.com/pinecone-io/terraform-provider-pinecone/releases/tag/v1.0.0) of the [Terraform Provider for Pinecone](/integrations/terraform). This version adds support for [sparse indexes](/guides/index-data/indexing-overview#sparse-indexes), [indexes with integrated embedding and reranking](/guides/index-data/indexing-overview#integrated-embedding), [index tags](/guides/manage-data/manage-indexes#configure-index-tags), and [index deletion protection](/guides/manage-data/manage-indexes#configure-deletion-protection).
</Update>

<Update label="2025-04-13" tags={["SDK"]}>
  ### Released .NET SDK v3.1.0

  Released [`v3.1.0`](https://github.com/pinecone-io/pinecone-dotnet-client/releases/tag/3.1.0) of the [Pinecone .NET SDK](/reference/sdks/dotnet/overview). This version adds support for [indexes with integrated embedding and reranking](/guides/index-data/indexing-overview#integrated-embedding).
</Update>

<Update label="2025-04-04" tags={["Docs"]}>
  ### LLM shortcuts for Pinecone docs

  You can now use the "Copy page" options at the top of every page of the Pinecone documentation to quickly ground LLMs with Pinecone-specific context.
</Update>

## March 2025

<Update label="2025-03-26" tags={["Assistant"]}>
  ### Control the context snippets the assistant sends to the LLM

  You can [control the context snippets sent to the LLM](/guides/assistant/chat-with-assistant#control-the-context-snippets-sent-to-the-llm) by setting `context_options` in the request.
</Update>

<Update label="2025-03-24" tags={["SDK"]}>
  ### Released Go SDK v3.1.0

  Released [`v3.1.0`](https://github.com/pinecone-io/go-pinecone/releases/tag/v3.1.0) of the [Pinecone Go SDK](/reference/sdks/go/overview). This version adds support for [indexes with integrated embedding and reranking](/guides/index-data/indexing-overview#integrated-embedding).
</Update>

<Update label="2025-03-21" tags={["General"]}>
  <img className="block max-w-full" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week-march-2025.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=a5a2d145f08f688f6ae41242a0b5ff45" data-og-width="2400" width="2400" data-og-height="1260" height="1260" data-path="images/release-notes/launch-week-march-2025.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week-march-2025.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=027b2cf2bccfc3cd36eb24b22a192199 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week-march-2025.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=ddedc6d1859e46a1e5d46351b1d58331 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week-march-2025.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=99024d848edae3477f5ce28037ea1d68 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week-march-2025.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=9b522179310daff93683d900a63b05ee 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week-march-2025.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=b0331b0c5e94a1f07f2e1b8b7b7e9b03 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/launch-week-march-2025.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=54db6d1b12f6b6dfc588d0491c72e484 2500w" />

  ### Launch week: Dark mode

  Dark mode is now out for Pinecone's website, docs, and console. You can change your theme at the top right of each site.
</Update>

<Update label="2025-03-20" tags={["General"]}>
  ### Launch week: Self-service audit logs

  You can now enable and [configure audit logs](/guides/production/configure-audit-logs) for your Pinecone organization. [Audit logs](/guides/production/security-overview#audit-logs) provide a detailed record of user, service account, and API actions that occur within Pinecone. This feature is in [public preview](/release-notes/feature-availability) and available only on [Enterprise plans](https://www.pinecone.io/pricing/).
</Update>

<Update label="2025-03-19" tags={["General"]}>
  ### Launch week: Introducing the Admin API and service accounts

  You can now use [service accounts](/guides/organizations/understanding-organizations#service-accounts) to programmatically manage your Pinecone organization through the Admin API. Use the Admin API to [create](/guides/projects/create-a-project) and [manage projects](/guides/projects/manage-projects), as well as [create and manage API keys](/guides/projects/manage-api-keys). The Admin API and service accounts are in [public preview](/release-notes/feature-availability).
</Update>

<Update label="2025-03-18" tags={["Database"]}>
  ### Launch week: Back up an index through the API

  You can now [back up an index](/guides/manage-data/back-up-an-index) and [restore an index](/guides/manage-data/restore-an-index) through the Pinecone API. This feature is in [public preview](/release-notes/feature-availability).
</Update>

<Update label="2025-03-17" tags={["Database"]}>
  ### Launch week: Optimized database architecture

  Pinecone has optimized its [serverless database architecture](/guides/get-started/database-architecture) to meet the growing demand for large-scale agentic workloads and improved performance for search and recommendation workloads. New customers will use this architecture by default, and existing customers will gain access over the next month.
</Update>

<Update label="2025-03-13" tags={["Docs"]}>
  ### Firebase Genkit integration

  Added the [Firebase Genkit](/integrations/genkit) integration page.
</Update>

<Update label="2025-03-10" tags={["Database"]}>
  ### Bring Your Own Cloud (BYOC) in public preview

  [Bring Your Own Cloud (BYOC)](/guides/production/bring-your-own-cloud) lets you deploy Pinecone Database in your private AWS account to ensure data sovereignty and compliance, with Pinecone handling provisioning, operations, and maintenance. This feature is in [public preview](/release-notes/feature-availability) on AWS.
</Update>

## February 2025

<Update label="2025-02-21" tags={["Docs"]}>
  ### Docs site refresh

  We've refreshed the look and layout of the [Pinecone documentation](https://docs.pinecone.io) site. You can now use the dropdown at the top of the side navigation to view documentation for either [Pinecone Database](/guides/get-started/overview) or [Pinecone Assistant](/guides/assistant/overview).
</Update>

<Update label="2025-02-18" tags={["Assistant"]}>
  ### Limit the number of chunks retrieved

  You can now limit the number of chunks the reranker sends to the LLM. To do this, set the `top_k` parameter (default is 15) when [retrieving context snippets](/guides/assistant/retrieve-context-snippets).
</Update>

<Update label="2025-02-21" tags={["Docs"]}>
  ### Assistant Quickstart colab notebook

  Added the [Assistant Quickstart colab notebook](https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/assistant-quickstart.ipynb). This notebook shows you how to set up and use [Pinecone Assistant](/guides/assistant/overview) in your browser.
</Update>

<Update label="2025-02-18" tags={["SDK"]}>
  ### Released Node.js SDK v5.0.0

  Released [`v5.0.0`](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/v5.0.0) of the [Pinecone Node.js SDK](/reference/sdks/node/overview). This version uses the latest stable API version, `2025-01`, and includes support for [Pinecone Assistant](/guides/assistant/overview) and [sparse-only indexes](/guides/index-data/indexing-overview#sparse-indexes).
</Update>

<Update label="2025-02-11" tags={["Docs"]}>
  ### New integrations

  Added the [Box](/integrations/box) and [Cloudera AI](/integrations/cloudera) integration pages.
</Update>

<Update label="2025-02-10" tags={["Assistant"]}>
  ### Citation highlights in assistant responses

  You can now include [highlights](/guides/assistant/chat-with-assistant#include-citation-highlights-in-the-response) in an assistant's citations. Highlights are the specific parts of the document that the assistant used to generate the response.

  Citation highlights are available in the Pinecone console or API versions `2025-04` and later.
</Update>

<Update label="2025-02-07" tags={["API"]}>
  ### Pinecone API version `2025-01` is now the latest stable version

  `2025-01` is now the latest [stable version](/reference/api/versioning#release-schedule) of the [Pinecone APIs](/reference/api/introduction).
</Update>

<Update label="2025-02-07" tags={["SDK"]}>
  ### Released Python SDK v6.0.0

  Released [`v6.0.0`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v6.0.0) of the [Pinecone Python SDK](/reference/sdks/python/overview). This version uses the latest stable API version, `2025-01`, and includes support for the following:

  * [Index tags](/guides/manage-data/manage-indexes#configure-index-tags) to categorize and identify your indexes.
  * [Integrated inference](/reference/api/introduction#inference) without the need for extra plugins. If you were using the preview functionality of integrated inference, you must uninstall the `pinecone-plugin-records` package to use the `v6.0.0` release.
  * Enum objects to help with the discoverability of some configuration options, for example, `Metric`, `AwsRegion`, `GcpRegion`, `PodType`, `EmbedModel`, `RerankModel`. This is a backwards compatible change; you can still pass string values for affected fields.
  * New client variants, `PineconeAsyncio` and `IndexAsyncio`, which provide `async` methods for use with [asyncio](https://docs.python.org/3/library/asyncio.html). This makes it possible to use Pinecone with modern async web frameworks such as [FastAPI](https://fastapi.tiangolo.com/), [Quart](https://quart.palletsprojects.com/en/latest/), and [Sanic](https://sanic.dev/en/). Async support should significantly increase the efficiency of running many upserts in parallel.

  <Warning>
    Before upgrading to `v6.0.0`, update all relevant code to account for the following [breaking changes](/reference/api/versioning#breaking-changes). See the [`v6.0.0`](https://github.com/pinecone-io/pinecone-python-client/releases/tag/v6.0.0) release notes for full details.

    * Incorporated the `pinecone-plugin-records` and `pinecone-plugin-inference` plugins into the `pinecone` package. If you are using these plugins, you must unstall them to use `v6.0.0`.
    * Dropped support for Python 3.8, which has now reached official end of life, and added support for Python 3.13.
    * Removed the explicit dependency on `tqdm`, which is used to provide a progress bar when upserting data into Pinecone. If `tqdm` is available in the environment, the Pinecone SDK will detect and use it, but `tdqm` is no longer required to run the SDK. Popular notebook platforms such as [Jupyter](https://jupyter.org/) and [Google Colab](https://colab.google/) already include `tqdm` in the environment by default, but if you are running small scripts in other environments and want to continue seeing progress bars, you will need to separately install the `tqdm` package.
    * Removed some previously deprecated and rarely used keyword arguments (`config`, `openapi_config`, and `index_api`) to instead prefer dedicated keyword arguments for individual settings such as `api_key`, `proxy_url`, etc.
  </Warning>
</Update>

<Update label="2025-02-07" tags={["SDK"]}>
  ### Released Java SDK v4.0.0

  Released [`v4.0.0`](https://github.com/pinecone-io/pinecone-java-client/releases/tag/v4.0.0) of the [Pinecone Java SDK](/reference/sdks/java/overview). This version uses the latest stable API version, `2025-01`, and adds support for [sparse-only indexes](/guides/index-data/indexing-overview#sparse-indexes).

  <Warning>
    Before upgrading to `v4.0.0`, update all relevant code to account for the following [breaking changes](/reference/api/versioning#breaking-changes). See the [`v4.0.0`](https://github.com/pinecone-io/pinecone-java-client/releases/tag/v4.0.0) release notes for full details.

    * [`embed` method](/reference/api/2025-01/inference/generate-embeddings):
      * `parameters` now accepts `Map<String, Object>` instead of `EmbedRequestParameters`.
      * The `Embeddings` response class now has dense and sparse embeddings. You now must use `getDenseEmbedding()` or `getSparseEmbedding()`. For example, instead of `embeddings.getData().get(0).getValues()`, you would use `embeddings.getData().get(0).getDenseEmbedding().getValues()`.

    * [`rerank` method](/guides/search/rerank-results):
      * `documents` now accepts `List<Map<String, Object>>` instead of `List<Map<String, String>>`.
      * `parameters` now accepts `Map<String, Object>` instead of `Map<String, String>`.
  </Warning>
</Update>

<Update label="2025-02-07" tags={["SDK"]}>
  ### Released Go SDK v3.0.0

  Released [`v3.0.0`](https://github.com/pinecone-io/go-pinecone/releases/tag/v3.0.0) of the [Pinecone Go SDK](/reference/sdks/go/overview). This version uses the latest stable API version, `2025-01`, and adds support for [sparse-only indexes](/guides/index-data/indexing-overview#sparse-indexes).

  <Warning>
    Before upgrading to `v3.0.0`, update all relevant code to account for the following [breaking changes](/reference/api/versioning#breaking-changes). See the [`v3.0.0`](https://github.com/pinecone-io/go-pinecone/releases/tag/v3.0.0) release notes for full details.

    * [`embed` operation](/reference/api/2025-01/inference/generate-embeddings):
      * `EmbedParameters` is no longer typed as a pointer.
    * [`create_index` operation](/guides/index-data/create-an-index):
      * `CreateServerlessIndexRequest` and `CreatePodIndexRequest` structs have been updated, and fields are now classified as pointers to better denote optionality around creating specific types of indexes: `Metric`, `Dimension`, `VectorType`, and `DeletionProtection`.
    * Various data operation:
      * `Values` in the `Vector` type are now a pointer to allow flexibility when working with sparse-only indexes.
  </Warning>
</Update>

<Update label="2025-02-07" tags={["SDK"]}>
  ### Released .NET SDK v3.0.0

  Released [`v3.0.0`](https://github.com/pinecone-io/pinecone-dotnet-client/releases/tag/3.0.0) of the [Pinecone .NET SDK](/reference/sdks/dotnet/overview). This version uses the latest stable API version, `2025-01`, and adds support for [sparse-only indexes](/guides/index-data/indexing-overview#sparse-indexes).

  <Warning>
    Before upgrading to `v3.0.0`, update all relevant code to account for the following [breaking changes](/reference/api/versioning#breaking-changes). See the [`v3.0.0`](https://github.com/pinecone-io/pinecone-dotnet-client/releases/tag/3.0.0) release notes for full details.

    * [`embed` operation](/reference/api/2025-01/inference/generate-embeddings):
      * The `Embedding` type has changed from a simple object to a discriminated union, supporting both `DenseEmbedding` and `SparseEmbedding`. New helper methods available on the Embedding type: `IsDense` and `IsSparse` for type checking, `AsDense()` and `AsSparse()` for type conversion, and `Match()` and `Visit()` for pattern matching.
      * The `Parameters` property now uses `Dictionary<string, object?>?` instead of `EmbedRequestParameters`.

    * `rerank` operation:
      * The `Document` property now uses `Dictionary<string, object?>?` instead of `Dictionary<string, string>?`.
      * The `Parameters` property now uses `Dictionary<string, object?>?` instead of `Dictionary<string, string>?`.
  </Warning>
</Update>

## January 2025

<Update label="2025-01-29" tags={["General"]}>
  ### Update to the API keys page

  Added the **Created by** column on the [API keys page](https://app.pinecone.io/organizations/-/projects/-/keys) in the Pinecone Console. This column shows the email of the user who created the API key.
</Update>

<Update label="2025-01-29" tags={["Database"]}>
  ### Sparse-only indexes in early access

  You can now use [sparse-only indexes](/guides/index-data/indexing-overview#sparse-indexes) for the storage and retrieval of sparse vectors. This feature is in [early access](/release-notes/feature-availability).
</Update>

<Update label="2025-01-22" tags={["Assistant"]}>
  <img className="block max-w-full" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/assistant-ga.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=aa42277fa4b0cb3fafd1544c9b65ee24" data-og-width="1774" width="1774" data-og-height="994" height="994" data-path="images/release-notes/assistant-ga.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/assistant-ga.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=62a2e6dcf304bc790c6d14098402592d 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/assistant-ga.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=533b1a0ceed95b2f6def26f206cea23a 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/assistant-ga.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=ad26bef0b1c7d1ed25535bb7d460883c 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/assistant-ga.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=0a04a321697583cf350ba392aad30858 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/assistant-ga.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=e4ce39f7e891321f28f44b446e87c47c 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/release-notes/assistant-ga.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=325c235b9a924917cf64afc7c3330f9e 2500w" />

  Pinecone Assistant is generally available (GA) for all users.

  [Read more](https://www.pinecone.io/blog/pinecone-assistant-generally-available) about the release on our blog.
</Update>

<Update label="2025-01-09" tags={["SDK"]}>
  ### Released Node SDK v4.1.0

  Released [`v4.1.0`](https://github.com/pinecone-io/pinecone-ts-client/releases/tag/4.1.0) of the [Pinecone Node.js SDK](/reference/sdks/node/overview). This version adds support for [index tags](/guides/manage-data/manage-indexes#configure-index-tags) when creating or configuring indexes. It also adds a new `RetryOnServerFailure` class that automatically retries asynchronous operations with exponential backoff when the server responds with a `500` or `503` [error](/reference/api/errors).
</Update>

<Update label="2025-01-09" tags={["General"]}>
  ### New Billing Admin user role

  Added the Billing Admin [user role](/guides/organizations/understanding-organizations#organization-roles). Billing Admins have permissions to view billing details, usage details, and support plans.
</Update>

<Update label="2025-01-07" tags={["SDK"]}>
  ### Released Go SDK v2.2.0

  Released [`v2.2.0`](https://github.com/pinecone-io/go-pinecone/releases/tag/v2.2.0) of the [Pinecone Go SDK](/reference/sdks/go/overview). This version adds support for [index tags](/guides/manage-data/manage-indexes#configure-index-tags) when creating or configuring indexes.
</Update>

---

# Source: https://docs.pinecone.io/release-notes/2026.md

# Source: https://docs.pinecone.io/assistant-release-notes/2026.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# 2026 releases

## January 2026

<Update label="2026-01-28" tags={["Assistant"]}>
  ### Claude model deprecation for Assistant

  Anthropic has [deprecated](https://platform.claude.com/docs/en/about-claude/model-deprecations) the Claude 3.5 Sonnet and Claude 3.7 Sonnet models. Pinecone Assistant automatically routes all chat requests that specify `claude-3-5-sonnet` or `claude-3-7-sonnet` to Claude Sonnet 4.5, which provides enhanced intelligence at the same price. No code changes are required.

  To update your code to explicitly use Claude Sonnet 4.5, set `model: "claude-sonnet-4-5"` in your chat requests. For more information, see [Choose a model](/guides/assistant/chat-with-assistant#choose-a-model).
</Update>

<Update label="2026-01-27" tags={["Assistant"]}>
  ### Claude Sonnet 4.5 now available for Assistant chat

  Pinecone Assistant now supports Anthropic's Claude Sonnet 4.5 model. To use this model, set `model: "claude-sonnet-4-5"` in your [chat](/reference/api/latest/assistant/chat_assistant) requests. In the Pinecone console, Claude Sonnet 4.5 is also available as a selection in the **Chat model** dropdown menu in the playground for each assistant.

  For more information, see [Choose a model](/guides/assistant/chat-with-assistant#choose-a-model).
</Update>

<Update label="2026-01-23" tags={["API"]}>
  ### Metadata filter limit: 10,000 values per `$in`/`$nin` operator

  Pinecone now enforces a limit of 10,000 values per `$in` or `$nin` operator in metadata filter expressions. This limit helps ensure consistent query performance and protects shared infrastructure from excessive load caused by very large filters.

  Requests that exceed this limit will fail with a `400 - BAD_REQUEST` error.

  If your application currently uses large `$in` filters (especially for access control), consider these approaches:

  * **Namespace-based isolation** (recommended): Create separate namespaces for each tenant instead of filtering by thousands of tenant IDs. This can also reduce query costs (queries on a 1 GB namespace cost 1 RU instead of 100 RUs for a 100 GB namespace with filtering).
  * **Access control groups**: Filter by organization, project, or role identifiers instead of individual user IDs.
  * **Post-filter client-side**: Retrieve a larger top K without filtering, then filter results in your application.

  For more information, see [Metadata filter limits](/reference/api/database-limits#metadata-filter-limits) and [Design for multi-tenancy](/guides/index-data/data-modeling#design-for-multi-tenancy).
</Update>

<Update label="2026-01-16" tags={["API"]}>
  ### Request-per-second limits for data plane operations

  Pinecone now enforces request-per-second rate limits on data plane operations (query, upsert, delete, and update) at the namespace level. These limits are set to 100 requests per second per namespace for all plans and provide protection against excessive request rates.

  Request-per-second limits are enforced in addition to existing read unit and write unit limits. If you exceed a request-per-second limit, you'll receive a `429 - TOO_MANY_REQUESTS` error.

  For more information, see [Database limits](/reference/api/database-limits#data-plane-operations-requests-per-second-limits).
</Update>

<Update label="2026-01-15" tags={["API"]}>
  ### Pagination support for fetch by metadata

  The [Fetch by metadata](/reference/api/latest/data-plane/fetch_by_metadata) operation now supports pagination, allowing you to fetch large result sets in multiple requests. Use the `paginationToken` parameter to retrieve the next page of results.

  When there are more results available, the response includes a `pagination` object with a `next` token. Pass this token as the `paginationToken` parameter in subsequent requests to fetch the next page. When there are no more results, the response does not include a `pagination` object.

  For more information, see [Fetch records by metadata](/guides/manage-data/fetch-data#fetch-records-by-metadata).
</Update>

---

# Source: https://docs.pinecone.io/guides/organizations/manage-billing/access-your-invoices.md

# Source: https://docs.pinecone.io/guides/assistant/admin/access-your-invoices.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Access your invoices

> View and download billing invoices from Pinecone.

You can access your billing history and invoices in the Pinecone console:

1. Go to [**Settings > Billing > Overview**](https://app.pinecone.io/organizations/-/settings/billing).
2. Scroll down to the **Payment history and invoices** section.
3. For each billing period, you can download the invoice by clicking the **Download** button.

Each invoice includes line items for the services used during the billing period. If the total cost of that usage is below the monthly minimum, the invoice also includes a line item covering the rest of the minimum usage commitment.

---

# Source: https://docs.pinecone.io/integrations/ai-engine.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Engine

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

AI Engine seamlessly connects WordPress with the world's leading AI models. Create intelligent chatbots, generate content, build AI forms, and automate tasks—all from your WordPress dashboard.

With AI Engine, you can create a chatbot to assist your visitors, answer support questions, or guide users through your products and services. Need fresh content? AI Engine can write posts in your voice, help rewrite existing ones, or translate them naturally into other languages. It can also generate custom images for your articles, refine messy text, or just lend a hand when you're stuck.

For developers and power users, AI Engine offers internal APIs, shortcode flexibility, and advanced features like function calling and real-time audio chat. You can build your own AI-powered tools, automate tasks, or even create AI-driven SaaS applications on top of WordPress. And with support for a wide range of providers — OpenAI, Anthropic, Google, Hugging Face, and more — you have full control over the models you want to use.

Everything is designed to feel native to WordPress. Whether you're exploring ideas in the AI Playground, using Copilot to help in the editor, or letting an AI agent manage your content through MCP, AI Engine is built to grow with you — and shaped by real user feedback every step of the way.

<PrimarySecondaryCTA primaryHref={"https://ai.thehiddendocs.com/services/pinecone/"} primaryLabel={"Get started"} primaryTarget={"_blank"} />

---

# Source: https://docs.pinecone.io/integrations/airbyte.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Airbyte

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Airbyte offers a platform for creating ETL pipelines without writing integration code. It supports integrations with hundreds of systems, including databases, data warehouses, and SaaS products.

The Pinecone connector for Airbyte allows users to connect these systems directly into Pinecone. The connector fetches data from the connected source, embeds the data using an large language model (LLM), and then upserts it into Pinecone. From enhancing semantic search capabilities to building intelligent recommendation systems, the Pinecone Airbyte connector offers a versatile solution. By tapping into Airbyte's extensive array of source connectors, you can explore new ways to enrich your data-driven projects and achieve your specific goals.

<PrimarySecondaryCTA primaryHref={"https://docs.airbyte.com/integrations/destinations/pinecone"} primaryLabel={"Get started"} secondaryHref={"https://www.pinecone.io/learn/series/airbyte/"} secondaryLabel={"View setup guide"} />

## Related articles

* [Data Sync and Search: Pinecone and Airbyte](https://www.pinecone.io/learn/series/airbyte/)
* [Introduction to Airbyte and the Pinecone connector](https://www.pinecone.io/learn/series/airbyte/airbyte-and-pinecone-intro/)
* [Postgres to Pinecone Syncing](https://www.pinecone.io/learn/series/airbyte/airbyte-postgres-to-pinecone/)

---

# Source: https://docs.pinecone.io/integrations/amazon-bedrock.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Amazon Bedrock

> Pinecone as a Knowledge Base for Amazon Bedrock

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Users can now select Pinecone as a Knowledge Base for [Amazon Bedrock](https://aws.amazon.com/bedrock/), a fully managed service from Amazon Web Services (AWS) for building GenAI applications.

The Pinecone vector database is a key component of the AI tech stack, helping companies solve one of the biggest challenges in deploying GenAI solutions — hallucinations — by allowing them to store, search, and find the most relevant and up-to-date information from company data and send that context to Large Language Models (LLMs) with every query. This workflow is called Retrieval Augmented Generation (RAG), and with Pinecone, it aids in providing relevant, accurate, and fast responses from search or GenAI applications to end users.

With the release of Knowledge Bases for Amazon Bedrock, developers can integrate their enterprise data into Amazon Bedrock using Pinecone as the fully-managed vector database to build GenAI applications that are:

* **Highly performant:** Speed through data in milliseconds. Leverage metadata filters and sparse-dense index support for top-notch relevance, ensuring quick, accurate, and grounded results across diverse search tasks.
* **Cost effective at scale:** Start for free on the starter plan and seamlessly scale usage with transparent usage-based pricing. Add or remove resources to meet your desired capacity and performance, upwards of billions of embeddings.
* **Enterprise ready:** Launch, use, and scale your AI solution without needing to maintain infrastructure, monitor services, or troubleshoot algorithms. Pinecone meets the security and operational requirements of enterprises.

<PrimarySecondaryCTA secondaryHref={"#setup-guide"} secondaryLabel={"View setup guide"} />

## What are Agents for Amazon Bedrock?

In Bedrock, users interact with **Agents** that are capable of combining the natural language interface of the supported LLMs with those of a **Knowledge Base.** Bedrock's Knowledge Base feature uses the supported LLMs to generate **embeddings** from the original data source. These embeddings are stored in Pinecone, and the Pinecone index is used to retrieve semantically relevant content upon the user's query to the agent.

**Note:** the LLM used for embeddings may be different than the one used for the natural language generation. For example, you may choose to use Amazon Titan to generate embeddings and use Anthropic's Claude to generate natural language responses.

Additionally, Agents for Amazon Bedrock may be configured to execute various actions in the context of responding to a user's query - but we won't get into this functionality in this post.

## What is a Knowledge Base for Amazon Bedrock?

A Bedrock Knowledge base ingests raw text data or documents found in Amazon S3, embeds the content and upserts the embeddings into Pinecone. Then, a Bedrock agent can interact with the knowledge base to retrieve the most semantically relevant content given a user's query.

Overall, the Knowledge Base feature is a valuable tool for users who want to improve their AI models' performance. With Bedrock's LLMs and Pinecone, users can easily integrate their data from AWS storage solutions and enhance the accuracy and relevance of their AI models.

In this post, we'll go through the steps required for creating a Knowledge Base for Amazon Bedrock as well as an agent that will retrieve information from the knowledge base.

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/AWS_Bedrock_Specific.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=f63ffcbacba8c0fbf31eeadd78a90cb5" alt="" data-og-width="1798" width="1798" data-og-height="874" height="874" data-path="images/AWS_Bedrock_Specific.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/AWS_Bedrock_Specific.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6ff781c2bc24fc6deab65f3551446201 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/AWS_Bedrock_Specific.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=545dad2b676b279106b630da9db055ec 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/AWS_Bedrock_Specific.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=31d71145b968d1f6befeebaff1ab33cc 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/AWS_Bedrock_Specific.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=ba3457f9baf0f4c8f68dfee120847435 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/AWS_Bedrock_Specific.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=edd6013ece4fd91988e82fb58c40d1c5 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/AWS_Bedrock_Specific.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=b988d58bc2c709233bdb1ff827c7d938 2500w" />

## Setup guide

The process of using a Bedrock knowledge base with Pinecone works as follows:

<Steps>
  <Step title="Create a Pinecone index.">
    Create an empty Pinecone index with an embedding model in mind. The index must be empty for Bedrock integration.
  </Step>

  <Step title="Set up a data source.">
    Upload sample data to Amazon S3.
  </Step>

  <Step title="Create a Bedrock knowledge base.">
    Sync data with Bedrock to create embeddings saved in Pinecone.
  </Step>

  <Step title="Connect Pinecone to Bedrock.">
    Use the knowledge base to reference the data saved in Pinecone.
  </Step>

  <Step title="Create and link agents to Bedrock.">
    Agents can interact directly with the Bedrock knowledge base, which will retrieve the semantically relevant content.
  </Step>
</Steps>

### 1. Create a Pinecone index

The knowledge base stores data in a Pinecone index. Decide which [supported embedding model](https://docs.aws.amazon.com/bedrock/latest/userguide/models-supported.html) to use with Bedrock before you create the index, as your index's dimensions will need to match the model's. For example, the AWS Titan Text Embeddings V2 model can use dimension sizes 1024, 384, and 256.

After signing up to Pinecone, follow the [quickstart guide](/guides/get-started/quickstart) to create your Pinecone index and retrieve your `apiKey` and index host from the [Pinecone console](https://app.pinecone.io).

<Note>
  Your index must have the same dimensions as the model you will later select for creating your embeddings. Also, your index must be empty. All data must be ingested through Bedrock's sync process.
</Note>

### 2. Set up your data source

#### Set up secrets

After setting up your Pinecone index, you'll have to create a secret in [AWS Secrets Manager](https://console.aws.amazon.com/secretsmanager/newsecret):

1. In the **Secret type** section, select **Other type of secret**.
2. In the **Key/value pairs** section, enter a key value pair for the Pinecone API key name and its respective value. For example, use `apiKey` and the API key value.
   <img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/kv.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=78ec35c6ece3f8c3a636f5c5de531fb2" alt="" data-og-width="1736" width="1736" data-og-height="448" height="448" data-path="images/bedrock/kv.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/kv.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=e3923c940bf9a654e45323403d063a54 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/kv.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=32223b794ff61a6a22274cd3f0b23798 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/kv.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=c968fd5d4f10554a5de34eb10f5e6fb9 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/kv.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=b7c9283b8372fe7192bda11d1a68051c 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/kv.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=3170bd1acc0c4d024c405ee56ee72748 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/kv.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=3d0002f3e9bf1d953007c6f92ae9c202 2500w" />
3. Click **Next**.
4. Enter a **Secret name** and **Description**.
5. Click **Next** to save your key.
6. On the **Configure rotation** page, select all the default options in the next screen, and click **Next**.
7. Click **Store**.
8. Click on the new secret you created and save the secret ARN for a later step.

#### Set up S3

The knowledge base is going to draw on data saved in S3. For this example, we use a [sample of research papers](https://huggingface.co/datasets/jamescalam/ai-arxiv2-semantic-chunks) obtained from a dataset. This data will be embedded and then saved in Pinecone.

1. Create a new general purpose bucket in [Amazon S3](https://console.aws.amazon.com/s3/home).

2. Once the bucket is created, upload a CSV file.

   <Note>
     The CSV file must have a field for text that will be embedded, and a field for metadata to upload with each embedded text.
   </Note>

3. Save your bucket's address (`s3://…`) for the following configuration steps.

### 3. Create a Bedrock knowledge base

To [create a Bedrock knowledge base](https://console.aws.amazon.com/bedrock/home?#/knowledge-bases/create-knowledge-base), use the following steps:

1. Enter a **Knowledge Base name**.
2. In the **Choose data source** section, select **Amazon S3**.
3. Click **Next**.
4. On the **Configure data source** page, enter the **S3 URI** for the bucket you created.
5. If you do not want to use the default chunking strategy, select a chunking strategy.
6. Click **Next**.

### 4. Connect Pinecone to the knowledge base

Now you will need to select an embedding model to configure with Bedrock and configure the data sources.

1. Select the embedding model you decided on earlier.
2. For the **Vector database**, select **Choose a vector store you have created** and select **Pinecone**.
3. Mark the check box for authorizing AWS to access your Pinecone index.

   <Note>
     Ensure your Pinecone index is empty before proceeding. Bedrock cannot work with indexes that contain existing data. All data must be ingested through Bedrock's sync process.
   </Note>
4. For the **Endpoint URL**, enter the Pinecone index host retrieved from the Pinecone console.
5. For the **Credentials secret ARN**, enter the secret ARN you created earlier.
6. In the **Metadata field mapping** section, enter the **Text field name** you want to embed and the **Bedrock-managed metadata field name** that will be used for metadata managed by Bedrock (e.g., `metadata`).
7. Click **Next**.
8. Review your selections and complete the creation of the knowledge base.
9. On the [Knowledge Bases](https://console.aws.amazon.com/bedrock/home?#/knowledge-bases) page select the knowledge base you just created to view its details.
10. Click **Sync** for the newly created data source.
    <Note>
      Sync the data source whenever you add new data to the data source to start the ingestion workflow of converting your Amazon S3 data into vector embeddings and upserting the embeddings into the vector database. Depending on the amount of data, this whole workflow can take some time.
    </Note>

### 5. Create and link an agent to Bedrock

Lastly, [create an agent](https://console.aws.amazon.com/bedrock/home?#/agents) that will use the knowledge base for retrieval:

1. Click **Create Agent**.
2. Enter an **Name** and **Description**.
3. Click **Create**.
4. Select the LLM provider and model you'd like to use.
5. Provide instructions for the agent. These will define what the agent is trying to accomplish.
6. In the **Knowledge Bases** section, select the knowledge base you created.
7. Prepare the agent by clicking **Prepare** near the top of the builder page.
8. Test the agent after preparing it to verify it is using the knowledge base.
9. Click **Save and exit**.

Your agent is now set up and ready to go! In the next section, we'll show how to interact with the newly created agent.

#### Create an alias for your agent

In order to deploy the agent, create an alias for it that points to a specific version of the agent. Once the alias is created, it will display in the agent view.

1. On the [Agents](https://console.aws.amazon.com/bedrock/home?#/agents) page, select the agent you created.
2. Click **Create Alias**.
3. Enter an **Alias name** and **Description**.
4. Click **Create alias**.

#### Test the Bedrock agent

To test the newly created agent, use the playground on the right of the screen when we open the agent.

In this example, we used a dataset of research papers for our source data. We can ask a question about those papers and retrieve a detailed response, this time with the deployed version.

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-query.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=bcc833861883b073434c43817f0c5c84" alt="" data-og-width="3142" width="3142" data-og-height="1796" height="1796" data-path="images/bedrock/alias-testing-query.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-query.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=96efbe192b82b95625698e95545b8f68 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-query.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=e267b66dbe0919337252cb2abf0bc6c7 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-query.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6927f3eb7510e0e7600028a6e8a1e83f 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-query.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=4fdd8b786767b97280c93fd2a648cc4f 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-query.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=ec4b8236c85faf371655e41b66fc67f9 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-query.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=5ee4801c0f09e8c0344a258141898677 2500w" />

By inspecting the trace, we can see what chunks were used by the Agent and diagnose issues with responses.

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-trace.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=e8aa228c1ad062975f12e56ea22cd899" alt="" data-og-width="3072" width="3072" data-og-height="1592" height="1592" data-path="images/bedrock/alias-testing-trace.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-trace.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=67036b99f9f7fa5ed38273f9597e0421 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-trace.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=8b4429bfc8471f314fde89e19bd8a790 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-trace.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=fe78ef34350cee516e9d15839ac20701 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-trace.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=f6ac178d952aad0482d21f7bff975a39 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-trace.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=c5e913f349a760f11c95ad1eec4279dc 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/bedrock/alias-testing-trace.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6e3c2fe335a26f3a05a996e62fefbe60 2500w" />

## Related articles

* [Pinecone as a Knowledge Base for Amazon Bedrock](https://www.pinecone.io/blog/amazon-bedrock-integration/)

---

# Source: https://docs.pinecone.io/integrations/amazon-sagemaker.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Amazon SageMaker

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Amazon SageMaker is a fully managed service that brings together a broad set of tools to enable high-performance, low-cost machine learning (ML) for any use case. With SageMaker, you can build, train and deploy ML models at scale-- all in one integrated development environment (IDE). SageMaker supports governance requirements with simplified access control and transparency over your ML projects. Amazon SageMaker offers access to hundreds of pretrained models, including publicly available foundational models (FMs), and you can build your own FMs with purpose-built tools to fine-tune, experiment, retrain, and deploy FMs.

Amazon SageMaker and Pinecone can be used together for high-performance, scalable, and reliable retrieval augmented generation (RAG) use cases. The integration uses Amazon SageMaker to compute and host models for large language Models (LLMs), and uses Pinecone as the knowledge base that keeps the LLMs up-to-date with the latest information, reducing the likelihood of hallucinations.

<PrimarySecondaryCTA primaryHref={"https://www.pinecone.io/learn/sagemaker-rag/"} primaryLabel={"Get started"} />

## Related articles

* [Mitigate hallucinations through Retrieval Augmented Generation using Pinecone vector database & Llama-2 from Amazon SageMaker JumpStart](https://aws.amazon.com/blogs/machine-learning/mitigate-hallucinations-through-retrieval-augmented-generation-using-pinecone-vector-database-llama-2-from-amazon-sagemaker-jumpstart/)

---

# Source: https://docs.pinecone.io/integrations/anyscale.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Anyscale

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Anyscale Endpoints offers open-source large language models (LLMs) as fully managed API endpoints. This allows you to focus on building applications powered by LLMs without the need to worry about the underlying infrastructure.

Use Anyscale Endpoints with Canopy, an open source SDK that allows you to test, build, and package retrieval augmented generation (RAG) applications with Pinecone vector database.

<PrimarySecondaryCTA primaryHref={"https://github.com/anyscale/endpoint-cookbook/blob/main/Integration_Canopy.ipynb"} primaryLabel={"Get started"} />

---

# Source: https://docs.pinecone.io/integrations/apify.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Apify

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

[Apify](https://apify.com) is a web scraping and data extraction platform. It provides an app store with more than a thousand ready-made cloud tools called Actors. These tools are suitable for use cases including extracting structured data from e-commerce sites, social media, search engines, online maps, or any other website.

For example, the [Website Content Crawler](https://apify.com/apify/website-content-crawler) Actor can deeply crawl websites, clean their HTML by removing a cookies modal, footer, or navigation, and then transform the HTML into Markdown. This Markdown can then be used as training data for AI models or to feed LLM and generative AI applications with web content.

The Apify integration for Pinecone makes it easy to transfer results from Actors to the Pinecone vector database, enabling Retrieval-Augmented Generation (RAG) or semantic search over data extracted from the web.

<PrimarySecondaryCTA primaryHref={"https://docs.apify.com/platform/integrations/pinecone"} primaryLabel={"Get started"} />

---

# Source: https://docs.pinecone.io/integrations/aryn.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Aryn

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Aryn is an AI-powered ETL system for complex, unstructured documents like PDFs, HTML, presentations, and more. It's purpose-built for building RAG and GenAI applications, providing up to 6x better accuracy in chunking and extracting information from documents. This can lead to 30% better recall and 2x improvement in answer accuracy for real-world use cases.

Aryn's ETL system has two components: Sycamore and the Aryn Partitioning Service. Sycamore is Aryn's open source document processing engine, available as a Python library. It contains a set of transforms for information extraction, LLM-powered enrichment, data cleaning, creating vector embeddings, and loading Pinecone indexes.

The Aryn Partitioning Service is used as a first step in a Sycamore data processing pipeline, and it identifies and extracts parts of documents, like text, tables, images, and more. It uses a state-of-the-art vision segmentation AI model, trained on hundreds of thousands of human-annotated documents.

The Pinecone integration with Aryn enables developers to easily chunk documents, create vector embeddings, and load Pinecone with high-quality data.

<PrimarySecondaryCTA primaryHref={"https://sycamore.readthedocs.io/en/stable/sycamore/connectors/pinecone.html"} primaryLabel={"Get started"} primaryTarget={"_blank"} />

---

# Source: https://docs.pinecone.io/reference/architecture/assistant-architecture.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Pinecone Assistant architecture

This page describes the architecture for [Pinecone Assistant](/guides/assistant/overview).

## Overview

[Pinecone Assistant](/guides/assistant/overview) runs as a managed service on the Pinecone platform. It uses a combination of machine learning models and information retrieval techniques to provide responses that are informed by your documents. The assistant is designed to be easy to use, requiring minimal setup and no machine learning expertise.

Pinecone Assistant simplifies complex tasks like data chunking, vector search, embedding, and querying while ensuring privacy and security.

<img className="block max-w-full dark:hidden" noZoom src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_light.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=fac1ffcff3fc58b8c48abb41070f76bd" data-og-width="1520" width="1520" data-og-height="1320" height="1320" data-path="images/assistant-architecture_light.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_light.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=da402176da803658c8ffbdda0287e1d8 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_light.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=05a4da048111f0f38a919f39252e2cc1 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_light.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=47eb95307afaca43eea69a8914f52ec1 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_light.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=96dfe1589dda495073c114a8fc01b9fd 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_light.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=1bc84f97b47498ae2ab7c6969c4ab37d 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_light.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=52df04932cfe358add8585cdbfcca0f4 2500w" />

<img className="hidden max-w-full dark:block" noZoom src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_dark.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=52f8183973f02f2f1eb8f88d929035c8" data-og-width="1520" width="1520" data-og-height="1320" height="1320" data-path="images/assistant-architecture_dark.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_dark.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6f77677652e9e12a46ad36facc47c8e4 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_dark.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=c51b5d995b5f8c5125a7f3f1bbd12995 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_dark.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=9d1740644d873ac940d8fb5340c9becb 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_dark.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=cdfa4ba7f3ae803cdcdd807455cfb165 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_dark.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=c431d99fde890a6f016b06ae1a7f985f 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/assistant-architecture_dark.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=3c4cbac2096c93879d340740a731d119 2500w" />

## Data ingestion

When a [document is uploaded](/guides/assistant/manage-files), the assistant processes the content by chunking it into smaller parts and generating [vector embeddings](https://www.pinecone.io/learn/vector-embeddings-for-developers/) for each chunk. These embeddings are stored in an [index](/guides/index-data/indexing-overview), making them ready for retrieval.

## Data retrieval

During a [chat](/guides/assistant/chat-with-assistant), the assistant processes the message to formulate relevant search queries, which are used to query the index and identify the most relevant chunks from the uploaded content.

## Response generation

After retrieving these chunks, the assistant performs a ranking step to determine which information is most relevant. This [context](/guides/assistant/context-snippets-overview), along with the chat history and [assistant instructions](/guides/assistant/manage-assistants#add-instructions-to-an-assistant), is then used by a large language model (LLM) to generate responses that are informed by your documents.

---

# Source: https://docs.pinecone.io/reference/api/assistant/assistant-limits.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Pinecone Assistant limits

Pinecone Assistant limits vary based on [subscription plan](https://www.pinecone.io/pricing/).

### Object limits

Object limits are restrictions on the number or size of assistant-related objects.

| Metric                               | Starter plan  | Standard plan | Enterprise plan |
| :----------------------------------- | :------------ | :------------ | :-------------- |
| Assistants per project               | 5             | Unlimited     | Unlimited       |
| File storage per project             | 1 GB          | Unlimited     | Unlimited       |
| Chat input tokens per project        | 1,500,000     | Unlimited     | Unlimited       |
| Chat output tokens per project       | 200,000       | Unlimited     | Unlimited       |
| Context retrieval tokens per project | 500,000       | Unlimited     | Unlimited       |
| Evaluation input tokens per project  | Not available | 150,000       | 500,000         |
| Files per assistant                  | 100           | 10,000        | 10,000          |
| File size (.docx, .json, .md, .txt)  | 10 MB         | 10 MB         | 10 MB           |
| File size (.pdf)                     | 10 MB         | 100 MB        | 100 MB          |
| Metadata size per file               | 16 KB         | 16 KB         | 16 KB           |

Additionally, the following limits apply to [multimodal PDFs](/guides/assistant/multimodal) (currently in [public preview](/release-notes/feature-availability)):

| Metric                        | Starter plan | Standard plan | Enterprise plan |
| :---------------------------- | :----------- | :------------ | :-------------- |
| Max file size                 | 10 MB        | 50 MB         | 50 MB           |
| Page limit                    | 100          | 100           | 100             |
| Multimodal PDFs per assistant | 10           | 20            | 20              |

### Rate limits

Rate limits help protect your applications from misuse and maintain the health of our shared infrastructure. These limits are designed to support typical production workloads while ensuring reliable performance for all users.

**Most rate limits can be adjusted upon request.** If you need higher limits to scale your application, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket) with details about your use case.

Requests that exceed a rate limit fail and return a `429 - TOO_MANY_REQUESTS` status.

<Tip>To handle rate limits, implement [retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic).</Tip>

| Metric                                      | Starter plan | Standard plan | Enterprise plan |
| :------------------------------------------ | :----------- | :------------ | :-------------- |
| Assistant list/get requests per minute      | 40           | 100           | 500             |
| Assistant create/update requests per minute | 20           | 50            | 100             |
| Assistant delete requests per minute        | 20           | 50            | 100             |
| File get requests per minute                | 100          | 300           | 6,000           |
| File list requests per minute               | 50           | 150           | 3,000           |
| File upload requests per minute             | 5            | 20            | 300             |
| File delete requests per minute             | 5            | 20            | 300             |
| Chat input tokens per minute                | 100,000      | 300,000       | 1,000,000       |
| Chat history tokens per query               | 64,000       | 64,000        | 64,000          |

---

# Source: https://docs.pinecone.io/examples/assistant.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Assistant examples

export const Tag = ({text, icon}) => {
  return <span className="card-tag">
            {icon && <img src={icon} className={`w-4 h-4 object-contain ${icon.includes("openai") ? "dark-inverted" : ""}`} />}
            {text}
        </span>;
};

export const UtilityExampleCard = ({title, text, link}) => {
  return <a href={link} className="example-card group">
            <h2 className="font-bold" style={{
    fontSize: "0.875rem"
  }}>{title}</h2>
            <p style={{
    fontSize: "0.875rem",
    marginTop: "0"
  }}>{text}</p>
        </a>;
};

export const ExampleCard = ({title, text, link, children, arrow, vectors, namespaces}) => {
  return <a href={link} className="example-card group">
            <h2 className="font-semibold text-base">{title}</h2>
            <p>{text}</p>

            {children && <div className="tags">{children}</div>}

            {arrow && <svg xmlns="http://www.w3.org/2000/svg" className="arrow" width="24" height="24" viewBox="0 0 24 24" fill="none">
                <path d="M5.30739 20L4 18.6926L16.8249 5.8677H5.05837V4H20V18.9416H18.1323V7.1751L5.30739 20Z" fill="var(--text-secondary)" />
            </svg>}

            {(vectors || namespaces) && <div className="vectors">
                {vectors && <span>{vectors} vectors</span>}
                {namespaces && <span>{namespaces} namespaces</span>}
            </div>}
        </a>;
};

<div className="card-grid not-prose">
  <ExampleCard title="Quickstart" text="Get started with Pinecone Assistant" link="https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/assistant-quickstart.ipynb" arrow>
    <Tag text="Python" />
  </ExampleCard>

  <ExampleCard title="Retrieve context snippets" text="Retrieve relevant information snippets for your queries" link="https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/assistant-retrieve-context-snippets.ipynb" arrow>
    <Tag text="Python" />
  </ExampleCard>

  <ExampleCard title="Pinecone Assistant" text="Connect your existing Pinecone Assistant to a chat interface with citations and file references." link="/examples/sample-apps/pinecone-assistant">
    <Tag text="NextJS" icon="/images/examples/nextjs.svg" />
  </ExampleCard>
</div>

---

# Source: https://docs.pinecone.io/integrations/build-integration/attribute-usage-to-your-integration.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Attribute usage to your integration

Once you have created your integration with Pinecone, specify a **source tag** when instantiating clients with Pinecone SDKs, or pass a source tag as part of the `User-Agent` header when using the API directly.

<Note>
  Anyone can create an integration, but [becoming an official Pinecone partner](/integrations/build-integration/integration-ecosystem) can help accelerate your go-to-market and add value to your customers.
</Note>

### Source tag naming conventions

Your source tag must follow these conventions:

* Clearly identify your integration.
* Use only lowercase letters, numbers, underscores, and colons.

For example, for an integration called "New Framework", `"new_framework"` is valid, but `"new framework"` and `"New_framework"` are not valid.

### Specify a source tag

| Pinecone SDK                              | Required version |
| ----------------------------------------- | ---------------- |
| [Python](/reference/sdks/python/overview) | v3.2.1+          |
| [Node.js](/reference/sdks/node/overview)  | v2.2.0+          |
| [Java](/reference/sdks/java/overview)     | v1.0.0+          |
| [Go](/reference/sdks/go/overview)         | v0.4.1+          |
| [.NET](/reference/sdks/dotnet/overview)   | v1.0.0+          |

<CodeGroup>
  ```python Python theme={null}
  # REST client
  from pinecone import Pinecone

  pc = Pinecone(
      api_key="YOUR_API_KEY", 
      source_tag="YOUR_SOURCE_TAG"
  )

  # gRPC client
  from pinecone.grpc import PineconeGRPC

  pc = PineconeGRPC(
      api_key="YOUR_API_KEY", 
      source_tag="YOUR_SOURCE_TAG"
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ 
      apiKey: 'YOUR_API_KEY', 
      sourceTag: 'YOUR_SOURCE_TAG' 
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;

  public class IntegrationExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY")
                  .withSourceTag("YOUR_SOURCE_TAG")
                  .build();
      }
  }
  ```

  ```go Go theme={null}
  import "github.com/pinecone-io/go-pinecone/v4/pinecone"

  client, err := pinecone.NewClient(pinecone.NewClientParams{
  	ApiKey: "YOUR_API_KEY",
  	SourceTag: "YOUR_SOURCE_TAG",
  })
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY", new ClientOptions
  {
      SourceTag = "YOUR_SOURCE_TAG",
  });
  ```

  ```shell curl theme={null}
  curl -i -X GET "https://api.pinecone.io/indexes" \
    -H "Accept: application/json" \
    -H "Api-Key: YOUR_API_KEY" \
    -H "User-Agent: source_tag=YOUR_SOURCE_TAG" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

---

# Source: https://docs.pinecone.io/reference/cli/authentication.md

# Source: https://docs.pinecone.io/reference/api/authentication.md

# Source: https://docs.pinecone.io/reference/api/assistant/authentication.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Authentication

All requests to the [Pinecone Assistant API](/reference/api/assistant/introduction) must contain a valid [API key](/guides/production/security-overview#api-keys) for the target project.

## Get an API key

[Create a new API key](https://app.pinecone.io/organizations/-/projects/-/keys) in the Pinecone console, or use the connect widget below to generate a key.

<div style={{minWidth: '450px', minHeight:'152px'}}>
  <div id="pinecone-connect-widget">
    <div class="connect-widget-skeleton">
      <div class="skeleton-content" />
    </div>
  </div>
</div>

Copy your generated key:

```
PINECONE_API_KEY="{{YOUR_API_KEY}}"

# This API key has ReadWrite access to all indexes in your project.
```

## Initialize a client

When using a Pinecone SDK, initialize a client object with your API key and then reuse the authenicated client in subsquent function calls. For example:

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # Creates an assistant using the API key stored in the client 'pc'.
  assistant = pc.assistant.create_assistant(
      assistant_name="example-assistant",
      instructions="Use American English for spelling and grammar.",
      region="us" 
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  // Creates an index using the API key stored in the client 'pc'.
  const assistant = await pc.createAssistant({
    name: 'example-assistant',
    instructions: 'Use American English for spelling and grammar.',
    region: 'us'
  });
  ```
</CodeGroup>

## Add headers to an HTTP request

All HTTP requests to the Pinecone Assistant API must contain an `Api-Key` header that specifies a valid [API key](/guides/production/security-overview#api-keys) and must be encoded as JSON with the `Content-Type: application/json` header. For example:

```bash curl theme={null}
PINECONE_API_KEY="YOUR_API_KEY"

curl "https://api.pinecone.io/assistant/assistants" \
  -H "Api-Key: $PINECONE_API_KEY" \
  -H "Content-Type: application/json" \
  -H "X-Pinecone-Api-Version: 2025-01" \
  -d '{
  "name": "example-assistant",
  "instructions": "Use American English for spelling and grammar.",
  "region":"us"
}'
```

---

# Source: https://docs.pinecone.io/guides/production/automated-testing.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# CI/CD with Pinecone Local and GitHub Actions

> Test Pinecone integration with CI/CD workflows.

Pinecone Local is an in-memory Pinecone Database emulator available as a Docker image.

This page shows you how to build a CI/CD workflow with Pinecone Local and [GitHub Actions](https://docs.github.com/en/actions) to test your integration without connecting to your Pinecone account, affecting production data, or incurring any usage or storage fees.

<Warning>
  Pinecone Local is not suitable for production. See [Limitations](#limitations) for details.
</Warning>

<Note>
  This feature is in [public preview](/release-notes/feature-availability).
</Note>

## Limitations

Pinecone Local has the following limitations:

* Pinecone Local uses the `2025-01` API version, which is not the latest stable version.
* Pinecone Local is available in Docker only.
* Pinecone Local is an in-memory emulator and is not suitable for production. Records loaded into Pinecone Local do not persist after it is stopped.
* Pinecone Local does not authenticate client requests. API keys are ignored.
* Max number of records per index: 100,000.

Pinecone Local does not currently support the following features:

* [Import from object storage](/guides/index-data/import-data)
* [Backup/restore of serverless indexes](/guides/manage-data/backups-overview)
* [Collections for pod-based indexes](/guides/indexes/pods/understanding-collections)
* [Namespace management](/guides/manage-data/manage-namespaces)
* [Pinecone Inference](/reference/api/introduction#inference)
* [Pinecone Assistant](/guides/assistant/overview)

## 1. Write your tests

Running code against Pinecone Local is just like running code against your Pinecone account, with the following differences:

* Pinecone Local does not authenticate client requests. API keys are ignored.

* The latest version of Pinecone Local uses [Pinecone API version](/reference/api/versioning) `2025-01` and requires [Python SDK](/reference/sdks/python/overview) `v6.x` or later, [Node.js SDK](/reference/sdks/node/overview) `v5.x` or later, [Java SDK](/reference/sdks/java/overview) `v4.x` or later, [Go SDK](/reference/sdks/go/overview) `v3.x` or later, and [.NET SDK](/reference/sdks/dotnet/overview) `v3.x` or later.

<Note>
  Be sure to review the [limitations](#limitations) of Pinecone Local before using it for development or testing.
</Note>

**Example**

The following example assumes that you have [started Pinecone Local without indexes](/guides/operations/local-development#database-emulator). It initializes a client, creates a [dense index](/guides/index-data/indexing-overview#dense-indexes) and a [sparse index](/guides/index-data/indexing-overview#sparse-indexes), upserts records into the indexes, checks their record counts, and queries the indexes.

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC, GRPCClientConfig
  from pinecone import ServerlessSpec

  # Initialize a client.
  # API key is required, but the value does not matter.
  # Host and port of the Pinecone Local instance
  # is required when starting without indexes. 
  pc = PineconeGRPC(
      api_key="pclocal", 
      host="http://localhost:5080" 
  )                                    

  # Create two indexes, one dense and one sparse
  dense_index_name = "dense-index"
  sparse_index_name = "sparse-index"

  if not pc.has_index(dense_index_name):  
      dense_index_model = pc.create_index(
          name=dense_index_name,
          vector_type="dense",
          dimension=2,
          metric="cosine",
          spec=ServerlessSpec(cloud="aws", region="us-east-1"),
          deletion_protection="disabled",
          tags={"environment": "development"}
      )

  print("Dense index model:\n", dense_index_model)

  if not pc.has_index(sparse_index_name):  
      sparse_index_model = pc.create_index(
          name=sparse_index_name,
          vector_type="sparse",
          metric="dotproduct",
          spec=ServerlessSpec(cloud="aws", region="us-east-1"),
          deletion_protection="disabled",
          tags={"environment": "development"}
      )

  print("\nSparse index model:\n", sparse_index_model)

  # Target each index, disabling tls
  dense_index_host = pc.describe_index(name=dense_index_name).host
  dense_index = pc.Index(host=dense_index_host, grpc_config=GRPCClientConfig(secure=False))
  sparse_index_host = pc.describe_index(name=sparse_index_name).host
  sparse_index = pc.Index(host=sparse_index_host, grpc_config=GRPCClientConfig(secure=False))

  # Upsert records into the dense index
  dense_index.upsert(
      vectors=[
          {
              "id": "vec1", 
              "values": [1.0, -2.5],
              "metadata": {"genre": "drama"}
          },
          {
              "id": "vec2", 
              "values": [3.0, -2.0],
              "metadata": {"genre": "documentary"}
          },
          {
              "id": "vec3", 
              "values": [0.5, -1.5],
              "metadata": {"genre": "documentary"}
          }
      ],
      namespace="example-namespace"
  )

  # Upsert records into the sparse index
  sparse_index.upsert(
      namespace="example-namespace",
      vectors=[
          {
              "id": "vec1",
              "sparse_values": {
                  "values": [1.7958984, 0.41577148, 2.828125, 2.8027344, 2.8691406, 1.6533203, 5.3671875, 1.3046875, 0.49780273, 0.5722656, 2.71875, 3.0820312, 2.5019531, 4.4414062, 3.3554688],
                  "indices": [822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191]
              },
              "metadata": {
                  "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
                  "category": "technology",
                  "quarter": "Q3"
              }
          },
          {
              "id": "vec2",
              "sparse_values": {
                  "values": [0.4362793, 3.3457031, 2.7714844, 3.0273438, 3.3164062, 5.6015625, 2.4863281, 0.38134766, 1.25, 2.9609375, 0.34179688, 1.4306641, 0.34375, 3.3613281, 1.4404297, 2.2558594, 2.2597656, 4.8710938, 0.5605469],
                  "indices": [131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697]
              },
              "metadata": {
                  "chunk_text": "Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
                  "category": "technology",
                  "quarter": "Q4"
              }
          },
          {
              "id": "vec3",
              "sparse_values": {
                  "values": [2.6875, 4.2929688, 3.609375, 3.0722656, 2.1152344, 5.78125, 3.7460938, 3.7363281, 1.2695312, 3.4824219, 0.7207031, 0.0826416, 4.671875, 3.7011719, 2.796875, 0.61621094],
                  "indices": [8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697]
              },
              "metadata": {
                  "chunk_text": "AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production",
                  "category": "technology",
                  "quarter": "Q3"
              }
          }
      ]
  )

  # Check the number of records in each index
  print("\nDense index stats:\n", dense_index.describe_index_stats())
  print("\nSparse index stats:\n", sparse_index.describe_index_stats())

  # Query the dense index with a metadata filter
  dense_response = dense_index.query(
      namespace="example-namespace",
      vector=[3.0, -2.0],
      filter={"genre": {"$eq": "documentary"}},
      top_k=1,
      include_values=False,
      include_metadata=True
  )

  print("\nDense query response:\n", dense_response)

  # Query the sparse index with a metadata filter
  sparse_response = sparse_index.query(
      namespace="example-namespace",
      sparse_vector={
        "values": [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],
        "indices": [767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697]
      }, 
      filter={
          "quarter": {"$eq": "Q4"}
      },
      top_k=1,
      include_values=False,
      include_metadata=True
  )

  print("/nSparse query response:\n", sparse_response)

  # Delete the indexes
  pc.delete_index(name=dense_index_name)
  pc.delete_index(name=sparse_index_name)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  // Initialize a client.
  // API key is required, but the value does not matter.
  // Host and port of the Pinecone Local instance
  // is required when starting without indexes.
  const pc = new Pinecone({ 
      apiKey: 'pclocal', 
      controllerHostUrl: 'http://localhost:5080' 
  });

  // Create two indexes, one dense and one sparse
  const denseIndexName = 'dense-index';
  const sparseIndexName = 'sparse-index';

  const denseIndexModel = await pc.createIndex({
    name: denseIndexName,
    vectorType: 'dense',
    dimension: 2,
    metric: 'cosine',
    spec: {
      serverless: {
        cloud: 'aws',
        region: 'us-east-1'
      }
    },
    deletionProtection: 'disabled',
    tags: { environment: 'development' }, 
  });

  console.log('Dense index model:', denseIndexModel);

  const sparseIndexModel = await pc.createIndex({
      name: sparseIndexName,
      vectorType: 'sparse',
      metric: 'dotproduct',
      spec: {
        serverless: {
          cloud: 'aws',
          region: 'us-east-1'
        }
      },
      deletionProtection: 'disabled',
      tags: { environment: 'development' }, 
    });
    
  console.log('\nSparse index model:', sparseIndexModel);
    
  // Target each index
  const denseIndexHost = (await pc.describeIndex(denseIndexName)).host;
  const denseIndex = await pc.index(denseIndexName, 'http://' + denseIndexHost);
  const sparseIndexHost = (await pc.describeIndex(sparseIndexName)).host;
  const sparseIndex = await pc.index(sparseIndexName, 'http://' + sparseIndexHost);

  // Upsert records into the dense index
  await denseIndex.namespace('example-namespace').upsert([
      {
          id: 'vec1', 
          values: [1.0, -2.5],
          metadata: { genre: 'drama' },
      },
      {
          id: 'vec2', 
          values: [3.0, -2.0],
          metadata: { genre: 'documentary' },
      },
      {
          id: 'vec3', 
          values: [0.5, -1.5],
          metadata: { genre: 'documentary' },
      }
  ]);

  // Upsert records into the sparse index
  await sparseIndex.namespace('example-namespace').upsert([
      {
          id: 'vec1',
          sparseValues: {
              indices: [822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191],
              values: [1.7958984, 0.41577148, 2.828125, 2.8027344, 2.8691406, 1.6533203, 5.3671875, 1.3046875, 0.49780273, 0.5722656, 2.71875, 3.0820312, 2.5019531, 4.4414062, 3.3554688]
          },
          metadata: { 
              chunk_text: 'AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.', 
              category: 'technology',
              quarter: 'Q3' 
          }
      },
      {
          id: 'vec2',
          sparseValues: {
              indices: [131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697],
              values: [0.4362793, 3.3457031, 2.7714844, 3.0273438, 3.3164062, 5.6015625, 2.4863281, 0.38134766, 1.25, 2.9609375, 0.34179688, 1.4306641, 0.34375, 3.3613281, 1.4404297, 2.2558594, 2.2597656, 4.8710938, 0.5605469]
          },
          metadata: { 
              chunk_text: "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market.", 
              category: 'technology',
              quarter: 'Q4' 
          }
      },
      {
          id: 'vec3',
          sparseValues: {
              indices: [8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697],
              values: [2.6875, 4.2929688, 3.609375, 3.0722656, 2.1152344, 5.78125, 3.7460938, 3.7363281, 1.2695312, 3.4824219, 0.7207031, 0.0826416, 4.671875, 3.7011719, 2.796875, 0.61621094]
          },
          metadata: { 
              chunk_text: "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production", 
              category: 'technology',
              quarter: 'Q3' 
          }
      }
  ]);
    
  // Check the number of records in each index
  console.log('\nDense index stats:', await denseIndex.describeIndexStats());
  console.log('\nSparse index stats:', await sparseIndex.describeIndexStats());

  // Query the dense index with a metadata filter
  const denseQueryResponse = await denseIndex.namespace('example-namespace').query({
      vector: [3.0, -2.0],
      filter: {
          'genre': {'$eq': 'documentary'}
      },
      topK: 1,
      includeValues: false,
      includeMetadata: true,
  });

  console.log('\nDense query response:', denseQueryResponse);

  const sparseQueryResponse = await sparseIndex.namespace('example-namespace').query({
      sparseVector: {
          indices: [767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697],
          values: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
      },
      topK: 1,
      includeValues: false,
      includeMetadata: true
  });

  console.log('\nSparse query response:', sparseQueryResponse);

  // Delete the index
  await pc.deleteIndex(denseIndexName);
  await pc.deleteIndex(sparseIndexName);
  ```

  ```java Java theme={null}
  import com.google.protobuf.Struct;
  import com.google.protobuf.Value;
  import io.pinecone.clients.Index;
  import io.pinecone.clients.Pinecone;
  import io.pinecone.proto.DescribeIndexStatsResponse;
  import org.openapitools.db_control.client.model.DeletionProtection;
  import io.pinecone.unsigned_indices_model.QueryResponseWithUnsignedIndices;

  import java.util.*;

  public class PineconeLocalExample {
      public static void main(String[] args) {
          // Initialize a client.
          // API key is required, but the value does not matter.
          // When starting without indexes, disable TLS and
          // provide the host and port of the Pinecone Local instance.
          String host = "http://localhost:5080";
          Pinecone pc = new Pinecone.Builder("pclocal")
                  .withHost(host)
                  .withTlsEnabled(false)
                  .build();

          // Create two indexes, one dense and one sparse
          String denseIndexName = "dense-index";
          String sparseIndexName = "sparse-index";
          HashMap<String, String> tags = new HashMap<>();
          tags.put("environment", "development");
          pc.createServerlessIndex(
                  denseIndexName,
                  "cosine",
                  2,
                  "aws",
                  "us-east-1",
                  DeletionProtection.DISABLED,
                  tags
          );
          pc.createSparseServelessIndex(
                  sparseIndexName,
                  "aws",
                  "us-east-1",
                  DeletionProtection.DISABLED,
                  tags,
                  "sparse"
          );

          // Get index connection objects
          Index denseIndexConnection = pc.getIndexConnection(denseIndexName);
          Index sparseIndexConnection = pc.getIndexConnection(sparseIndexName);

          // Upsert records into the dense index
          Struct metaData1 = Struct.newBuilder()
                  .putFields("genre", Value.newBuilder().setStringValue("drama").build())
                  .build();
          Struct metaData2 = Struct.newBuilder()
                  .putFields("genre", Value.newBuilder().setStringValue("documentary").build())
                  .build();
          Struct metaData3 = Struct.newBuilder()
                  .putFields("genre", Value.newBuilder().setStringValue("documentary").build())
                  .build();

          denseIndexConnection.upsert("vec1", Arrays.asList(1.0f, -2.5f),  null, null, metaData1, "example-namespace");
          denseIndexConnection.upsert("vec2", Arrays.asList(3.0f, -2.0f),  null, null, metaData2, "example-namespace");
          denseIndexConnection.upsert("vec3", Arrays.asList(0.5f, -1.5f),  null, null, metaData3, "example-namespace");

          // Upsert records into the sparse index
          ArrayList<Long> indices1 = new ArrayList<>(Arrays.asList(
                  822745112L, 1009084850L, 1221765879L, 1408993854L, 1504846510L,
                  1596856843L, 1640781426L, 1656251611L, 1807131503L, 2543655733L,
                  2902766088L, 2909307736L, 3246437992L, 3517203014L, 3590924191L
          ));

          ArrayList<Float> values1 = new ArrayList<>(Arrays.asList(
                  1.7958984f, 0.41577148f, 2.828125f, 2.8027344f, 2.8691406f,
                  1.6533203f, 5.3671875f, 1.3046875f, 0.49780273f, 0.5722656f,
                  2.71875f, 3.0820312f, 2.5019531f, 4.4414062f, 3.3554688f
          ));

          Struct sparseMetaData1 = Struct.newBuilder()
                  .putFields("chunk_text", Value.newBuilder().setStringValue("AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.").build())
                  .putFields("category", Value.newBuilder().setStringValue("technology").build())
                  .putFields("quarter", Value.newBuilder().setStringValue("Q3").build())
                  .build();

          ArrayList<Long> indices2 = new ArrayList<>(Arrays.asList(
                  131900689L, 592326839L, 710158994L, 838729363L, 1304885087L,
                  1640781426L, 1690623792L, 1807131503L, 2066971792L, 2428553208L,
                  2548600401L, 2577534050L, 3162218338L, 3319279674L, 3343062801L,
                  3476647774L, 3485013322L, 3517203014L, 4283091697L
          ));

          ArrayList<Float> values2 = new ArrayList<>(Arrays.asList(
                  0.4362793f, 3.3457031f, 2.7714844f, 3.0273438f, 3.3164062f,
                  5.6015625f, 2.4863281f, 0.38134766f, 1.25f, 2.9609375f,
                  0.34179688f, 1.4306641f, 0.34375f, 3.3613281f, 1.4404297f,
                  2.2558594f, 2.2597656f, 4.8710938f, 0.5605469f
          ));

          Struct sparseMetaData2 = Struct.newBuilder()
                  .putFields("chunk_text", Value.newBuilder().setStringValue("Analysts suggest that AAPL'\\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.").build())
                  .putFields("category", Value.newBuilder().setStringValue("technology").build())
                  .putFields("quarter", Value.newBuilder().setStringValue("Q4").build())
                  .build();

          ArrayList<Long> indices3 = new ArrayList<>(Arrays.asList(
                  8661920L, 350356213L, 391213188L, 554637446L, 1024951234L,
                  1640781426L, 1780689102L, 1799010313L, 2194093370L, 2632344667L,
                  2641553256L, 2779594451L, 3517203014L, 3543799498L,
                  3837503950L, 4283091697L
          ));

          ArrayList<Float> values3 = new ArrayList<>(Arrays.asList(
                  2.6875f, 4.2929688f, 3.609375f, 3.0722656f, 2.1152344f,
                  5.78125f, 3.7460938f, 3.7363281f, 1.2695312f, 3.4824219f,
                  0.7207031f, 0.0826416f, 4.671875f, 3.7011719f, 2.796875f,
                  0.61621094f
          ));

          Struct sparseMetaData3 = Struct.newBuilder()
                  .putFields("chunk_text", Value.newBuilder().setStringValue("AAPL'\\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production").build())
                  .putFields("category", Value.newBuilder().setStringValue("technology").build())
                  .putFields("quarter", Value.newBuilder().setStringValue("Q3").build())
                  .build();

          sparseIndexConnection.upsert("vec1", Collections.emptyList(), indices1, values1, sparseMetaData1, "example-namespace");
          sparseIndexConnection.upsert("vec2", Collections.emptyList(), indices2, values2, sparseMetaData2, "example-namespace");
          sparseIndexConnection.upsert("vec3", Collections.emptyList(), indices3, values3, sparseMetaData3, "example-namespace");

          // Check the number of records each the index
          DescribeIndexStatsResponse denseIndexStatsResponse = denseIndexConnection.describeIndexStats(null);
          System.out.println("Dense index stats:");
          System.out.println(denseIndexStatsResponse);
          DescribeIndexStatsResponse sparseIndexStatsResponse = sparseIndexConnection.describeIndexStats(null);
          System.out.println("Sparse index stats:");
          System.out.println(sparseIndexStatsResponse);

          // Query the dense index with a metadata filter
          List<Float> queryVector = Arrays.asList(1.0f, 1.5f);
          QueryResponseWithUnsignedIndices denseQueryResponse = denseIndexConnection.query(1, queryVector, null, null, null, "example-namespace", null, false, true);
          System.out.println("Dense query response:");
          System.out.println(denseQueryResponse);

          // Query the sparse index with a metadata filter
          List<Long> sparseIndices = Arrays.asList(
                  767227209L, 1640781426L, 1690623792L, 2021799277L, 2152645940L,
                  2295025838L, 2443437770L, 2779594451L, 2956155693L, 3476647774L,
                  3818127854L, 428309169L);
          List<Float> sparseValues = Arrays.asList(
                  1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f,
                  1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f);

          QueryResponseWithUnsignedIndices sparseQueryResponse = sparseIndexConnection.query(1, null, sparseIndices, sparseValues, null, "example-namespace", null, false, true);
          System.out.println("Sparse query response:");
          System.out.println(sparseQueryResponse);

          // Delete the indexes
          pc.deleteIndex(denseIndexName);
          pc.deleteIndex(sparseIndexName);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  	"google.golang.org/protobuf/types/known/structpb"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
  	ctx := context.Background()

  	// Initialize a client.
  	// No API key is required.
  	// Host and port of the Pinecone Local instance
  	// is required when starting without indexes.
  	pc, err := pinecone.NewClientBase(pinecone.NewClientBaseParams{
  		Host: "http://localhost:5080",
  	})
  	if err != nil {
  		log.Fatalf("Failed to create Client: %v", err)
  	}

  	// Create two indexes, one dense and one sparse
  	denseIndexName := "dense-index"
  	denseVectorType := "dense"
  	dimension := int32(2)
  	denseMetric := pinecone.Cosine
  	deletionProtection := pinecone.DeletionProtectionDisabled
  	denseIdx, err := pc.CreateServerlessIndex(ctx, &pinecone.CreateServerlessIndexRequest{
  		Name:               denseIndexName,
  		VectorType:         &denseVectorType,
  		Dimension:          &dimension,
  		Metric:             &denseMetric,
  		Cloud:              pinecone.Aws,
  		Region:             "us-east-1",
  		DeletionProtection: &deletionProtection,
  		Tags:               &pinecone.IndexTags{"environment": "development"},
  	})
  	if err != nil {
  		log.Fatalf("Failed to create serverless index: %v", denseIdx.Name)
  	} else {
  		fmt.Printf("Successfully created serverless index: %v\n", denseIdx.Name)
  	}

  	sparseIndexName := "sparse-index"
  	sparseVectorType := "sparse"
  	sparseMetric := pinecone.Dotproduct
  	sparseIdx, err := pc.CreateServerlessIndex(ctx, &pinecone.CreateServerlessIndexRequest{
  		Name:               sparseIndexName,
  		VectorType:         &sparseVectorType,
  		Metric:             &sparseMetric,
  		Cloud:              pinecone.Aws,
  		Region:             "us-east-1",
  		DeletionProtection: &deletionProtection,
  		Tags:               &pinecone.IndexTags{"environment": "development"},
  	})
  	if err != nil {
  		log.Fatalf("Failed to create serverless index: %v", sparseIdx.Name)
  	} else {
  		fmt.Printf("\nSuccessfully created serverless index: %v\n", sparseIdx.Name)
  	}

  	// Get the index hosts
  	denseIdxModel, err := pc.DescribeIndex(ctx, denseIndexName)
  	if err != nil {
  		log.Fatalf("Failed to describe index \"%v\": %v", denseIndexName, err)
  	}

  	sparseIdxModel, err := pc.DescribeIndex(ctx, sparseIndexName)
  	if err != nil {
  		log.Fatalf("Failed to describe index \"%v\": %v", sparseIndexName, err)
  	}

  	// Target the indexes.
  	// Make sure to prefix the hosts with http:// to let the SDK know to disable tls.
  	denseIdxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "http://" + denseIdxModel.Host, Namespace: "example-namespace"})
  	if err != nil {
  		log.Fatalf("Failed to create IndexConnection for Host: %v", err)
  	}

  	sparseIdxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "http://" + sparseIdxModel.Host, Namespace: "example-namespace"})
  	if err != nil {
  		log.Fatalf("Failed to create IndexConnection for Host: %v", err)
  	}

  	// Upsert records into the dense index
  	denseMetadataMap1 := map[string]interface{}{
  		"genre": "drama",
  	}

  	denseMetadata1, err := structpb.NewStruct(denseMetadataMap1)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	denseMetadataMap2 := map[string]interface{}{
  		"genre": "documentary",
  	}

  	denseMetadata2, err := structpb.NewStruct(denseMetadataMap2)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	denseMetadataMap3 := map[string]interface{}{
  		"genre": "documentary",
  	}

  	denseMetadata3, err := structpb.NewStruct(denseMetadataMap3)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	denseVectors := []*pinecone.Vector{
  		{
  			Id:       "vec1",
  			Values:   &[]float32{1.0, -2.5},
  			Metadata: denseMetadata1,
  		},
  		{
  			Id:       "vec2",
  			Values:   &[]float32{3.0, -2.0},
  			Metadata: denseMetadata2,
  		},
  		{
  			Id:       "vec3",
  			Values:   &[]float32{0.5, -1.5},
  			Metadata: denseMetadata3,
  		},
  	}

  	denseCount, err := denseIdxConnection.UpsertVectors(ctx, denseVectors)
  	if err != nil {
  		log.Fatalf("Failed to upsert vectors: %v", err)
  	} else {
  		fmt.Printf("\nSuccessfully upserted %d vector(s)!\n", denseCount)
  	}

  	// Upsert records into the sparse index
  	sparseValues1 := pinecone.SparseValues{
  		Indices: []uint32{822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191},
  		Values:  []float32{1.7958984, 0.41577148, 2.828125, 2.8027344, 2.8691406, 1.6533203, 5.3671875, 1.3046875, 0.49780273, 0.5722656, 2.71875, 3.0820312, 2.5019531, 4.4414062, 3.3554688},
  	}

  	sparseMetadataMap1 := map[string]interface{}{
  		"chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones",
  		"category":   "technology",
  		"quarter":    "Q3",
  	}

  	sparseMetadata1, err := structpb.NewStruct(sparseMetadataMap1)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	sparseValues2 := pinecone.SparseValues{
  		Indices: []uint32{131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697},
  		Values:  []float32{0.4362793, 3.3457031, 2.7714844, 3.0273438, 3.3164062, 5.6015625, 2.4863281, 0.38134766, 1.25, 2.9609375, 0.34179688, 1.4306641, 0.34375, 3.3613281, 1.4404297, 2.2558594, 2.2597656, 4.8710938, 0.560546},
  	}

  	sparseMetadataMap2 := map[string]interface{}{
  		"chunk_text": "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
  		"category":   "technology",
  		"quarter":    "Q4",
  	}

  	sparseMetadata2, err := structpb.NewStruct(sparseMetadataMap2)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	sparseValues3 := pinecone.SparseValues{
  		Indices: []uint32{8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697},
  		Values:  []float32{2.6875, 4.2929688, 3.609375, 3.0722656, 2.1152344, 5.78125, 3.7460938, 3.7363281, 1.2695312, 3.4824219, 0.7207031, 0.0826416, 4.671875, 3.7011719, 2.796875, 0.61621094},
  	}

  	sparseMetadataMap3 := map[string]interface{}{
  		"chunk_text": "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production",
  		"category":   "technology",
  		"quarter":    "Q3",
  	}

  	sparseMetadata3, err := structpb.NewStruct(sparseMetadataMap3)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	sparseVectors := []*pinecone.Vector{
  		{
  			Id:           "vec1",
  			SparseValues: &sparseValues1,
  			Metadata:     sparseMetadata1,
  		},
  		{
  			Id:           "vec2",
  			SparseValues: &sparseValues2,
  			Metadata:     sparseMetadata2,
  		},
  		{
  			Id:           "vec3",
  			SparseValues: &sparseValues3,
  			Metadata:     sparseMetadata3,
  		},
  	}

  	sparseCount, err := sparseIdxConnection.UpsertVectors(ctx, sparseVectors)
  	if err != nil {
  		log.Fatalf("Failed to upsert vectors: %v", err)
  	} else {
  		fmt.Printf("\nSuccessfully upserted %d vector(s)!\n", sparseCount)
  	}

  	// Check the number of records in each index
  	denseStats, err := denseIdxConnection.DescribeIndexStats(ctx)
  	if err != nil {
  		log.Fatalf("Failed to describe index: %v", err)
  	} else {
  		fmt.Printf("\nDense index stats: %+v\n", prettifyStruct(*denseStats))
  	}

  	sparseStats, err := sparseIdxConnection.DescribeIndexStats(ctx)
  	if err != nil {
  		log.Fatalf("Failed to describe index: %v", err)
  	} else {
  		fmt.Printf("\nSparse index stats: %+v\n", prettifyStruct(*sparseStats))
  	}

  	// Query the dense index with a metadata filter
  	queryVector := []float32{3.0, -2.0}

  	queryMetadataMap := map[string]interface{}{
  		"genre": map[string]interface{}{
  			"$eq": "documentary",
  		},
  	}

  	metadataFilter, err := structpb.NewStruct(queryMetadataMap)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	denseRes, err := denseIdxConnection.QueryByVectorValues(ctx, &pinecone.QueryByVectorValuesRequest{
  		Vector:          queryVector,
  		TopK:            1,
  		MetadataFilter:  metadataFilter,
  		IncludeValues:   false,
  		IncludeMetadata: true,
  	})
  	if err != nil {
  		log.Fatalf("Error encountered when querying by vector: %v", err)
  	} else {
  		fmt.Printf("\nDense query response: %v\n", prettifyStruct(denseRes))
  	}

  	// Query the sparse index with a metadata filter

  	sparseValues := pinecone.SparseValues{
  		Indices: []uint32{767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697},
  		Values:  []float32{1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0},
  	}

  	sparseRes, err := sparseIdxConnection.QueryByVectorValues(ctx, &pinecone.QueryByVectorValuesRequest{
  		SparseValues:    &sparseValues,
  		TopK:            1,
  		IncludeValues:   false,
  		IncludeMetadata: true,
  	})
  	if err != nil {
  		log.Fatalf("Error encountered when querying by vector: %v", err)
  	} else {
  		fmt.Printf("\nSparse query response: %v\n", prettifyStruct(sparseRes))
  	}
  	// Delete the indexes
  	err = pc.DeleteIndex(ctx, denseIndexName)
  	if err != nil {
  		log.Fatalf("Failed to delete index: %v", err)
  	} else {
  		fmt.Printf("\nIndex \"%v\" deleted successfully\n", denseIndexName)
  	}

  	err = pc.DeleteIndex(ctx, sparseIndexName)
  	if err != nil {
  		log.Fatalf("Failed to delete index: %v", err)
  	} else {
  		fmt.Printf("\nIndex \"%v\" deleted successfully\n", sparseIndexName)
  	}
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  // Initialize a client.
  // API key is required, but the value does not matter.
  // When starting without indexes, disable TLS and
  // provide the host and port of the Pinecone Local instance.
  var pc = new PineconeClient("pclocal",
      new ClientOptions
      {
          BaseUrl = "http://localhost:5080",
          IsTlsEnabled = false
      }
  );

  // Create two indexes, one dense and one sparse
  var denseIndexName = "dense-index";
  var sparseIndexName = "sparse-index";

  var createDenseIndexRequest = await pc.CreateIndexAsync(new CreateIndexRequest
  {
      Name = denseIndexName,
      VectorType = VectorType.Dense,
      Dimension = 2,
      Metric = MetricType.Cosine,
      Spec = new ServerlessIndexSpec
      {
          Serverless = new ServerlessSpec
          {
              Cloud = ServerlessSpecCloud.Aws,
              Region = "us-east-1"
          }
      },
      DeletionProtection = DeletionProtection.Disabled,
      Tags = new Dictionary<string, string> 
      {  
          { "environment", "development" }
      }
  });

  Console.WriteLine("Dense index model:" + createDenseIndexRequest);

  var createSparseIndexRequest = await pc.CreateIndexAsync(new CreateIndexRequest
  {
      Name = sparseIndexName,
      VectorType = VectorType.Sparse,
      Metric = MetricType.Dotproduct,
      Spec = new ServerlessIndexSpec
      {
          Serverless = new ServerlessSpec
          {
              Cloud = ServerlessSpecCloud.Aws,
              Region = "us-east-1"
          }
      },
      DeletionProtection = DeletionProtection.Disabled,
      Tags = new Dictionary<string, string> 
      {  
          { "environment", "development" }
      }
  });

  Console.WriteLine("\nSparse index model:" + createSparseIndexRequest);

  // Target the indexes
  var denseIndex = pc.Index(denseIndexName);
  var sparseIndex = pc.Index(sparseIndexName);

  // Upsert records into the dense index
  var denseUpsertResponse = await denseIndex.UpsertAsync(new UpsertRequest()
  {
      Namespace = "example-namespace",
      Vectors = new List<Vector>
      {
          new Vector 
          { 
              Id = "vec1", 
              Values = new ReadOnlyMemory<float>([1.0f, -2.5f]),
              Metadata = new Metadata {
                  ["genre"] = new("drama"),
              },
          },
          new Vector 
          { 
              Id = "vec2", 
              Values = new ReadOnlyMemory<float>([3.0f, -2.0f]), 
              Metadata = new Metadata {
                  ["genre"] = new("documentary"),
              },
          },
          new Vector 
          { 
              Id = "vec3", 
              Values = new ReadOnlyMemory<float>([0.5f, -1.5f]),
              Metadata = new Metadata {
                  ["genre"] = new("documentary"),
              } 
          }
      }
  });
  Console.WriteLine($"\nUpserted {denseUpsertResponse.UpsertedCount} dense vectors");

  // Upsert records into the sparse index
  var sparseVector1 = new Vector
  {
      Id = "vec1",
      SparseValues = new SparseValues
      {
          Indices = new uint[] { 822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191 },
          Values = new ReadOnlyMemory<float>([1.7958984f, 0.41577148f, 2.828125f, 2.8027344f, 2.8691406f, 1.6533203f, 5.3671875f, 1.3046875f, 0.49780273f, 0.5722656f, 2.71875f, 3.0820312f, 2.5019531f, 4.4414062f, 3.3554688f])
      },
      Metadata = new Metadata {
          ["chunk_text"] = new("AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones."),
          ["category"] = new("technology"),
          ["quarter"] = new("Q3"),
      },
  };

  var sparseVector2 = new Vector
  {
      Id = "vec2",
      SparseValues = new SparseValues
      {
          Indices = new uint[] { 131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697 },
          Values = new ReadOnlyMemory<float>([0.4362793f, 3.3457031f, 2.7714844f, 3.0273438f, 3.3164062f, 5.6015625f, 2.4863281f, 0.38134766f, 1.25f, 2.9609375f, 0.34179688f, 1.4306641f, 0.34375f, 3.3613281f, 1.4404297f, 2.2558594f, 2.2597656f, 4.8710938f, 0.5605469f])
      },
      Metadata = new Metadata {
          ["chunk_text"] = new("Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market."),
          ["category"] = new("technology"),
          ["quarter"] = new("Q4"),
      },
  };

  var sparseVector3 = new Vector
  {
      Id = "vec3",
      SparseValues = new SparseValues
      {
          Indices = new uint[] { 8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697 },
          Values = new ReadOnlyMemory<float>([2.6875f, 4.2929688f, 3.609375f, 3.0722656f, 2.1152344f, 5.78125f, 3.7460938f, 3.7363281f, 1.2695312f, 3.4824219f, 0.7207031f, 0.0826416f, 4.671875f, 3.7011719f, 2.796875f, 0.61621094f])
      },
      Metadata = new Metadata {
          ["chunk_text"] = new("AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production"),
          ["category"] = new("technology"),
          ["quarter"] = new("Q3"),
      },    
  };

  var sparseUpsertResponse = await sparseIndex.UpsertAsync(new UpsertRequest
  {
      Vectors = new List<Vector> { sparseVector1, sparseVector2, sparseVector3 },
      Namespace = "example-namespace"
  });
  Console.WriteLine($"\nUpserted {sparseUpsertResponse.UpsertedCount} sparse vectors");

  // Check the number of records in each index
  var denseIndexStatsResponse = await denseIndex.DescribeIndexStatsAsync(new DescribeIndexStatsRequest());
  Console.WriteLine("\nDense index stats:" + denseIndexStatsResponse);
  var sparseIndexStatsResponse = await sparseIndex.DescribeIndexStatsAsync(new DescribeIndexStatsRequest());
  Console.WriteLine("\nSparse index stats:" + sparseIndexStatsResponse);

  // Query the dense index with a metadata filter
  var denseQueryResponse = await denseIndex.QueryAsync(new QueryRequest
  {
      Vector = new ReadOnlyMemory<float>([3.0f, -2.0f]),
      TopK = 1,
      Namespace = "example-namespace",
      Filter = new Metadata
      {
          ["genre"] = new Metadata
          {
              ["$eq"] = "documentary",
          }
      },
      IncludeValues = false,
      IncludeMetadata = true
  });
  Console.WriteLine("\nDense query response:" + denseQueryResponse);

  // Query the sparse index with a metadata filter
  var sparseQueryResponse = await sparseIndex.QueryAsync(new QueryRequest {
      Namespace = "example-namespace",
      TopK = 1,
      SparseVector = new SparseValues
      {
          Indices = [767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697],
          Values = new[] { 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f },
      },
      Filter = new Metadata
      {
          ["quarter"] = new Metadata
          {
              ["$eq"] = "Q4",
          }
      },
      IncludeValues = false,
      IncludeMetadata = true
  });
  Console.WriteLine("\nSparse query response:" + sparseQueryResponse);

  // Delete the indexes
  await pc.DeleteIndexAsync(denseIndexName);
  await pc.DeleteIndexAsync(sparseIndexName);
  ```

  ```shell curl theme={null}
  PINECONE_LOCAL_HOST="localhost:5080"
  DENSE_INDEX_HOST="localhost:5081"
  SPARSE_INDEX_HOST="localhost:5082"

  # Create two indexes, one dense and one sparse
  curl -X POST "http://$PINECONE_LOCAL_HOST/indexes" \
      -H "Accept: application/json" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "name": "dense-index",
              "vector_type": "dense",
              "dimension": 2,
              "metric": "cosine",
              "spec": {
                  "serverless": {
                      "cloud": "aws",
                      "region": "us-east-1"
                  }
              },
              "tags": {
                  "environment": "development"
              },
              "deletion_protection": "disabled"
          }'

  curl -X POST "http://$PINECONE_LOCAL_HOST/indexes" \
      -H "Accept: application/json" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "name": "sparse-index",
              "vector_type": "sparse",
              "metric": "dotproduct",
              "spec": {
                  "serverless": {
                      "cloud": "aws",
                      "region": "us-east-1"
                  }
              },
              "tags": {
                  "environment": "development"
              },
              "deletion_protection": "disabled"
          }'

  # Upsert records into the dense index
  curl -X POST "http://$DENSE_INDEX_HOST/vectors/upsert" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "namespace": "example-namespace",
              "vectors": [
                  {
                      "id": "vec1", 
                      "values": [1.0, -2.5],
                      "metadata": {"genre": "drama"}
                  },
                  {
                      "id": "vec2", 
                      "values": [3.0, -2.0],
                      "metadata": {"genre": "documentary"}
                  },
                  {
                      "id": "vec3", 
                      "values": [0.5, -1.5],
                      "metadata": {"genre": "documentary"}
                  }
              ]
          }'

  # Upsert records into the sparse index
  curl -X POST "http://$SPARSE_INDEX_HOST/vectors/upsert" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "namespace": "example-namespace",
              "vectors": [
                  {
                      "id": "vec1",
                      "sparseValues": {
                          "values": [1.7958984, 0.41577148, 2.828125, 2.8027344, 2.8691406, 1.6533203, 5.3671875, 1.3046875, 0.49780273, 0.5722656, 2.71875, 3.0820312, 2.5019531, 4.4414062, 3.3554688],
                          "indices": [822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191]
                      },
                      "metadata": {
                          "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
                          "category": "technology",
                          "quarter": "Q3"
                      }
                  },
                  {
                      "id": "vec2",
                      "sparseValues": {
                          "values": [0.4362793, 3.3457031, 2.7714844, 3.0273438, 3.3164062, 5.6015625, 2.4863281, 0.38134766, 1.25, 2.9609375, 0.34179688, 1.4306641, 0.34375, 3.3613281, 1.4404297, 2.2558594, 2.2597656, 4.8710938, 0.5605469],
                          "indices": [131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697]
                      },
                      "metadata": {
                          "chunk_text": "Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
                          "category": "technology",
                          "quarter": "Q4"
                      }
                  },
                  {
                      "id": "vec3",
                      "sparseValues": {
                          "values": [2.6875, 4.2929688, 3.609375, 3.0722656, 2.1152344, 5.78125, 3.7460938, 3.7363281, 1.2695312, 3.4824219, 0.7207031, 0.0826416, 4.671875, 3.7011719, 2.796875, 0.61621094],
                          "indices": [8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697]
                      },
                      "metadata": {
                          "chunk_text": "AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production",
                          "category": "technology",
                          "quarter": "Q3"
                      }
                  }
              ]
          }'

  # Check the number of records in each index
  curl -X POST "http://$DENSE_INDEX_HOST/describe_index_stats" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{}'

  curl -X POST "http://$SPARSE_INDEX_HOST/describe_index_stats" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{}'

  # Query the dense index with a metadata filter
  curl "http://$DENSE_INDEX_HOST/query" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "vector": [3.0, -2.0],
              "filter": {"genre": {"$eq": "documentary"}},
              "topK": 1,
              "includeMetadata": true,
              "includeValues": false,
              "namespace": "example-namespace"
          }'

  # Query the sparse index with a metadata filter
  curl "http://$SPARSE_INDEX_HOST/query" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "sparseVector": {
                  "values": [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],
                  "indices": [767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697]
              },
              "filter": {"quarter": {"$eq": "Q4"}},
              "namespace": "example-namespace",
              "topK": 1,
              "includeMetadata": true,
              "includeValues": false
          }'

  # Delete the index
  curl -X DELETE "http://$PINECONE_LOCAL_HOST/indexes/dense-index" \
      -H "X-Pinecone-Api-Version: 2025-10"

  curl -X DELETE "http://$PINECONE_LOCAL_HOST/indexes/sparse-index" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

## 2. Set up GitHub Actions

[Set up a GitHub Actions workflow](https://docs.github.com/en/actions/writing-workflows/quickstart) to do the following:

1. Pull the Pinecone Local Docker image.
2. Start a Pinecone Local instance for each test run.
3. Execute tests against the local instance.
4. Tear down the instance after tests complete.

Here's a sample GitHub Actions workflow that you can extend for your own needs:

```yaml  theme={null}
name: CI/CD with Pinecone Local
on:
  pull_request:
    branches:
      - main
  push:
    branches:
      - main

jobs:
  pc-local-tests:
    name: Pinecone Local tests
    runs-on: ubuntu-latest
    services:
      pc-local:
        image: ghcr.io/pinecone-io/pinecone-local:latest
        env:
          PORT: 5080
        ports:
          - "5080-6000:5080-6000"
    steps:
      - name: Check out repository code
        uses: actions/checkout@v4

      - name: Install dependencies
        run: |
          python -m pip install --upgrade pip
          pip install "pinecone[grpc]"

      - name: Run tests
        run: |
          pytest test/
```

## 3. Run your tests

GitHub Actions will automaticaly run your tests against Pinecone Local when the events you specified in your workflow occur.

For a list of the events that can trigger a workflow and more details about using GitHub Actions for CI/CD, see the [GitHub Actions documentation](https://docs.github.com/en/actions).

---

# Source: https://docs.pinecone.io/integrations/aws-marketplace.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# AWS Marketplace

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Access Pinecone through our AWS Marketplace listing. AWS Marketplace allows you to manage Pinecone and other third-party software from a centralized location, and simplifies software licensing and procurement with flexible pricing options and multiple deployment methods.

You can set up pay-as-you-go billing for a Pinecone organization through the AWS Marketplace.

<PrimarySecondaryCTA
  primaryLabel={"Get started"}
  primaryHref={
  "https://aws.amazon.com/marketplace/pp/prodview-xhgyscinlz4jk?trk=44c6de22-4392-4c04-9465-644542cd14a8&sc_channel=el"
}
  secondaryLabel={"View setup guide"}
  secondaryHref={
  "/guides/organizations/manage-billing/upgrade-billing-plan"
}
/>

---

# Source: https://docs.pinecone.io/guides/indexes/pods/back-up-a-pod-based-index.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Back up a pod-based index

> Backup pod-based indexes using Pinecone collections

<Warning>
  Customers who sign up for a Standard or Enterprise plan on or after August 18, 2025 cannot create pod-based indexes. Instead, create [serverless indexes](/guides/index-data/create-an-index), and consider using [dedicated read nodes](/guides/index-data/dedicated-read-nodes) for large workloads (millions of records or more, and moderate or high query rates).
</Warning>

This page describes how to create a static copy of a pod-based index, also known as a [collection](/guides/indexes/pods/understanding-collections).

## Create a collection

To create a backup of your pod-based index, use the [`create_collection`](/reference/api/latest/control-plane/create_collection) operation.

The following example creates a [collection](/guides/indexes/pods/understanding-collections) named `example-collection` from an index named `docs-example`:

<CodeGroup>
  ```python Python theme={null}
  # pip install "pinecone[grpc]"
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="API_KEY")
  pc.create_collection("example-collection", "docs-example")
  ```

  ```javascript JavaScript theme={null}
  // npm install @pinecone-database/pinecone
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({
    apiKey: 'YOUR_API_KEY'
  });

  await pc.createCollection({
    name: "example-collection",
    source: "docs-example",
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;

  public class CreateCollectionExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();

          pc.createCollection("example-collection", "docs-example");
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      collection, err := pc.CreateCollection(ctx, &pinecone.CreateCollectionRequest{
          Name: "example-collection", 
          Source: "docs-example",
      })
      if err != nil {
          log.Fatalf("Failed to create collection: %v", err)
      } else {
          fmt.Printf("Successfully created collection: %v", collection.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var collectionModel = await pinecone.CreateCollectionAsync(new CreateCollectionRequest {
      Name = "example-collection",
      Source = "docs-example",
  });
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s POST "https://api.pinecone.io/collections" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "name": "example-collection",
          "source": "docs-example"
    }'
  ```
</CodeGroup>

<Tip>
  You can create a collection using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/backups).
</Tip>

## Check the status of a collection

To retrieve the status of the process creating a collection and the size of the collection, use the [`describe_collection`](/reference/api/latest/control-plane/describe_collection) operation. Specify the name of the collection to check. You can only call `describe_collection` on a collection in the current project.

The `describe_collection` operation returns an object containing key-value pairs representing the name of the collection, the size in bytes, and the creation status of the collection.

The following example gets the creation status and size of a collection named `example-collection`.

<CodeGroup>
  ```python Python theme={null}
  # pip install "pinecone[grpc]"
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key='API_KEY')
  pc.describe_collection(name="example-collection")
  ```

  ```javascript JavaScript theme={null}
  // npm install @pinecone-database/pinecone
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await pc.describeCollection('example-collection');
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.client.model.CollectionModel;

  public class DescribeCollectionExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          CollectionModel collectionModel = pc.describeCollection("example-collection");
          System.out.println(collectionModel);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      collectionName := "example-collection"

      collection, err := pc.DescribeCollection(ctx, collectionName)
      if err != nil {
          log.Fatalf("Error describing collection %v: %v", collectionName, err)
      } else {
          fmt.Printf("Collection: %+v", collection)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var collectionModel = await pinecone.DescribeCollectionAsync("example-collection");

  Console.WriteLine(collectionModel);
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X GET "https://api.pinecone.io/collections/example-collection" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

<Tip>
  You can check the status of a collection using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/backups).
</Tip>

## List your collections

To get a list of the collections in the current project, use the [`list_collections`](/reference/api/latest/control-plane/list_collections) operation.

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key='API_KEY')
  pc.list_collections()
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  await pc.listCollections();
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.client.model.CollectionModel;

  public class ListCollectionsExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          List<CollectionModel> collectionList = pc.listCollections().getCollections();
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
      bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      collections, err := pc.ListCollections(ctx)
      if err != nil {
  	    log.Fatalf("Failed to list collections: %v", err)
      } else {
          if len(collections) == 0 {
              fmt.Printf("No collections found in project")
          } else {
              for _, collection := range collections {
                  fmt.Printf("collection: %v\n", prettifyStruct(collection))
              }
          }
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var collectionList = await pinecone.ListCollectionsAsync();

  Console.WriteLine(collectionList);
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X GET "https://api.pinecone.io/collections" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```

  <Tip>
    You can view a list of your collections using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/backups).
  </Tip>
</CodeGroup>

<Tip>
  You can view a list of your collections using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/backups).
</Tip>

## Delete a collection

To delete a collection, use the [`delete_collection`](/reference/api/latest/control-plane/delete_collection) operation. Specify the name of the collection to delete.

Deleting the collection takes several minutes. During this time, the [`describe_collection`](#check-the-status-of-a-collection) operation returns the status "deleting".

<CodeGroup>
  ```python Python theme={null}
  # pip install "pinecone[grpc]"
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key='API_KEY')
  pc.delete_collection("example-collection")
  ```

  ```javascript JavaScript theme={null}
  // npm install @pinecone-database/pinecone
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  await pc.deleteCollection("example-collection");
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;

  public class DeleteCollectionExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          pc.deleteCollection("example-collection");
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      collectionName := "example-collection"

      err = pc.DeleteCollection(ctx, collectionName)
      if err != nil {
  	    log.Fatalf("Failed to delete collection: %v\n", err)
      } else {
          if len(collections) == 0 {
              fmt.Printf("No collections found in project")
          } else {
              fmt.Printf("Successfully deleted collection \"%v\"\n", collectionName)
          }
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  await pinecone.DeleteCollectionAsync("example-collection");
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X DELETE "https://api.pinecone.io/collections/example-collection" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

<Tip>
  You can delete a collection using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/backups).
</Tip>

---

# Source: https://docs.pinecone.io/guides/manage-data/back-up-an-index.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Back up an index

> Create backups of serverless indexes for protection

## Create a backup

You can [create a backup from a serverless index](/reference/api/latest/control-plane/create_backup) as follows.

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  backup = pc.create_backup(
      index_name="docs-example", 
      backup_name="example-backup", 
      description="Monthly backup of production index"
  )

  print(backup)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  const backup = await pc.createBackup({
    indexName: 'docs-example',
    name: 'example-backup',
    description: 'Monthly backup of production index',
  });

  console.log(backup);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.ApiException;
  import org.openapitools.db_control.client.model.*;

  public class CreateBackup {
      public static void main(String[] args) throws ApiException {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();

          String indexName = "docs-example";
          String backupName = "example-backup";
          String backupDescription = "Monthly backup of production index";

          BackupModel backup = pc.createBackup(indexName,backupName, backupDescription);

          System.out.println(backup);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
      bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      indexName := "docs-example"
      backupName := "example-backup"
      backupDesc := "Monthly backup of production index"

      backup, err := pc.CreateBackup(ctx, &pinecone.CreateBackupParams{
          IndexName:   indexName,
          Name:        &backupName,
          Description: &backupDesc,
      })
      if err != nil {
          log.Fatalf("Failed to create backup: %v", err)
      }

      fmt.Printf(prettifyStruct(backup))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("PINECONE_API_KEY");

  var backup = await pinecone.Backups.BackupIndexAsync(
      "docs-example", 
      new BackupIndexRequest
      {
          Name = "example-backup",
          Description = "Monthly backup of production index"
      }
  );

  Console.WriteLine(backup);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_NAME="docs-example"

  curl "https://api.pinecone.io/indexes/$INDEX_NAME/backups" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H 'accept: application/json' \
      -H 'Content-Type: application/json' \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
        "name": "example-backup", 
        "description": "Monthly backup of production index"
        }'
  ```
</CodeGroup>

The example returns a response like the following:

<CodeGroup>
  ```python Python theme={null}
  {'backup_id': '8c85e612-ed1c-4f97-9f8c-8194e07bcf71',
   'cloud': 'aws',
   'created_at': '2025-05-15T00:52:10.809305882Z',
   'description': 'Monthly backup of production index',
   'dimension': 1024,
   'name': 'example-backup',
   'namespace_count': 3,
   'record_count': 98,
   'region': 'us-east-1',
   'size_bytes': 1069169,
   'source_index_id': 'f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74',
   'source_index_name': 'docs-example',
   'status': 'Ready',
   'tags': {}}
  ```

  ```javascript JavaScript theme={null}
  {
    backupId: '8c85e612-ed1c-4f97-9f8c-8194e07bcf71',
    sourceIndexName: 'docs-example',
    sourceIndexId: 'f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74',
    name: 'example-backup',
    description: 'Monthly backup of production index',
    status: 'Ready',
    cloud: 'aws',
    region: 'us-east-1',
    dimension: 1024,
    metric: undefined,
    recordCount: 98,
    namespaceCount: 3,
    sizeBytes: 1069169,
    tags: {},
    createdAt: '2025-05-14T16:37:25.625540Z'
  }
  ```

  ```java Java theme={null}
  class BackupModel {
      backupId: 0d75b99f-be61-4a93-905e-77201286c02e
      sourceIndexName: docs-example
      sourceIndexId: f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74
      name: example-backup
      description: Monthly backup of production index
      status: Initializing
      cloud: aws
      region: us-east-1
      dimension: null
      metric: null
      recordCount: null
      namespaceCount: null
      sizeBytes: null
      tags: {}
      createdAt: 2025-05-16T19:42:23.804787550Z
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  {
    "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
    "cloud": "aws",
    "created_at": "2025-05-15T00:52:10.809305882Z",
    "description": "Monthly backup of production index",
    "dimension": 1024,
    "name": "example-backup",
    "region": "us-east-1",
    "source_index_id": "f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
    "source_index_name": "docs-example",
    "status": "Initializing",
    "tags": {}
  }
  ```

  ```csharp C# theme={null}
  {
      "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
      "source_index_name": "docs-example",
      "source_index_id": "f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
      "name": "example-backup",
      "description": "Monthly backup of production index",
      "status": "Ready",
      "cloud": "aws",
      "region": "us-east-1",
      "tags": {},
      "created_at": "2025-05-15T00:52:10.809305882Z"
  }
  ```

  ```json curl theme={null}
  {
    "backup_id":"8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
    "source_index_id":"f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
    "source_index_name":"docs-example",
    "tags":{},
    "name":"example-backup",
    "description":"Monthly backup of production index",
    "status":"Ready",
    "cloud":"aws",
    "region":"us-east-1",
    "dimension":1024,
    "record_count":96,
    "namespace_count":3,
    "size_bytes":1069169,
    "created_at":"2025-05-14T16:37:25.625540Z"
    }
  ```
</CodeGroup>

<Tip>
  You can create a backup using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/backups).
</Tip>

## Describe a backup

You can [view the details of a backup](/reference/api/latest/control-plane/describe_backup) as follows.

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  backup = pc.describe_backup(backup_id="8c85e612-ed1c-4f97-9f8c-8194e07bcf71")

  print(backup)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  const backupDesc = await pc.describeBackup('8c85e612-ed1c-4f97-9f8c-8194e07bcf71');

  console.log(backupDesc);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.ApiException;
  import org.openapitools.db_control.client.model.*;

  public class CreateBackup {
      public static void main(String[] args) throws ApiException {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();

          BackupModel backupModel = pc.describeBackup("8c85e612-ed1c-4f97-9f8c-8194e07bcf71");

          System.out.println(backupModel);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
  	ctx := context.Background()

  	pc, err := pinecone.NewClient(pinecone.NewClientParams{
  		ApiKey: "YOUR_API_KEY",
  	})
  	if err != nil {
  		log.Fatalf("Failed to create Client: %v", err)
  	}

  	backup, err := pc.DescribeBackup(ctx, "8c85e612-ed1c-4f97-9f8c-8194e07bcf71")
  	if err != nil {
  		log.Fatalf("Failed to describe backup: %v", err)
  	}
  	fmt.Printf(prettifyStruct(backup))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("PINECONE_API_KEY");

  var backup = await pinecone.Backups.GetAsync("8c85e612-ed1c-4f97-9f8c-8194e07bcf71");

  Console.WriteLine(backup);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  BACKUP_ID="8c85e612-ed1c-4f97-9f8c-8194e07bcf71"

  curl -X GET "https://api.pinecone.io/backups/$BACKUP_ID" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -H "accept: application/json"
  ```
</CodeGroup>

The example returns a response like the following:

<CodeGroup>
  ```python Python theme={null}
  {'backup_id': '8c85e612-ed1c-4f97-9f8c-8194e07bcf71',
   'cloud': 'aws',
   'created_at': '2025-05-15T00:52:10.809354Z',
   'description': 'Monthly backup of production index',
   'dimension': 1024,
   'name': 'example-backup',
   'namespace_count': 3,
   'record_count': 98,
   'region': 'us-east-1',
   'size_bytes': 1069169,
   'source_index_id': 'f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74',
   'source_index_name': 'docs-example',
   'status': 'Ready',
   'tags': {}}
  ```

  ```javascript JavaScript theme={null}
  {
    backupId: '8c85e612-ed1c-4f97-9f8c-8194e07bcf71',
    sourceIndexName: 'docs-example',
    sourceIndexId: 'f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74',
    name: 'example-backup',
    description: 'Monthly backup of production index',
    status: 'Ready',
    cloud: 'aws',
    region: 'us-east-1',
    dimension: 1024,
    metric: undefined,
    recordCount: 98,
    namespaceCount: 3,
    sizeBytes: 1069169,
    tags: {},
    createdAt: '2025-05-14T16:37:25.625540Z'
  }
  ```

  ```java Java theme={null}
  class BackupList {
      data: [class BackupModel {
          backupId: 95707edb-e482-49cf-b5a5-312219a51a97
          sourceIndexName: docs-example
          sourceIndexId: f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74
          name: example-backup
          description: Monthly backup of production index
          status: Initializing
          cloud: aws
          region: us-east-1
          dimension: null
          metric: null
          recordCount: null
          namespaceCount: null
          sizeBytes: null
          tags: {}
          createdAt: 2025-05-16T19:46:26.248428Z
          additionalProperties: null
      }]
      pagination: null
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  {
    "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
    "cloud": "aws",
    "created_at": "2025-05-15T00:52:10.809354Z",
    "description": "Monthly backup of production index",
    "dimension": 1024,
    "name": "example-backup",
    "namespace_count": 3,
    "record_count": 98,
    "region": "us-east-1",
    "size_bytes": 1069169,
    "source_index_id": "f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
    "source_index_name": "docs-example",
    "status": "Ready",
    "tags": {}
  }
  ```

  ```csharp C# theme={null}
  {
    "backup_id": "95707edb-e482-49cf-b5a5-312219a51a97",
    "source_index_name": "docs-example",
    "source_index_id": "f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
    "name": "example-backup",
    "description": "Monthly backup of production index",
    "status": "Ready",
    "cloud": "aws",
    "region": "us-east-1",
    "dimension": 1024,
    "record_count": 97,
    "namespace_count": 2,
    "size_bytes": 1069169,
    "tags": {},
    "created_at": "2025-05-15T00:52:10.809354Z"
  }
  ```

  ```json curl theme={null}
  {
    "backup_id":"8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
    "source_index_id":"f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
    "source_index_name":"docs-example",
    "tags":{},
    "name":"example-backup",
    "description":"Monthly backup of production index",
    "status":"Ready",
    "cloud":"aws",
    "region":"us-east-1",
    "dimension":1024,
    "record_count":98,
    "namespace_count":3,
    "size_bytes":1069169,
    "created_at":"2025-03-11T18:29:50.549505Z"
  }
  ```
</CodeGroup>

<Tip>
  You can view backup details using the [Pinecone console](https://app.pinecone.io/organizations/-/projects-/backups).
</Tip>

## List backups for an index

You can [list backups for a specific index](/reference/api/latest/control-plane/list_index_backups) as follows.

Up to 100 backups are returned at a time by default, in sorted order (bitwise “C” collation). If the `limit` parameter is set, up to that number of backups are returned instead. Whenever there are additional backups to return, the response also includes a `pagination_token` that you can use to get the next batch of backups. When the response does not include a `pagination_token`, there are no more backups to return.

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  index_backups = pc.list_backups(index_name="docs-example")

  print(index_backups)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  const indexBackups = await pc.listBackups({ indexName: 'docs-example' });

  console.log(indexBackups);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.ApiException;
  import org.openapitools.db_control.client.model.*;

  public class CreateBackup {
      public static void main(String[] args) throws ApiException {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();

          String indexName = "docs-example";
          BackupList indexBackupList = pc.listIndexBackups(indexName);

          System.out.println(indexBackupList);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
  	ctx := context.Background()

  	pc, err := pinecone.NewClient(pinecone.NewClientParams{
  		ApiKey: "YOUR_API_KEY",
  	})
  	if err != nil {
  		log.Fatalf("Failed to create Client: %v", err)
  	}

  	indexName := "docs-example"
  	limit := 2
  	indexBackups, err := pc.ListBackups(ctx, &pinecone.ListBackupsParams{
  		Limit: &limit,
  		IndexName: &indexName,
  	})
  	if err != nil {
  		log.Fatalf("Failed to list backups: %v", err)
  	}
  	fmt.Printf(prettifyStruct(indexBackups))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var indexBackups = await pinecone.Backups.ListByIndexAsync( "docs-example", new ListBackupsByIndexRequest());

  Console.WriteLine(indexBackups);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_NAME="docs-example"

  curl -X GET "https://api.pinecone.io/indexes/$INDEX_NAME/backups" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -H "accept: application/json"
  ```
</CodeGroup>

The example returns a response like the following:

<CodeGroup>
  ```python Python theme={null}
  [{
      "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
      "source_index_name": "docs-example",
      "source_index_id": "f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
      "status": "Ready",
      "cloud": "aws",
      "region": "us-east-1",
      "tags": {},
      "name": "example-backup",
      "description": "Monthly backup of production index",
      "dimension": 1024,
      "record_count": 98,
      "namespace_count": 3,
      "size_bytes": 1069169,
      "created_at": "2025-05-15T00:52:10.809305882Z"
  }]
  ```

  ```javascript JavaScript theme={null}
  {
    data: [
      {
        backupId: '8c85e612-ed1c-4f97-9f8c-8194e07bcf71',
        sourceIndexName: 'docs-example',
        sourceIndexId: 'f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74',
        name: 'example-backup',
        description: 'Monthly backup of production index',
        status: 'Ready',
        cloud: 'aws',
        region: 'us-east-1',
        dimension: 1024,
        metric: undefined,
        recordCount: 98,
        namespaceCount: 3,
        sizeBytes: 1069169,
        tags: {},
        createdAt: '2025-05-14T16:37:25.625540Z'
      }
    ],
    pagination: undefined
  }
  ```

  ```java Java theme={null}
  class BackupList {
      data: [class BackupModel {
          backupId: 8c85e612-ed1c-4f97-9f8c-8194e07bcf71
          sourceIndexName: docs-example
          sourceIndexId: f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74
          name: example-backup
          description: Monthly backup of production index
          status: Initializing
          cloud: aws
          region: us-east-1
          dimension: null
          metric: null
          recordCount: null
          namespaceCount: null
          sizeBytes: null
          tags: {}
          createdAt: 2025-05-16T19:46:26.248428Z
          additionalProperties: null
      }]
      pagination: null
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  {
    "data": [
      {
        "backup_id": "bf2cda5d-b233-4a0a-aae9-b592780ad3ff",
        "cloud": "aws",
        "created_at": "2025-05-16T18:01:51.531129Z",
        "description": "Monthly backup of production index",
        "dimension": 0,
        "name": "example-backup",
        "namespace_count": 1,
        "record_count": 96,
        "region": "us-east-1",
        "size_bytes": 86393,
        "source_index_id": "bcb5b3c9-903e-4cb6-8b37-a6072aeb874f",
        "source_index_name": "docs-example",
        "status": "Ready",
        "tags": {}
      },
      {
        "backup_id": "e12269b0-a29b-4af0-9729-c7771dec03e3",
        "cloud": "aws",
        "created_at": "2025-05-14T17:00:45.803146Z",
        "dimension": 0,
        "name": "example-backup2",
        "namespace_count": 1,
        "record_count": 96,
        "region": "us-east-1",
        "size_bytes": 86393,
        "source_index_id": "bcb5b3c9-903e-4cb6-8b37-a6072aeb874f",
        "source_index_name": "docs-example",
        "status": "Ready"
      }
    ],
    "pagination": {
      "next": "eyJsaW1pdCI6Miwib2Zmc2V0IjoyfQ=="
    }
  }
  ```

  ```csharp C# theme={null}
  {
    "data":
    [
      {
        "backup_id":"9947520e-d5a1-4418-a78d-9f464c9969da",
        "source_index_id":"8433941a-dae7-43b5-ac2c-d3dab4a56b2b",
        "source_index_name":"docs-example",
        "tags":{},
        "name":"example-backup",
        "description":"Monthly backup of production index",
        "status":"Pending",
        "cloud":"aws",
        "region":"us-east-1",
        "dimension":1024,
        "record_count":98,
        "namespace_count":3,
        "size_bytes":1069169,
        "created_at":"2025-03-11T18:29:50.549505Z"
      }
    ]
  }
  ```

  ```json curl theme={null}
  {
    "data":
    [
      {
        "backup_id":"9947520e-d5a1-4418-a78d-9f464c9969da",
        "source_index_id":"8433941a-dae7-43b5-ac2c-d3dab4a56b2b",
        "source_index_name":"docs-example",
        "tags":{},
        "name":"example-backup",
        "description":"Monthly backup of production index",
        "status":"Pending",
        "cloud":"aws",
        "region":"us-east-1",
        "dimension":1024,
        "record_count":98,
        "namespace_count":3,
        "size_bytes":1069169,
        "created_at":"2025-03-11T18:29:50.549505Z"
        }
      ],
    "pagination":null
  }
  ```
</CodeGroup>

<Tip>
  You can view the backups for a specific index from either the [Backups](https://app.pinecone.io/organizations/-/projects/-/backups) tab or the [Indexes](https://app.pinecone.io/organizations/-/projects/-/indexes) tab in the Pinecone console.
</Tip>

## List backups in a project

You can [list backups for all indexes in a project](/reference/api/latest/control-plane/list_project_backups) as follows.

Up to 100 backups are returned at a time by default, in sorted order (bitwise “C” collation). If the `limit` parameter is set, up to that number of backups are returned instead. Whenever there are additional backups to return, the response also includes a `pagination_token` that you can use to get the next batch of backups. When the response does not include a `pagination_token`, there are no more backups to return.

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  project_backups = pc.list_backups()

  print(project_backups)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  const projectBackups = await pc.listBackups();

  console.log(projectBackups);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.ApiException;
  import org.openapitools.db_control.client.model.*;

  public class CreateBackup {
      public static void main(String[] args) throws ApiException {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();

          String indexName = "docs-example";
          BackupList projectBackupList = pc.listProjectBackups();

          System.out.println(projectBackupList);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
  	ctx := context.Background()

  	pc, err := pinecone.NewClient(pinecone.NewClientParams{
  		ApiKey: "YOUR_API_KEY",
  	})
  	if err != nil {
  		log.Fatalf("Failed to create Client: %v", err)
  	}

  	limit := 3
  	backups, err := pc.ListBackups(ctx, &pinecone.ListBackupsParams{
  		Limit:     &limit,
  	})
  	if err != nil {
  		log.Fatalf("Failed to list backups: %v", err)
  	}
  	fmt.Printf(prettifyStruct(backups))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("PINECONE_API_KEY");

  var backups = await pinecone.Backups.ListAsync();

  Console.WriteLine(backups);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -X GET "https://api.pinecone.io/backups" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -H "accept: application/json"
  ```
</CodeGroup>

The example returns a response like the following:

<CodeGroup>
  ```python Python theme={null}
  [{
      "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
      "source_index_name": "docs-example",
      "source_index_id": "f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
      "status": "Ready",
      "cloud": "aws",
      "region": "us-east-1",
      "tags": {},
      "name": "example-backup",
      "description": "Monthly backup of production index",
      "dimension": 1024,
      "record_count": 98,
      "namespace_count": 3,
      "size_bytes": 1069169,
      "created_at": "2025-05-15T20:26:21.248515Z"
  }, {
      "backup_id": "95707edb-e482-49cf-b5a5-312219a51a97",
      "source_index_name": "docs-example2",
      "source_index_id": "b49f27d1-1bf3-49c6-82b5-4ae46f00f0e6",
      "status": "Ready",
      "cloud": "aws",
      "region": "us-east-1",
      "tags": {},
      "name": "example-backup2",
      "description": "Monthly backup of production index",
      "dimension": 1024,
      "record_count": 97,
      "namespace_count": 2,
      "size_bytes": 1069169,
      "created_at": "2025-05-15T00:52:10.809354Z"
  }, {
      "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
      "source_index_name": "docs-example3",
      "source_index_id": "f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
      "status": "Ready",
      "cloud": "aws",
      "region": "us-east-1",
      "tags": {},
      "name": "example-backup3",
      "description": "Monthly backup of production index",
      "dimension": 1024,
      "record_count": 98,
      "namespace_count": 3,
      "size_bytes": 1069169,
      "created_at": "2025-05-14T16:37:25.625540Z"
  }]
  ```

  ```javascript JavaScript theme={null}
  {
    data: [
      {
        backupId: 'e12269b0-a29b-4af0-9729-c7771dec03e3',
        sourceIndexName: 'docs-example',
        sourceIndexId: 'bcb5b3c9-903e-4cb6-8b37-a6072aeb874f',
        name: 'example-backup',
        description: undefined,
        status: 'Ready',
        cloud: 'aws',
        region: 'us-east-1',
        dimension: 0,
        metric: undefined,
        recordCount: 96,
        namespaceCount: 1,
        sizeBytes: 86393,
        tags: undefined,
        createdAt: '2025-05-14T17:00:45.803146Z'
      },
      {
        backupId: 'd686451d-1ede-4004-9f72-7d22cc799b6e',
        sourceIndexName: 'docs-example2',
        sourceIndexId: 'b49f27d1-1bf3-49c6-82b5-4ae46f00f0e6',
        name: 'example-backup2',
        description: undefined,
        status: 'Ready',
        cloud: 'aws',
        region: 'us-east-1',
        dimension: 1024,
        metric: undefined,
        recordCount: 50,
        namespaceCount: 1,
        sizeBytes: 545171,
        tags: undefined,
        createdAt: '2025-05-14T17:00:34.814371Z'
      },
      {
        backupId: '8c85e612-ed1c-4f97-9f8c-8194e07bcf71',
        sourceIndexName: 'docs-example3',
        sourceIndexId: 'f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74',
        name: 'example-backup3',
        description: 'Monthly backup of production index',
        status: 'Ready',
        cloud: 'aws',
        region: 'us-east-1',
        dimension: 1024,
        metric: undefined,
        recordCount: 98,
        namespaceCount: 3,
        sizeBytes: 1069169,
        tags: {},
        createdAt: '2025-05-14T16:37:25.625540Z'
      }
    ],
    pagination: undefined
  }
  ```

  ```java Java theme={null}
  class BackupList {
      data: [class BackupModel {
          backupId: 13761d20-7a0b-4778-ac27-36dd91c4be43
          sourceIndexName: example-dense-index
          sourceIndexId: f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74
          name: example-backup
          description: Monthly backup of production index
          status: Initializing
          cloud: aws
          region: us-east-1
          dimension: null
          metric: null
          recordCount: null
          namespaceCount: null
          sizeBytes: null
          tags: {}
          createdAt: 2025-05-16T19:46:26.248428Z
          additionalProperties: null
      }, class BackupModel {
          backupId: 0d75b99f-be61-4a93-905e-77201286c02e
          sourceIndexName: example-dense-index
          sourceIndexId: f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74
          name: example-backup2
          description: Monthly backup of production index
          status: Initializing
          cloud: aws
          region: us-east-1
          dimension: null
          metric: null
          recordCount: null
          namespaceCount: null
          sizeBytes: null
          tags: {}
          createdAt: 2025-05-16T19:42:23.804820Z
          additionalProperties: null
      }, class BackupModel {
          backupId: bf2cda5d-b233-4a0a-aae9-b592780ad3ff
          sourceIndexName: example-sparse-index
          sourceIndexId: bcb5b3c9-903e-4cb6-8b37-a6072aeb874f
          name: example-backup3
          description: Monthly backup of production index
          status: Ready
          cloud: aws
          region: us-east-1
          dimension: 0
          metric: null
          recordCount: 96
          namespaceCount: 1
          sizeBytes: 86393
          tags: {}
          createdAt: 2025-05-16T18:01:51.531129Z
          additionalProperties: null
      }]
      pagination: null
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  {
    "data": [
      {
        "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
        "cloud": "aws",
        "created_at": "2025-05-15T00:52:10.809305882Z",
        "description": "Monthly backup of production index",
        "dimension": 1024,
        "name": "example-backup",
        "namespace_count": 3,
        "record_count": 98,
        "region": "us-east-1",
        "size_bytes": 1069169,
        "source_index_id": "f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
        "source_index_name": "docs-example",
        "status": "Ready",
        "tags": {}
      },
      {
        "backup_id": "bf2cda5d-b233-4a0a-aae9-b592780ad3ff",
        "cloud": "aws",
        "created_at": "2025-05-15T00:52:10.809305882Z",
        "description": "Monthly backup of production index",
        "dimension": 0,
        "name": "example-backup2",
        "namespace_count": 1,
        "record_count": 96,
        "region": "us-east-1",
        "size_bytes": 86393,
        "source_index_id": "bcb5b3c9-903e-4cb6-8b37-a6072aeb874f",
        "source_index_name": "example-sparse-index",
        "status": "Ready",
        "tags": {}
      },
      {
        "backup_id": "f73028f6-1746-410e-ab6d-9dd2519df4de",
        "cloud": "aws",
        "created_at": "2025-05-15T20:26:21.248515Z",
        "description": "Monthly backup of production index",
        "dimension": 1024,
        "name": "example-backup3",
        "namespace_count": 2,
        "record_count": 97,
        "region": "us-east-1",
        "size_bytes": 1069169,
        "source_index_id": "f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
        "source_index_name": "example-dense-index",
        "status": "Ready",
        "tags": {}
      }
    ],
    "pagination": {
      "next": "eyJsaW1pdCI6Miwib2Zmc2V0IjoyfQ=="
    }  
  }
  ```

  ```csharp C# theme={null}
  {
    "data": [
      {
        "backup_id": "95707edb-e482-49cf-b5a5-312219a51a97",
        "source_index_name": "docs-example",
        "source_index_id": "f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
        "name": "example-backup",
        "description": "Monthly backup of production index",
        "status": "Ready",
        "cloud": "aws",
        "region": "us-east-1",
        "dimension": 1024,
        "record_count": 97,
        "namespace_count": 2,
        "size_bytes": 1069169,
        "tags": {},
        "created_at": "2025-05-15T00:52:10.809354Z"
      },
      {
        "backup_id": "e12269b0-a29b-4af0-9729-c7771dec03e3",
        "source_index_name": "docs-example2",
        "source_index_id": "bcb5b3c9-903e-4cb6-8b37-a6072aeb874f",
        "name": "example-backup2",
        "status": "Ready",
        "cloud": "aws",
        "region": "us-east-1",
        "dimension": 0,
        "record_count": 96,
        "namespace_count": 1,
        "size_bytes": 86393,
        "created_at": "2025-05-14T17:00:45.803146Z"
      },
      {
        "backup_id": "d686451d-1ede-4004-9f72-7d22cc799b6e",
        "source_index_name": "docs-example3",
        "source_index_id": "b49f27d1-1bf3-49c6-82b5-4ae46f00f0e6",
        "name": "example-backup3",
        "status": "Ready",
        "cloud": "aws",
        "region": "us-east-1",
        "dimension": 1024,
        "record_count": 50,
        "namespace_count": 1,
        "size_bytes": 545171,
        "created_at": "2025-05-14T17:00:34.814371Z"
      }
    ]
  }
  ```

  ```json curl  theme={null}
  {
    "data": [
      {
        "backup_id": "e12269b0-a29b-4af0-9729-c7771dec03e3",
        "source_index_id": "bcb5b3c9-903e-4cb6-8b37-a6072aeb874f",
        "source_index_name": "docs-example",
        "tags": null,
        "name": "example-backup",
        "description": null,
        "status": "Ready",
        "cloud": "aws",
        "region": "us-east-1",
        "dimension": 0,
        "record_count": 96,
        "namespace_count": 1,
        "size_bytes": 86393,
        "created_at": "2025-05-14T17:00:45.803146Z"
      },
      {
        "backup_id": "d686451d-1ede-4004-9f72-7d22cc799b6e",
        "source_index_id": "b49f27d1-1bf3-49c6-82b5-4ae46f00f0e6",
        "source_index_name": "docs-example2",
        "tags": null,
        "name": "example-backup2",
        "description": null,
        "status": "Ready",
        "cloud": "aws",
        "region": "us-east-1",
        "dimension": 1024,
        "record_count": 50,
        "namespace_count": 1,
        "size_bytes": 545171,
        "created_at": "2025-05-14T17:00:34.814371Z"
      },
      {
        "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
        "source_index_id": "f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
        "source_index_name": "docs-example3",
        "tags": {},
        "name": "example-backup3",
        "description": "Monthly backup of production index",
        "status": "Ready",
        "cloud": "aws",
        "region": "us-east-1",
        "dimension": 1024,
        "record_count": 98,
        "namespace_count": 3,
        "size_bytes": 1069169,
        "created_at": "2025-05-14T16:37:25.625540Z"
      }
    ],
    "pagination": null
  }
  ```
</CodeGroup>

<Tip>
  You can view all backups in a project using the [Pinecone console](https://app.pinecone.io/organizations/-/projects-/backups).
</Tip>

## Delete a backup

You can [delete a backup](/reference/api/latest/control-plane/delete_backup) as follows.

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.delete_backup(backup_id="9947520e-d5a1-4418-a78d-9f464c9969da")
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  await pc.deleteBackup('9947520e-d5a1-4418-a78d-9f464c9969da');
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.ApiException;
  import org.openapitools.db_control.client.model.*;

  public class CreateBackup {
      public static void main(String[] args) throws ApiException {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();

          pc.deleteBackup("9947520e-d5a1-4418-a78d-9f464c9969da");
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
  	ctx := context.Background()

  	pc, err := pinecone.NewClient(pinecone.NewClientParams{
  		ApiKey: "YOUR_API_KEY",
  	})
  	if err != nil {
  		log.Fatalf("Failed to create Client: %v", err)
  	}

  	err = pc.DeleteBackup(ctx, "8c85e612-ed1c-4f97-9f8c-8194e07bcf71")
  	if err != nil {
  		log.Fatalf("Failed to delete backup: %v", err)
  	} else {
  		fmt.Println("Backup deleted successfully")
  	}
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("PINECONE_API_KEY");

  await pinecone.Backups.DeleteAsync("9947520e-d5a1-4418-a78d-9f464c9969da");
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  BACKUP_ID="9947520e-d5a1-4418-a78d-9f464c9969da"

  curl -X DELETE "https://api.pinecone.io/backups/$BACKUP_ID" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

<Tip>
  You can delete a backup using the [Pinecone console](https://app.pinecone.io/organizations/-/projects-/backups).
</Tip>

---

# Source: https://docs.pinecone.io/guides/manage-data/backups-overview.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Backups overview

> Learn about backups of serverless indexes in Pinecone.

A backup is a static copy of a serverless [index](/guides/index-data/indexing-overview) that only consumes storage. It is a non-queryable representation of a set of records. You can [create a backup](/guides/manage-data/back-up-an-index) of a serverless index, and you can [create a new serverless index from a backup](/guides/manage-data/restore-an-index). This allows you to restore the index with the same or different configurations.

## Use cases

Creating a backup is useful when performing tasks like the following:

* Protecting an index from manual or system failures.
* Temporarily shutting down an index.
* Copying the data from one index into a different index.
* Making a backup of your index.
* Experimenting with different index configurations.

## Performance

Backup and restore times depend upon the size of the index and number of namespaces:

* For less than 1M vectors in a namespace, backups and restores take approximately 10 minutes.
* For 100,000,000 vectors, backups and restores can take up to 5 hours.

## Quotas

| Metric              | Starter plan | Standard plan | Enterprise plan |
| :------------------ | :----------- | :------------ | :-------------- |
| Backups per project | N/A          | 500           | 1000            |

## Limitations

Backup limitations are as follows:

* Backups are stored in the same project, cloud provider, and region as the source index.
* You can only restore an index to the same project, cloud provider, and region as the source index.
* Backups only include vectors that were in the index at least 15 minutes prior to the backup time. This means that if a vector was inserted into an index and a backup was immediately taken after, the recently inserted vector may not be backed up. More specifically, if a backup is created only a few minutes after the source index was created, the backup may have 0 vectors.
* You can only perform operations on backups in the current Pinecone project.

## Backup and restore cost

* To understand how cost is calculated for backups and restores, see [Understanding cost](/guides/manage-cost/understanding-cost#backups-and-restores).
* For up-to-date pricing information, see [Pricing](https://www.pinecone.io/pricing/).

---

# Source: https://docs.pinecone.io/integrations/build-integration/become-a-partner.md

# Become a Pinecone partner

Anyone can use the [Pinecone SDKs](/reference/pinecone-sdks) or the [Pinecone API](/reference/api/introduction) to build a native Pinecone integration into a third-party service or tool. However, becoming an official Pinecone partner can help accelerate your go-to-market and add value to your customers.

To start the process, fill out our [application form](https://www.pinecone.io/partners/#sales-contact-form-submissions).

## Additional information

* [Attribute usage to your integration](/integrations/build-integration/attribute-usage-to-your-integration)
* [Connect your users to Pinecone](/integrations/build-integration/connect-your-users-to-pinecone)

---

# Source: https://docs.pinecone.io/troubleshooting/billing-disputes-and-refunds.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Billing disputes and refunds

As a rule, Pinecone does not offer refunds for unused indexes. If you use a pod-based index, we charge only for the pods you use to create it, not per API call or query. Whether you have used your index or not does not factor into our billing.

Our serverless indexes are a better fit if you don't plan to use your index on a regular basis. Serverless indexes are billed by the number of reads and writes you run and how much storage your index consumes. If your bill for your pod-based index is too high, we recommend [migrating to a serverless index](/guides/indexes/pods/migrate-a-pod-based-index-to-serverless) instead.

Our billing policies are detailed in our [user agreement](https://www.pinecone.io/user-agreement/) and [pricing page](https://pinecone.io/pricing).

We have several resources available to help you manage your bill:

* [Change your billing plan](/guides/organizations/manage-billing/upgrade-billing-plan)
* [Monitor your usage and costs](/guides/manage-cost/monitor-usage-and-costs)
* [Understand cost](/guides/manage-cost/understanding-cost)
* [Manage cost](/guides/manage-cost/manage-cost)

---

# Source: https://docs.pinecone.io/integrations/box.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Box

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Connect a [Box](https://www.box.com/) account to a Pinecone vector database.

This integration allows embeddings generation based on content stored in a particular account and folder in Box. By default, the Pinecone Inference API is used, as well as OpenAI for the LLM. The integration can be used within a larger AI agent or workflow.

<PrimarySecondaryCTA primaryHref={"https://medium.com/box-developer-blog/demo-box-pinecone-f03783c412bb"} primaryLabel={"Get started"} />

---

# Source: https://docs.pinecone.io/guides/production/bring-your-own-cloud.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Bring your own cloud

> Deploy Pinecone in your AWS or GCP account

Bring your own cloud (BYOC) lets you deploy Pinecone Database in your own AWS or GCP account to ensure data sovereignty and compliance, with Pinecone handling provisioning, operations, and maintenance.

<Note>
  BYOC is in [public preview](/release-notes/feature-availability) on AWS and GCP. To learn more about the offering, [contact Pinecone](https://www.pinecone.io/contact/?contact_form_inquiry_type=Product+Information).
</Note>

## Use cases

Pinecone BYOC is designed for organizations with high security and compliance requirements, for example:

* **Data sovereignty**: If your organization has strict data governance policies, Pinecone BYOC can help ensure that all data is stored and processed locally and does not leave your security perimeter.
* **Data residency**: The standard Pinecone managed service can be deployed in several [AWS or GCP cloud regions](/guides/index-data/create-an-index#cloud-regions). If your organization has specific data residency or latency constraints that require you to deploy in regions that Pinecone does not yet support, Pinecone BYOC gives you that flexibility.

## Architecture

<img className="block max-w-full dark:hidden" noZoom src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=01b34857159d1eb8849f4250aa79ed00" data-og-width="2040" width="2040" data-og-height="1240" height="1240" data-path="images/byoc.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=e925a3467600c20f1c37e68e49d38acb 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=c2c1b4077bf7f8dc593553a399e5fff6 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6688be4f64b0b4db4fe2ef90e04d6c44 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=4aa2923e28afc6a167bd8dfb28662c58 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=c5c4c539a148f31e977c39c2ca813733 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=8669a2c64a678822e7f2abe7dde6fdf0 2500w" />

<img className="hidden max-w-full dark:block" noZoom src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc-dark.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=a27381222b26d1f31bc7d52b09f3ace0" data-og-width="2040" width="2040" data-og-height="1240" height="1240" data-path="images/byoc-dark.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc-dark.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=f18f81d0f2a9387d2fd10231a4952284 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc-dark.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=7be91ffb937303f692dac9f4c7a812b6 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc-dark.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=06ec40f990b9795d42cd670c1f80e603 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc-dark.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6cc520e19003068137c960d9064648fc 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc-dark.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=902b44ca748163f9b199f128c5c7f543 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/byoc-dark.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6d1fc020dbfb7fc4d144012711c10a2a 2500w" />

The BYOC architecture employs a split model:

* **Data plane**: The data plane is responsible for storing and processing your records, executing queries, and interacting with object storage for index data. In a BYOC deployment, the data plane is hosted in your own AWS or GCP account within a dedicated VPC, ensuring that all data is stored and processed locally and does not leave your organizational boundaries. You use a [private endpoint](#configure-a-private-endpoint) (AWS PrivateLink or GCP Private Service Connect) as an additional measure to secure requests to your indexes.
* **Control plane**: The control plane is responsible for managing the index lifecycle as well as region-agnostic services such as user management, authentication, and billing. The control plane does not hold or process any records. In a BYOC deployment, the control plane is managed by Pinecone and hosted globally. Communication between the data plane and control plane is encrypted using TLS and employs role-based access control (RBAC) with minimal IAM permissions.

## Onboarding

The onboarding process for BYOC in AWS or GCP involves the following general stages:

<Steps>
  <Step title="Set up AWS or GCP account">
    If you don't already have an AWS or GCP account where you want to deploy Pinecone, you create one for this purpose.
  </Step>

  <Step title="Execute Terraform template">
    You download and run a Terraform template provided by Pinecone. This template creates essential resources, including an IAM role with scoped-down permissions and a trust relationship with Pinecone's AWS or GCP account.
  </Step>

  <Step title="Create environment">
    Pinecone deploys a data plane cluster within a dedicated VPC in your AWS or GCP account, and you [configure a private endpoint](#configure-a-private-endpoint) for securely connecting to your indexes via AWS PrivateLink or GCP Private Service Connect.
  </Step>

  <Step title="Validate">
    Once the environment is operational, Pinecone performs validation tests to ensure proper functionality.
  </Step>
</Steps>

## Configure a private endpoint

You use a [private endpoint](#configure-a-private-endpoint) to securely connect to your BYOC indexes. On AWS, you use the [AWS PrivateLink](https://docs.aws.amazon.com/vpc/latest/privatelink/what-is-privatelink.html) service; on GCP, you use the [GCP Private Service Connect](https://cloud.google.com/vpc/docs/private-service-connect) service.

<Tabs>
  <Tab title="AWS">
    Follow the instructions in the AWS documentation to [create a VPC endpoint](https://docs.aws.amazon.com/vpc/latest/privatelink/create-interface-endpoint.html#create-interface-endpoint-aws) for connecting to your indexes via AWS PrivateLink.

    * For **Resource configurations**, select the relevant resource for your Pinecone BYOC deployment.

    * For **Network settings**, select the VPC for your BYOC deployment.

    * In **Additional settings**, select **Enable DNS name** to allow you to access your indexes using a DNS name.
  </Tab>

  <Tab title="GCP">
    <Steps>
      <Step title="Create a private endpoint">
        Follow the instructions in the GCP documentation to [create a private endpoint](https://cloud.google.com/vpc/docs/configure-private-service-connect-services#create-endpoint) for connecting to your indexes via GCP Private Service Connect.

        * Set the **Target service** to the following:

          ```
          projects/<YOUR-BYOC-PROJECT>/regions/<YOUR-BYOC-REGION>/serviceAttachments/pinecone-psc
          ```
        * Copy the IP address of the private endpoint. You'll need it later.
      </Step>

      <Step title="Create a private DNS zone">
        Follow the instructions in the GCP documentation to [create a private DNS zone](https://cloud.google.com/dns/docs/zones#create-private-zone).

        * Set the **DNS name** to the following:

          ```
          private.<YOUR-BYOC-ENVIRONMENT>.pinecone.io
          ```
        * Select the same VPC network as the private endpoint.
      </Step>

      <Step title="Add a resource record set">
        Follow the instructions in the GCP documentation to [add a resource record set](https://cloud.google.com/dns/docs/records#add-rrset).

        * Set the **DNS name** to **\***.

        * Set the **Resource record type** to **A**.

        * Set the **Ipv4 Address** to the IP address of the private endpoint.
      </Step>
    </Steps>
  </Tab>
</Tabs>

## Create an index

Once your BYOC environment is ready, you can create a BYOC index in the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/indexes) or via Pinecone API or [Python SDK](/reference/sdks/python/overview).

To create a BYOC index, set the `spec` parameter to the environment name provided to you during onboarding, for example:

<CodeGroup>
  ```python Python {9-11} theme={null}
  from pinecone import Pinecone, ByocSpec

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.create_index(
      name="example-byoc-index", 
      dimension=1536, 
      metric="cosine", 
      spec=ByocSpec(
          environment="aws-us-east-1-b921"
      ),
      deletion_protection="disabled",
      tags={
          "example": "tag"
      }
  )
  ```

  ```shell curl {11-15} theme={null}
  curl -s "https://api.pinecone.io/indexes" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "name": "example-byoc-index",
          "vector_type": "dense",
          "dimension": 1536,
          "metric": "cosine",
          "spec": {
              "byoc": {
                  "environment": "aws-us-east-1-b921"
              }
          },
          "tags": {
              "example": "tag"
          },
          "deletion_protection": "disabled"
        }'
  ```
</CodeGroup>

## Read and write data

<Note>
  BYOC does not support reading and writing data from the index browser in the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/indexes/-/browser).
</Note>

Once your private endpoint is configured, you can run data operations against an index as usual, but you must target the index using its private endpoint URL. The only difference in the URL is that `.svc.` is changed to `.svc.private.`.

You can get the private endpoint URL for an index from the Pinecone console or API.

<Tabs>
  <Tab title="Console">
    To get the private endpoint URL for an index from the Pinecone console:

    1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).
    2. Select the project containing the index.
    3. Select the index.
    4. Copy the URL under **PRIVATE ENDPOINT**.
  </Tab>

  <Tab title="API">
    To get the private endpoint URL for an index from the API, use the [`describe_index`](/reference/api/latest/control-plane/describe_index) operation, which returns the private endpoint URL as the `private_host` value:

    <CodeGroup>
      ```JavaScript JavaScript theme={null}
      import { Pinecone } from '@pinecone-database/pinecone';

      const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

      await pc.describeIndex('docs-example');
      ```

      ```go Go theme={null}
      package main

      import (
          "context"
          "encoding/json"
          "fmt"
          "log"

          "github.com/pinecone-io/go-pinecone/v4/pinecone"
      )

      func prettifyStruct(obj interface{}) string {
          bytes, _ := json.MarshalIndent(obj, "", "  ")
          return string(bytes)
      }

      func main() {
          ctx := context.Background()

          pc, err := pinecone.NewClient(pinecone.NewClientParams{
              ApiKey: "YOUR_API_KEY",
          })
          if err != nil {
              log.Fatalf("Failed to create Client: %v", err)
          }

          idx, err := pc.DescribeIndex(ctx, "docs-example")
          if err != nil {
              log.Fatalf("Failed to describe index \"%v\": %v", idx.Name, err)
          } else {
              fmt.Printf("index: %v\n", prettifyStruct(idx))
          }
      }
      ```

      ```bash curl theme={null}
      PINECONE_API_KEY="YOUR_API_KEY"

      curl -i -X GET "https://api.pinecone.io/indexes/docs-example" \
          -H "Api-Key: YOUR_API_KEY" \
          -H "X-Pinecone-Api-Version: 2025-10"
      ```
    </CodeGroup>

    The response includes the private endpoint URL as the `private_host` value:

    <CodeGroup>
      ```json JavaScript {6} theme={null}
      {
        name: 'docs-example',
        dimension: 1536,
        metric: 'cosine',
        host: 'docs-example-jl7boae.svc.aped-4627-b74a.pinecone.io',
        privateHost: 'docs-example-jl7boae.svc.private.aped-4627-b74a.pinecone.io',
        deletionProtection: 'disabled',
        tags: { environment: 'production' },
        embed: undefined,
        spec: {
          byoc: undefined,
          pod: undefined,
          serverless: { cloud: 'aws', region: 'us-east-1' }
        },
        status: { ready: true, state: 'Ready' },
        vectorType: 'dense'
      }
      ```

      ```go Go {5} theme={null}
      index: {
        "name": "docs-example",
        "dimension": 1536,
        "host": "docs-example-jl7boae.svc.aped-4627-b74a.pinecone.io",
        "private_host": "docs-example-jl7boae.svc.private.aped-4627-b74a.pinecone.io",
        "metric": "cosine",
        "deletion_protection": "disabled",
        "spec": {
          "serverless": {
            "cloud": "aws",
            "region": "us-east-1"
          }
        },
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "tags": {
          "environment": "production"
        }
      }
      ```

      ```json curl {12} theme={null}
      {
        "id": "025117b3-e683-423c-b2d1-6d30fbe5027f",
        "vector_type": "dense",
        "name": "docs-example",
        "metric": "cosine",
        "dimension": 1536,
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "host": "docs-example-jl7boae.svc.aped-4627-b74a.pinecone.io",
        "private_host": "docs-example-jl7boae.svc.private.aped-4627-b74a.pinecone.io",
        "spec": {
          "serverless": {
            "region": "us-east-1",
            "cloud": "aws"
          }
        },
        "deletion_protection": "disabled",
        "tags": {
          "environment": "production"
        }
      ```
    </CodeGroup>
  </Tab>
</Tabs>

<Note>
  If you run data operations against an index from outside the Private Endpoint, you will get an `Unauthorized` response.
</Note>

## Monitoring

Pinecone engineers monitor the state of your BYOC deployment and manage incidents if they arise. In addition, you can [monitor performance metrics](/guides/production/monitoring) for your BYOC indexes in the Pinecone Console or with Prometheus or Datadog.

<Note>
  To use Prometheus, your monitoring tool must have access to your VPC.
</Note>

## Limitations

BYOC does not support the following:

* [Integrated embedding](/guides/index-data/indexing-overview#integrated-embedding), which relies on models hosted by Pinecone that are outsite of your AWS or GCP account.

* Reading and writing data from the index browser in the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/indexes/-/browser). You must use the Pinecone API or SDKs instead.

* Using [customer-managed encryption keys (CMEK)](/guides/production/configure-cmek) to secure data within a Pinecone project.

Also, to [monitor performance metrics with Prometheus](/guides/production/monitoring#monitor-with-prometheus), you must configure Prometheus within your VPC.

## FAQs

<AccordionGroup>
  <Accordion title="What is the difference between BYOC and Pinecone's standard service?">
    In the standard service, Pinecone manages all cloud resources and includes their cost in the service fee. In BYOC, customers provision and pay for cloud resources directly through their AWS or GCP account, providing greater control and data sovereignty as well as access to available AWS or GCP credits or discounts.

    Also, BYOC does not support certain features. See [Limitations](#limitations) for details.
  </Accordion>

  <Accordion title="How is data secured in BYOC?">
    Data is stored and processed exclusively within the customer's AWS or GCP account, with encryption applied at rest and in transit. Communication between the data plane and control plane is encrypted using TLS, and access is controlled via RBAC and scoped IAM permissions. AWS PrivateLink or GCP Private Service Connect is used for secure data plane API calls.
  </Accordion>

  <Accordion title="Is BYOC available in other cloud providers?">
    Currently, BYOC is available in AWS and GCP. Support for Azure is planned for future releases.
  </Accordion>
</AccordionGroup>

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/cancel_import.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Cancel an import

> Cancel an import operation if it is not yet finished. It has no effect if the operation is already finished.

For guidance and examples, see [Import data](https://docs.pinecone.io/guides/index-data/import-data).

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Standard and Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

<RequestExample>
  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X DELETE "https://$INDEX_HOST/bulk/imports/101" \
    -H 'Api-Key: $PINECONE_API_KEY' \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {}
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml delete /bulk/imports/{id}
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /bulk/imports/{id}:
    delete:
      tags:
        - Bulk Operations
      summary: Cancel an import
      description: >-
        Cancel an import operation if it is not yet finished. It has no effect
        if the operation is already finished.


        For guidance and examples, see [Import
        data](https://docs.pinecone.io/guides/index-data/import-data).
      operationId: cancelBulkImport
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: id
          description: Unique identifier for the import operation.
          required: true
          schema:
            type: string
            minLength: 1
            maxLength: 1000
          example: '101'
          style: simple
      responses:
        '200':
          description: Operation cancelled successfully
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/CancelImportResponse'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    CancelImportResponse:
      description: The response for the `cancel_import` operation.
      type: object
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/guides/organizations/manage-billing/change-payment-method.md

# Source: https://docs.pinecone.io/guides/assistant/admin/change-payment-method.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Change your payment method

> Update billing payment method for your organization.

You can pay for the [Standard and Enterprise plans](https://www.pinecone.io/pricing/) with a credit/debit card or through the AWS Marketplace, Microsoft Marketplace, or Google Cloud Marketplace. This page describes how to switch between these payment methods.

<Note>
  To change your payment method, you must be an [organization owner or billing admin](/guides/organizations/understanding-organizations#organization-roles).
</Note>

## Credit card → marketplace

To change from credit card to marketplace billing, you'll need to:

1. Create a new Pinecone organization through the marketplace
2. Migrate your existing projects to the new Pinecone organization
3. Add your team members to the new Pinecone organization
4. Downgrade your original Pinecone organization once migration is complete

<Tabs>
  <Tab title="Credit card → Google Cloud">
    To change from paying with a credit card to paying through the Google Cloud Marketplace, do the following:

    1. Subscribe to Pinecone in the Google Cloud Marketplace:

       1. In the Google Cloud Marketplace, go to the [Pinecone listing](https://console.cloud.google.com/marketplace/product/pinecone-public/pinecone).
       2. Click **Subscribe**.
       3. On the **Order Summary** page, select a billing account, accept the terms and conditions, and click **Subscribe**.

          <Note>The billing unit listed does not reflect the actual cost or metering of costs for Pinecone. See the [Pinecone Pricing page](https://www.pinecone.io/pricing/) for accurate details.</Note>
       4. On the **Your order request has been sent to Pinecone** modal, click **Sign up with Pinecone**. This takes you to a Google-specific Pinecone sign-up page.
       5. Sign up using the same authentication method as your existing Pinecone organization.

    2. Create a new Pinecone organization and connect it to your Google Cloud Marketplace account:

       1. On the **Connect GCP to Pinecone** page, choose **Select an organization > + Create New Organization**.

          <Note>If you see a message saying that the subscription is still in process, wait a few minutes, refresh the page, and proceed only when the message has disappeared.</Note>
       2. Enter the name of the new organization and click **Connect to Pinecone**.
       3. On the **Confirm GCP marketplace Connection** modal, click **Connect**. This takes you to your new organization in the Pinecone console.

    3. Migrate your projects to the new Pinecone organization:

       1. In the Pinecone console, go to [**Settings > Access > Members**](https://app.pinecone.io/organizations/-/settings/access/members).
       2. Make sure the **Owner** email address for your original organization is set as an **Owner** or **Billing Admin** for your new organization. This allows Pinecone to verify that both the original and new organizations are owned by the same person.
       3. Go to [**Settings > Manage**](https://app.pinecone.io/organizations/-/settings/manage) and copy your new organization ID.
       4. Go to [**Settings > Support > Tickets**](https://app.pinecone.io/organizations/-/settings/support/ticket/create).
       5. For **Ticket category**, select **Project or Organization Management**.
       6. For **Subject**, enter "Migrate projects to a new organization".
       7. For **Description**, enter the following:

          ```
          I am changing my payment method from credit card to Google Cloud Marketplace. 
          Please migrate my projects to my new organization: `<NEW_ORG_ID>`
          ```
       8. Click **Submit**.

    4. Add your team members to the new organization:

       1. In the Pinecone console, go to [**Settings > Access > Members**](https://app.pinecone.io/organizations/-/settings/access/members).
       2. [Add your team members to the new organization](/guides/organizations/manage-organization-members#add-a-member-to-an-organization).

    5. Downgrade your original Pinecone organization:

       <Note>
         Do not downgrade your original organization until you receive a confirmation that Pinecone has finished the migration to your new organization.
       </Note>

       1. In the Pinecone console, go to your original organization.
       2. Go to **Settings > Billing > Plans**.
       3. In the **Starter** section, click **Downgrade**.
       4. Click **Confirm downgrade**.

    Going forward, your usage of Pinecone will be billed through the Google Cloud Marketplace.

    <Warning>
      You can [delete your original organization](/troubleshooting/delete-your-organization). However, before deleting, make sure to [download your past invoices](/guides/organizations/manage-billing/access-your-invoices) since you will lose access to them once the organization is deleted.
    </Warning>
  </Tab>

  <Tab title="Credit card → AWS">
    To change from paying with a credit card to paying through the AWS Marketplace, do the following:

    1. Subscribe to Pinecone in the AWS Marketplace:

       1. In the AWS Marketplace, go to the [Pinecone listing](https://aws.amazon.com/marketplace/pp/prodview-xhgyscinlz4jk).
       2. Click **View purchase options**.
       3. On the **Subscribe to Pinecone Vector Database** page, review the offer and then click **Subscribe**.

          <Note>The billing unit listed does not reflect the actual cost or metering of costs for Pinecone. See the [Pinecone Pricing page](https://www.pinecone.io/pricing/) for accurate details.</Note>
       4. You'll see a message stating that your subscription is in process. Click **Set up your account**. This takes you to an AWS-specific Pinecone sign-up page.
       5. Sign up using the same authentication method as your existing Pinecone organization.

    2. Create a new Pinecone organization and connect it to your AWS account:

       1. On the **Connect AWS to Pinecone** page, choose **Select an organization > + Create New Organization**.

       <Note>If you see a message saying that the subscription is still in process, wait a few minutes, refresh the page, and proceed only when the message has disappeared.</Note>

       1. Enter the name of the new organization and click **Connect to Pinecone**.
       2. On the **Confirm AWS Marketplace Connection** modal, click **Connect**. This takes you to your new organization in the Pinecone console.

    3. Migrate your projects to the new Pinecone organization:

       1. In the Pinecone console, go to [**Settings > Access > Members**](https://app.pinecone.io/organizations/-/settings/access/members).
       2. Make sure the **Owner** email address for your original organization is set as an **Owner** or **Billing Admin** for your new organization. This allows Pinecone to verify that both the original and new organizations are owned by the same person.
       3. Go to [**Settings > Manage**](https://app.pinecone.io/organizations/-/settings/manage) and copy your new organization ID.
       4. Go to [**Settings > Support > Tickets**](https://app.pinecone.io/organizations/-/settings/support/ticket/create).
       5. For **Ticket category**, select **Project or Organization Management**.
       6. For **Subject**, enter "Migrate projects to a new organization".
       7. For **Description**, enter the following:

          ```
          I am changing my payment method from credit card to Google Cloud Marketplace. 
          Please migrate my projects to my new organization: `<NEW_ORG_ID>`
          ```
       8. Click **Submit**.

    4. Add your team members to the new organization:

       1. In the Pinecone console, go to [**Settings > Access > Members**](https://app.pinecone.io/organizations/-/settings/access/members).
       2. [Add your team members to the new organization](/guides/organizations/manage-organization-members#add-a-member-to-an-organization).

    5. Downgrade your original Pinecone organization:

       <Note>
         Do not downgrade your original organization until you receive a confirmation that Pinecone has finished the migration to your new organization.
       </Note>

       1. In the Pinecone console, go to your original organization.
       2. Go to **Settings > Billing > Plans**.
       3. In the **Starter** section, click **Downgrade**.
       4. Click **Confirm downgrade**.

    Going forward, your usage of Pinecone will be billed through the AWS Marketplace.

    <Warning>
      You can [delete your original organization](/troubleshooting/delete-your-organization). However, before deleting, make sure to [download your past invoices](/guides/organizations/manage-billing/access-your-invoices) since you will lose access to them once the organization is deleted.
    </Warning>
  </Tab>

  <Tab title="Credit card → Microsoft">
    To change from paying with a credit card to paying through the Microsoft Marketplace, do the following:

    1. Subscribe to Pinecone in the Microsoft Marketplace:

       1. In the Microsoft Marketplace, go to the [Pinecone listing](https://marketplace.microsoft.com/product/saas/pineconesystemsinc1688761585469.pineconesaas).
       2. Click **Get it now**.
       3. Select the **Pinecone - Pay As You Go** plan.
       4. Click **Subscribe**.
       5. On the **Subscribe to Pinecone** page, select the required details and click **Review + subscribe**.

          <Note>The billing unit listed does not reflect the actual cost or metering of costs for Pinecone. See the [Pinecone Pricing page](https://www.pinecone.io/pricing/) for accurate details.</Note>
       6. Click **Subscribe**.
       7. After the subscription is approved, click **Configure account now**. This redirects you to an Microsoft-specific Pinecone login page.
       8. Sign up using the same authentication method as your existing Pinecone organization.

    2. Create a new Pinecone organization and connect it to your Microsoft Marketplace account:

       1. On the **Connect Azure to Pinecone** page, choose **Select an organization > + Create New Organization**.

       <Note>If you see a message saying that the subscription is still in process, wait a few minutes, refresh the page, and proceed only when the message has disappeared.</Note>

       1. Enter the name of the new organization and click **Connect to Pinecone**.
       2. On the **Connect Azure marketplace connection** modal, click **Connect**. This takes you to your new organization in the Pinecone console.

    3. Migrate your projects to the new Pinecone organization:

       1. In the Pinecone console, go to [**Settings > Access > Members**](https://app.pinecone.io/organizations/-/settings/access/members).
       2. Make sure the **Owner** email address for your original organization is set as an **Owner** or **Billing Admin** for your new organization. This allows Pinecone to verify that both the original and new organizations are owned by the same person.
       3. Go to [**Settings > Manage**](https://app.pinecone.io/organizations/-/settings/manage) and copy your new organization ID.
       4. Go to [**Settings > Support > Tickets**](https://app.pinecone.io/organizations/-/settings/support/ticket/create).
       5. For **Ticket category**, select **Project or Organization Management**.
       6. For **Subject**, enter "Migrate projects to a new organization".
       7. For **Description**, enter the following:

          ```
          I am changing my payment method from credit card to Microsoft Marketplace. 
          Please migrate my projects to my new organization: `<NEW_ORG_ID>`
          ```
       8. Click **Submit**.

    4. Add your team members to the new organization:

       1. In the Pinecone console, go to [**Settings > Access > Members**](https://app.pinecone.io/organizations/-/settings/access/members).
       2. [Add your team members to the new organization](/guides/organizations/manage-organization-members#add-a-member-to-an-organization).

    5. Downgrade your original Pinecone organization:

       <Note>
         Do not downgrade your original organization until you receive a confirmation that Pinecone has finished the migration to your new organization.
       </Note>

       1. In the Pinecone console, go to your original organization.
       2. Go to **Settings > Billing > Plans**.
       3. In the **Starter** section, click **Downgrade**.
       4. Click **Confirm downgrade**.

    Going forward, your usage of Pinecone will be billed through the Microsoft Marketplace.

    <Warning>
      You can [delete your original organization](/troubleshooting/delete-your-organization). However, before deleting, make sure to [download your past invoices](/guides/organizations/manage-billing/access-your-invoices) since you will lose access to them once the organization is deleted.
    </Warning>
  </Tab>
</Tabs>

## Marketplace → credit card

To change from marketplace billing to credit card, you'll need to:

1. Create a new organization in your Pinecone account
2. Upgrade the new organization to the Standard or Enterprise plan
3. Migrate your existing projects to the new organization
4. Add your team members to the new organization
5. Downgrade your original organization once migration is complete

<Tabs>
  <Tab title="Google Cloud → credit card">
    To change from paying through the Google Cloud Marketplace to paying with a credit card, do the following:

    1. Create a new organization in your Pinecone account:

       1. In the Pinecone console, go to [**Organizations**](https://app.pinecone.io/organizations/-/settings/account/organizations).
       2. Click **+ Create organization**.
       3. Enter the name of the new organization and click **Create**.

    2. Upgrade the new organization:

       1. Go to [Settings > Billing > Plans](https://app.pinecone.io/organizations/-/settings/billing/plans).
       2. Click **Upgrade** in the **Standard** or **Enterprise** plan section.
       3. Click **Credit / Debit card**.
       4. Enter your credit card information.
       5. Click **Upgrade**.

       The new organization is now set up with credit card billing. You'll use this organization after completing the rest of this process.

    3. Migrate your projects to the new Pinecone organization:

       1. Go to [**Settings > Manage**](https://app.pinecone.io/organizations/-/settings/manage) and copy your new organization ID.
       2. Go to [**Settings > Support > Tickets**](https://app.pinecone.io/organizations/-/settings/support/ticket/create).
       3. For **Ticket category**, select **Project or Organization Management**.
       4. For **Subject**, enter "Migrate projects to a new organization".
       5. For **Description**, enter the following:

          ```
          I am changing my payment method from Google Cloud Marketplace to credit card. 
          Please migrate my projects to my new organization: `<NEW_ORG_ID>`
          ```
       6. Click **Submit**.

    4. Add your team members to the new organization:

       1. In the Pinecone console, go to [**Settings > Access > Members**](https://app.pinecone.io/organizations/-/settings/access/members).
       2. [Add your team members to the new organization](/guides/organizations/manage-organization-members#add-a-member-to-an-organization).

    5. Downgrade your original Pinecone organization:

       <Note>
         Do not downgrade your original organization until you receive a confirmation that Pinecone has finished the migration to your new organization.
       </Note>

       1. In the Pinecone console, go to your original organization.
       2. Go to [**Settings > Billing > Plans**](https://app.pinecone.io/organizations/-/settings/billing/plans).
       3. In the **Starter** section, click **Downgrade**.
       4. Click **Confirm downgrade**.
       5. On the **Continue your downgrade on the GCP marketplace** modal, click **Continue to marketplace**. This takes you to your orders page in Google Cloud Marketplace.
       6. [Cancel the order](https://cloud.google.com/marketplace/docs/manage-billing#saas-products) for your original organization.

          <Tip>
            If you don't see the order, check that the correct billing account is selected.
          </Tip>

    Going forward, you'll use your new organization and your usage will be billed through the credit card you provided.

    <Warning>
      You can [delete your original organization](/troubleshooting/delete-your-organization). However, before deleting, make sure to [download your past invoices](/guides/organizations/manage-billing/access-your-invoices) since you will lose access to them once the organization is deleted.
    </Warning>
  </Tab>

  <Tab title="AWS → credit card">
    To change from paying through the AWS Marketplace to paying with a credit card, do the following:

    1. Create a new organization in your Pinecone account:

       1. In the Pinecone console, go to [**Organizations**](https://app.pinecone.io/organizations/-/settings/account/organizations).
       2. Click **+ Create organization**.
       3. Enter the name of the new organization and click **Create**.

    2. Upgrade the new organization:

       1. Go to [Settings > Billing > Plans](https://app.pinecone.io/organizations/-/settings/billing/plans).
       2. Click **Upgrade** in the **Standard** or **Enterprise** plan section.
       3. Click **Credit / Debit card**.
       4. Enter your credit card information.
       5. Click **Upgrade**.

       The new organization is now set up with credit card billing. You'll use this organization after completing the rest of this process.

    3. Migrate your projects to the new Pinecone organization:

       1. Go to [**Settings > Manage**](https://app.pinecone.io/organizations/-/settings/manage) and copy your new organization ID.
       2. Go to [**Settings > Support > Tickets**](https://app.pinecone.io/organizations/-/settings/support/ticket/create).
       3. For **Ticket category**, select **Project or Organization Management**.
       4. For **Subject**, enter "Migrate projects to a new organization".
       5. For **Description**, enter the following:

          ```
          I am changing my payment method from AWS Marketplace to credit card. 
          Please migrate my projects to my new organization: `<NEW_ORG_ID>`
          ```
       6. Click **Submit**.

    4. Add your team members to the new organization:

       1. In the Pinecone console, go to [**Settings > Access > Members**](https://app.pinecone.io/organizations/-/settings/access/members).
       2. [Add your team members to the new organization](/guides/organizations/manage-organization-members#add-a-member-to-an-organization).

    5. Downgrade your original Pinecone organization:

       <Note>
         Do not downgrade your original organization until you receive a confirmation that Pinecone has finished the migration to your new organization.
       </Note>

       1. In the Pinecone console, go to your original organization.
       2. Go to [**Settings > Billing > Plans**](https://app.pinecone.io/organizations/-/settings/billing/plans).
       3. In the **Starter** section, click **Downgrade**.
       4. Click **Confirm downgrade**.
       5. On the **Continue your downgrade on the AWS marketplace** modal, click **Continue to marketplace**. This takes you to the [Manage subscriptions](https://console.aws.amazon.com/marketplace) page in the AWS Marketplace.
       6. [Cancel the subscription](https://docs.aws.amazon.com/marketplace/latest/buyerguide/cancel-subscription.html#cancel-saas-subscription) to Pinecone.

    Going forward, you'll use your new organization and your usage will be billed through the credit card you provided.

    <Warning>
      You can [delete your original organization](/troubleshooting/delete-your-organization). However, before deleting, make sure to [download your past invoices](/guides/organizations/manage-billing/access-your-invoices) since you will lose access to them once the organization is deleted.
    </Warning>
  </Tab>

  <Tab title="Microsoft → credit card">
    To change from paying through the Microsoft Marketplace to paying with a credit card, do the following:

    1. Create a new organization in your Pinecone account:

       1. In the Pinecone console, go to [**Organizations**](https://app.pinecone.io/organizations/-/settings/account/organizations).
       2. Click **+ Create organization**.
       3. Enter the name of the new organization and click **Create**.

    2. Upgrade the new organization:

       1. Go to [Settings > Billing > Plans](https://app.pinecone.io/organizations/-/settings/billing/plans).
       2. Click **Upgrade** in the **Standard** or **Enterprise** plan section.
       3. Click **Credit / Debit card**.
       4. Enter your credit card information.
       5. Click **Upgrade**.

       The new organization is now set up with credit card billing. You'll use this organization after completing the rest of this process.

    3. Migrate your projects to the new Pinecone organization:

       1. Go to [**Settings > Manage**](https://app.pinecone.io/organizations/-/settings/manage) and copy your new organization ID.
       2. Go to [**Settings > Support > Tickets**](https://app.pinecone.io/organizations/-/settings/support/ticket/create).
       3. For **Ticket category**, select **Project or Organization Management**.
       4. For **Subject**, enter "Migrate projects to a new organization".
       5. For **Description**, enter the following:

          ```
          I am changing my payment method from Microsoft Marketplace to credit card. 
          Please migrate my projects to my new organization: `<NEW_ORG_ID>`
          ```
       6. Click **Submit**.

    4. Add your team members to the new organization:

       1. In the Pinecone console, go to [**Settings > Access > Members**](https://app.pinecone.io/organizations/-/settings/access/members).
       2. [Add your team members to the new organization](/guides/organizations/manage-organization-members#add-a-member-to-an-organization).

    5. Downgrade your original Pinecone organization:

       <Note>
         Do not downgrade your original organization until you receive a confirmation that Pinecone has finished the migration to your new organization.
       </Note>

       1. In the Pinecone console, go to your original organization.
       2. Go to [**Settings > Billing > Plans**](https://app.pinecone.io/organizations/-/settings/billing/plans).
       3. In the **Starter** section, click **Downgrade**.
       4. Click **Confirm downgrade**.
       5. On the **Continue your downgrade on Azure marketplace** modal, click **Continue to marketplace**.
       6. On the **SaaS** page, click your subscription to Pinecone.
       7. Click **Cancel subscription**.
       8. Confirm the cancellation.

    Going forward, you'll use your new organization and your usage will be billed through the credit card you provided.

    <Warning>
      You can [delete your original organization](/troubleshooting/delete-your-organization). However, before deleting, make sure to [download your past invoices](/guides/organizations/manage-billing/access-your-invoices) since you will lose access to them once the organization is deleted.
    </Warning>
  </Tab>
</Tabs>

## Marketplace → marketplace

To change from one marketplace to another, you'll need to:

1. Subscribe to Pinecone in the new marketplace
2. Connect your existing org to the new marketplace
3. Cancel your subscription in the old marketplace

<Tabs>
  <Tab title="Any marketplace → Google Cloud">
    To change to a Google Cloud Marketplace billing account, do the following:

    1. Subscribe to Pinecone in the Google Cloud Marketplace:

       1. In the Google Cloud Marketplace, go to the [Pinecone listing](https://console.cloud.google.com/marketplace/product/pinecone-public/pinecone).
       2. Click **Subscribe**.
       3. On the **Order Summary** page, select a billing account, accept the terms and conditions, and click **Subscribe**.

          <Note>The billing unit listed does not reflect the actual cost or metering of costs for Pinecone. See the [Pinecone Pricing page](https://www.pinecone.io/pricing/) for accurate details.</Note>
       4. On the **Your order request has been sent to Pinecone** modal, click **Sign up with Pinecone**. This takes you to a Google-specific Pinecone login page.
       5. Log in to your Pinecone account. Use the same authentication method as your existing Pinecone organization.

    2. Connect your existing org to your Google account:

       1. On the **Connect GCP to Pinecone** page, select the Pinecone organization that you want to use Google Cloud Marketplace.

          <Note>If you see a message saying that the subscription is still in process, wait a few minutes, refresh the page, and proceed only when the message has disappeared.</Note>
       2. On the **Confirm GCP marketplace connection** modal, click **Connect**. This takes you to your organization in the Pinecone console.

          Going forward, your usage of Pinecone will be billed through the Google Cloud Marketplace.

    3. Cancel your subscription in your previous marketplace:

       * For AWS:
         1. In the AWS Marketplace, go to the [Manage subscriptions](https://console.aws.amazon.com/marketplace) page in the AWS Marketplace.
         2. [Cancel the subscription](https://docs.aws.amazon.com/marketplace/latest/buyerguide/cancel-subscription.html#cancel-saas-subscription) to Pinecone.

       * For Microsoft:
         1. Go to [Azure SaaS Resource Management](https://portal.azure.com/#blade/HubsExtension/BrowseResourceBlade/resourceType/Microsoft.SaaS%2Fresources).
         2. Select your subscription to Pinecone.
         3. Click **Cancel subscription**.
         4. Confirm the cancellation.
  </Tab>

  <Tab title="Any marketplace → AWS">
    To change to an AWS Marketplace billing account, do the following:

    1. Subscribe to Pinecone in the AWS Marketplace:

       1. In the AWS Marketplace, go to the [Pinecone listing](https://aws.amazon.com/marketplace/pp/prodview-xhgyscinlz4jk) in the AWS Marketplace.
       2. Click **View purchase options**.
       3. On the **Subscribe to Pinecone Vector Database** page, review the offer and then click **Subscribe**.

          <Note>The billing unit listed does not reflect the actual cost or metering of costs for Pinecone. See the [Pinecone Pricing page](https://www.pinecone.io/pricing/) for accurate details.</Note>
       4. You'll see a message stating that your subscription is in process. Click **Set up your account**. This takes you to an AWS-specific Pinecone login page.
       5. Log in to your Pinecone account. Use the same authentication method as your existing Pinecone organization.

    2. Connect your existing org to your AWS account:

       1. On the **Connect AWS to Pinecone** page, select the Pinecone organization that you want to change to AWS Marketplace.

          <Note>If you see a message saying that the subscription is still in process, wait a few minutes, refresh the page, and proceed only when the message has disappeared.</Note>
       2. On the **Confirm AWS marketplace connection** modal, click **Connect**. This takes you to your organization in the Pinecone console.

          Going forward, your usage of Pinecone will be billed through the AWS Marketplace.

    3. Cancel your subscription in your previous marketplace:

       * For Google Cloud Marketplace:
         1. Go to the [Orders](https://console.cloud.google.com/marketplace/orders) page.
         2. [Cancel the order](https://cloud.google.com/marketplace/docs/manage-billing#saas-products) for Pinecone.

       * For Microsoft Marketplace:
         1. Go to [Azure SaaS Resource Management](https://portal.azure.com/#blade/HubsExtension/BrowseResourceBlade/resourceType/Microsoft.SaaS%2Fresources).
         2. Select your subscription to Pinecone.
         3. Click **Cancel subscription**.
         4. Confirm the cancellation.
  </Tab>

  <Tab title="Any marketplace → Microsoft">
    To change to a Microsoft Marketplace billing account, do the following:

    1. Subscribe to Pinecone in the Microsoft Marketplace:

       1. In the Microsoft Marketplace, go to the [Pinecone listing](https://marketplace.microsoft.com/product/saas/pineconesystemsinc1688761585469.pineconesaas).
       2. Click **Get it now**.
       3. Select the **Pinecone - Pay As You Go** plan.
       4. Click **Subscribe**.
       5. On the **Subscribe to Pinecone** page, select the required details and click **Review + subscribe**.

          <Note>The billing unit listed does not reflect the actual cost or metering of costs for Pinecone. See the [Pinecone Pricing page](https://www.pinecone.io/pricing/) for accurate details.</Note>
       6. Click **Subscribe**.
       7. After the subscription is approved, click **Configure account now**. This redirects you to an Microsoft-specific Pinecone login page.
       8. Log in to your Pinecone account. Use the same authentication method as your existing Pinecone organization.

    2. Connect your existing org to your Microsoft account:

       1. On the **Connect Azure to Pinecone** page, select the Pinecone organization that you want to change to Microsoft Marketplace.

          <Note>If you see a message saying that the subscription is still in process, wait a few minutes, refresh the page, and proceed only when the message has disappeared.</Note>
       2. On the **Confirm Azure marketplace connection** modal, click **Connect**. This takes you to your organization in the Pinecone console.

          Going forward, your usage of Pinecone will be billed through the Microsoft Marketplace.

    3. Cancel your subscription in your previous marketplace:

       * For Google Cloud Marketplace:
         1. Go to the [Orders](https://console.cloud.google.com/marketplace/orders) page.
         2. [Cancel the order](https://cloud.google.com/marketplace/docs/manage-billing#saas-products) for Pinecone.

       * For AWS Marketplace:
         1. Go to the [Manage subscriptions](https://console.aws.amazon.com/marketplace) page in the AWS Marketplace.
         2. [Cancel the subscription](https://docs.aws.amazon.com/marketplace/latest/buyerguide/cancel-subscription.html#cancel-saas-subscription) to Pinecone.
  </Tab>
</Tabs>

## Credit card → credit card

To update your credit card information in the Pinecone console, do the following:

1. Go to [**Settings > Billing > Overview**](https://app.pinecone.io/organizations/-/settings/billing).
2. In the **Billing Contact** section, click **Edit**.
3. Enter your new credit card information.
4. Click **Update**.

---

# Source: https://docs.pinecone.io/guides/assistant/chat-through-the-openai-compatible-interface.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Chat through the OpenAI-compatible interface

> Integrate OpenAI-compatible chat interface with Pinecone Assistant.

After [uploading files](/guides/assistant/manage-files) to an assistant, you can chat with the assistant.

This page shows you how to chat with an assistant using the [OpenAI-compatible chat interface](/reference/api/latest/assistant/chat_completion_assistant). This interface is based on the OpenAI Chat Completion API, a commonly used and adopted API. It is useful if you need inline citations or OpenAI-compatible responses, but has limited functionality compared to the [standard chat interface](/guides/assistant/chat-with-assistant).

<Tip>
  The [standard chat interface](/guides/assistant/chat-with-assistant) is the recommended way to chat with an assistant, as it offers more functionality and control over the assistant's responses and references.
</Tip>

## Chat with an assistant

The [OpenAI-compatible chat interface](/reference/api/latest/assistant/chat_completion_assistant) can return responses in two different formats:

* [Default response](#default-response): The assistant returns a response in a single string field, which includes citation information.
* [Streaming response](#streaming-response): The assistant returns the response as a text stream.

### Default response

The following example sends a message and requests a response in the default format:

<Note>
  The `content` parameter in the request cannot be empty.
</Note>

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone(api_key="YOUR_API_KEY")

  # Get your assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant", 
  )

  # Chat with the assistant.
  chat_context = [Message(role="user", content='What is the maximum height of a red pine?')]
  response = assistant.chat_completions(messages=chat_context)

  print(response)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
  const assistantName = 'example-assistant';

  const assistant = pc.Assistant(assistantName);
  const chatResp = await assistant.chatCompletion({
        messages: [{ role: 'user', content: 'Who is the CFO of Netflix?' }]
      });
  console.log(chatResp);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME/chat/completions" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "What is the maximum height of a red pine?"
      }
    ]
  }'
  ```
</CodeGroup>

The example above returns a result like the following:

```JSON  theme={null}
{"chat_completion":
  {
    "id":"chatcmpl-9OtJCcR0SJQdgbCDc9JfRZy8g7VJR",
    "choices":[
      {
        "finish_reason":"stop",
        "index":0,
        "message":{
          "role":"assistant",
          "content":"The maximum height of a red pine (Pinus resinosa) is up to 25 meters."
        }
      }
    ],
    "model":"my_assistant"
  }
}
```

### Streaming response

Streaming responses can improve perceived latency by allowing users to see content as it's generated, rather than waiting for the complete response. This creates a more responsive chat experience, especially for longer responses.

The following example sends a message and requests a streaming response:

<Note>
  The `content` parameter in the request cannot be empty.
</Note>

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone(api_key="YOUR_API_KEY")

  # Get your assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant" 
  )

  # Streaming chat with the Assistant.
  chat_context = [Message(role="user", content="What is the maximum height of a red pine?")]
  response = assistant.chat_completions(messages=[chat_context], stream=True)

  for data in response:
      if data:
          print(data)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const chatResp = await assistant.chatCompletionStream({
      messages: [{ role: 'user', content: 'Who is the CFO of Netflix?' }]
  });

  for await (const response of chatResp) {
      if (response) {
          console.log(response);
      }
  }
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME/chat/completions" \
    -H "Api-Key: $PINECONE_API_KEY "\
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "What is the maximum height of a red pine?"
      }
    ],
    "stream": true
  }'
  ```
</CodeGroup>

The example above returns a result like the following:

```json  theme={null}
{
  'id': '000000000000000009de65aa87adbcf0', 
  'choices': [
      {
      'index': 0, 
      'delta': 
        {
        'role': 'assistant', 
        'content': 'The'
        }, 
      'finish_reason': None
      }
    ], 
  'model': 'gpt-4o-2024-05-13'
}

...

{
  'id': '00000000000000007a927260910f5839',
  'choices': [
      {
      'index': 0,
      'delta':
        {
          'role': '', 
          'content': 'The'
        }, 
      'finish_reason': None
      }
    ], 
  'model': 'gpt-4o-2024-05-13'
}

...

{
  'id': '00000000000000007a927260910f5839', 
  'choices': [
    {
      'index': 0, 
      'delta': 
        {
        'role': None, 
        'content': None
        }, 
      'finish_reason': 'stop'
      }
    ], 
  'model': 'gpt-4o-2024-05-13'
}
```

There are three types of messages in a chat completion response:

* **Message start**: Includes `"role":"assistant"`, which indicates that the assistant is responding to the user's message.
* **Content**: Includes a value in the `content` field (e.g., `"content":"The"`), which is part of the assistant's streamed response to the user's message.
* **Message end**: Includes `"finish_reason":"stop"`, which indicates that the assistant has finished responding to the user's message.

## Extract the response content

In the assistant's response, the message string is contained in the following JSON object:

* `choices.[0].message.content` for the default chat response
* `choices[0].delta.content` for the streaming chat response

You can extract the message content and print it to the console:

<Tabs>
  <Tab title="Default response">
    <CodeGroup>
      ```python Python theme={null}
      print(str(response.choices[0].message.content))
      ```

      ```bash curl theme={null}
      | jq '.choices.[0].message.content'
      ```
    </CodeGroup>

    This creates output like the following:

    ```bash  theme={null}
    A red pine, scientifically known as *Pinus resinosa*, is a medium-sized tree that can grow up to 25 meters high and 75 centimeters in diameter. [1, pp. 1]
    ```
  </Tab>

  <Tab title="Streaming response">
    <CodeGroup>
      ```python Python theme={null}
      for data in response:
          if data:
              print(str(data.choices[0].delta.content))
      ```

      ```bash curl theme={null}
      |  sed -u 's/.*"content":"\([^"]*\)".*/\1/'
      ```
    </CodeGroup>

    This creates output like the following:

    ```bash Streaming response theme={null}
    The
     maximum
     height
     of
     a
     red
     pine
     (
    Pin
    us
     resin
    osa
    )
     is
     up
     to
     twenty
    -five
     meters

     [1, pp. 1]
    .
    ```
  </Tab>
</Tabs>

## Choose a model

Pinecone Assistant supports the following models:

* `gpt-4o` (default)
* `gpt-4.1`
* `gpt-5`
* `o4-mini`
* `claude-sonnet-4-5`
* `gemini-2.5-pro`

<Note>
  Anthropic has [deprecated](https://platform.claude.com/docs/en/about-claude/model-deprecations) the Claude 3.5 Sonnet and Claude 3.7 Sonnet models. Assistant automatically routes chat requests that specify `claude-3-5-sonnet` or `claude-3-7-sonnet` to `claude-sonnet-4-5` at the same price.
</Note>

<Tip>
  For chat applications, we recommend using GPT models (`gpt-4o`, `gpt-4.1`, `gpt-5`, or `o4-mini`) as they typically provide faster response times compared to other models.
</Tip>

To choose a non-default model for your assistant, set the `model` parameter in the request:

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone(api_key="YOUR_API_KEY")

  # Get your assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant", 
  )

  # Chat with the assistant.
  chat_context = [Message(role="user", content="What is the maximum height of a red pine?")]
  response = assistant.chat_completions(
      messages=chat_context, 
      model="gpt-4.1"
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const chatResp = await assistant.chatCompletion({
    messages: [{ role: 'user', content: 'What is the maximum height of a red pine?' }],
    model: 'gpt-4.1',
  });

  console.log(chatResp);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME/chat/completions" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "What is the maximum height of a red pine?"
      }
    ],
    "model": "gpt-4.1"
  }'
  ```
</CodeGroup>

## Filter chat with metadata

You can [filter which documents to use for chat completions](/guides/assistant/files-overview#file-metadata). The following example filters the responses to use only documents that include the metadata `"resource": "encyclopedia"`.

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone(api_key="YOUR_API_KEY")

  # Get your assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant", 
  )

  # Chat with the assistant.
  chat_context = [Message(role="user", content="What is the maximum height of a red pine?")]
  response = assistant.chat_completions(messages=chat_context, stream=True, filter={"resource": "encyclopedia"})
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const chatResp = await assistant.chatCompletion({
    messages: [{ role: 'user', content: 'What is the maximum height of a red pine?' }],
    filter: {
      'resource': 'encyclopedia'
    }
  });
  console.log(chatResp);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME/chat/completions" \
    -H "Api-Key: $PINECONE_API_KEY "\
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "What is the maximum height of a red pine?"
      }
    ],
    "stream": true,
    "filter": 
      {
      "resource": "encyclopedia"
      }
    }'
  ```
</CodeGroup>

## Set the sampling temperature

<Note>
  This is available in API versions `2025-04` and later.
</Note>

Temperature is a parameter that controls the randomness of a model's predictions during text generation. Lower temperatures (\~0.0) yield more consistent, predictable answers, while higher temperatures increase the model's explanatory power and is generally better for creative tasks.

To control the sampling temperature for a model, set the `temperarture` parameter in the request. If a model does not support a temperature parameter, the parameter is ignored.

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone(api_key="YOUR_API_KEY")
  assistant = pc.assistant.Assistant(assistant_name="example-assistant")

  msg = Message(role="user", content="Who is the CFO of Netflix?")
  response = assistant.chat_completions(
      messages=[msg], 
      temperature=0.8
  )

  print(response)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const chatResp = await assistant.chatCompletion({
    messages: [{ role: 'user', content: 'Who is the CFO of Netflix?' }],
    temperature: 0.8,
  });
  console.log(chatResp);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME/chat/completions" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "Who is the CFO of Netflix?"
      }
    ],
    "temperature": 0.8
  }'
  ```
</CodeGroup>

---

# Source: https://docs.pinecone.io/guides/assistant/chat-with-assistant.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Chat through the standard interface

> Chat with your assistant using the standard interface and API.

After [uploading files](/guides/assistant/manage-files) to an assistant, you can chat with the assistant.

<Tip>
  You can chat with an assistant using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/assistant). Select the assistant to chat with, and use the Assistant playground.
</Tip>

## Chat through the standard interface

The [standard chat interface](/reference/api/latest/assistant/chat_assistant) can return responses in three different formats:

* [Default response](#default-response): The assistant returns a structured response and separate citation information.
* [Streaming response](#streaming-response): The assistant returns the response as a text stream.
* [JSON response](#json-response): The assistant returns the response as JSON key-value pairs.

<Tip>
  This is the recommended way to chat with an assistant, as it offers more functionality and control over the assistant's responses and references. However, if you need your assistant to be OpenAI-compatible or need inline citations, use the [OpenAI-compatible chat interface](#chat-through-the-openai-compatible-interface).
</Tip>

### Default response

The following example sends a message and requests a default response:

<Note>
  The `content` parameter in the request cannot be empty.
</Note>

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone(api_key="YOUR_API_KEY")
  assistant = pc.assistant.Assistant(assistant_name="example-assistant")

  msg = Message(role="user", content="Who is the CFO of Netflix?")
  response = assistant.chat(messages=[msg])

  # Alternatively, you can provide a dictionary as the message:
  # msg = {"role": "user", "content": "Who is the CFO of Netflix?"}
  # response = assistant.chat(messages=[msg])

  print(response)

  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const chatResp = await assistant.chat({
    messages: [{ role: 'user', content: 'Who is the CFO of Netflix?' }],
  });
  console.log(chatResp);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "Who is the CFO of Netflix?"
      }
    ],
    "stream": false,
    "model": "gpt-4o"
  }'
  ```
</CodeGroup>

The example above returns a result like the following:

```json JSON theme={null}
{
    "finish_reason": "stop",
    "message": {
        "role": "assistant",
        "content": "The Chief Financial Officer (CFO) of Netflix is Spencer Neumann."
    },
    "id": "00000000...",
    "model": "gpt-4o-2024-11-20",
    "usage": {
        "prompt_tokens": 23633,
        "completion_tokens": 24,
        "total_tokens": 23657
    },
    "citations": [
        {
            "position": 63,
            "references": [
                {
                    "file": {
                        "status": "Available",
                        "id": "76a11dd1...",
                        "name": "Netflix-10-K-01262024.pdf",
                        "size": 1073470,
                        "metadata": {
                            "company": "netflix",
                            "document_type": "form 10k"
                        },
                        "updated_on": "2025-07-16T16:46:40.787204651Z",
                        "created_on": "2025-07-16T16:45:59.414273474Z",
                        "percent_done": 1.0,
                        "signed_url": "https://storage.googleapis.com/...",
                        "error_message": null
                    },
                    "pages": [
                        78,
                        79,
                        80
                    ],
                    "highlight": null
                }
            ]
        }
    ]
}
```

<Warning>
  [`signed_url`](https://cloud.google.com/storage/docs/access-control/signed-urls) provides temporary, read-only access to the relevant file. Anyone with the link can access the file, so treat it as sensitive data. Expires in one hour.
</Warning>

### Streaming response

Streaming responses can improve perceived latency by allowing users to see content as it's generated, rather than waiting for the complete response. This creates a more responsive chat experience, especially for longer responses.

The following example sends a message and requests a streaming response:

<Note>
  The `content` parameter in the request cannot be empty.
</Note>

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone(api_key="YOUR_API_KEY")

  assistant = pc.assistant.Assistant(assistant_name="example-assistant")

  msg = Message(role="user", content="What is the inciting incident of Pride and Prejudice?")

  response = assistant.chat(messages=[msg], stream=True)

  for data in response:
      if data:
          print(data)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
  const assistantName = 'example-assistant';

  const assistant = pc.Assistant(assistantName);
  const chatResp = await assistant.chatStream({
        messages: [{ role: 'user', content: 'Who is the CFO of Netflix?' }]
      });

  for await (const response of chatResp) {
      if (response) {
          console.log(response);
      }
  }
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "What is the inciting incident of Pride and Prejudice?"
      }
    ],
    "stream": true,
    "model": "gpt-4o"
  }'
  ```
</CodeGroup>

The example above returns a result like the following:

```shell  theme={null}
data:{"type":"message_start","id":"0000000000000000111b35de85e8a8f9","model":"gpt-4o-2024-05-13","role":"assistant"}

data:{"type":"content_chunk","id":"0000000000000000111b35de85e8a8f9","model":"gpt-4o-2024-05-13","delta":{"content":"The"}}

...

data:{"type":"citation","id":"0000000000000000111b35de85e8a8f9","model":"gpt-4o-2024-05-13","citation":{"position":406,"references":[{"file":{"status":"Available","id":"ae79e447-b89e-4994-994b-3232ca52a654","name":"Pride-and-Prejudice.pdf","size":2973077,"metadata":null,"updated_on":"2024-06-14T15:01:57.385425746Z","created_on":"2024-06-14T15:01:02.910452398Z","percent_done":0.0,"signed_url":"https://storage.googleapis.com/...", "error_message":null},"pages":[1]}]}}

data:{"type":"message_end","id":"0000000000000000111b35de85e8a8f9","model":"gpt-4o-2024-05-13","finish_reason":"stop","usage":{"prompt_tokens":9736,"completion_tokens":102,"total_tokens":9838}}
```

There are four types of messages in a streaming chat response:

* **Message start**: Includes `"role":"assistant"`, which indicates that the assistant is responding to the user's message.
* **Content**: Includes a value in the `content` field (e.g., `"content":"The"`), which is part of the assistant's streamed response to the user's message.
* **Citation**: Includes a citation to the document that the assistant used to generate the response.
* **Message end**: Includes `"finish_reason":"stop"`, which indicates that the assistant has finished responding to the user's message.

### JSON response

The following example uses the `json_response` parameter to instruct the assistant to return the response as JSON key-value pairs. This is useful if you need to parse the response programmatically.

<Note>
  JSON response cannot be used with the `stream` parameter.
</Note>

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  import json
  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone(api_key="YOUR_API_KEY")

  assistant = pc.assistant.Assistant(assistant_name="example-assistant")

  msg = Message(role="user", content="Who is the CFO and CEO of Netflix?")

  response = assistant.chat(messages=[msg], json_response=True)

  print(json.loads(response))
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const chatResp = await assistant.chat({
    messages: [{ role: 'user', content: 'Who is the CFO and CEO of Netflix?', json_response: true }],
  });
  console.log(chatResp);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "Who is the CFO and CEO of Netflix?"
      }
    ],
    "json_response": true,
    "model": "gpt-4o"
  }'
  ```
</CodeGroup>

The example above returns a result like the following:

```json  theme={null}
{
  "finish_reason": "stop",
  "message": {
    "role": "assistant",
    "content": "{\"CFO\": \"Spencer Neumann\", \"CEO\": \"Ted Sarandos and Greg Peters\"}"
  },
  "id": "0000000000000000680c95d2faab7aad",
  "model": "gpt-4o-2024-11-20",
  "usage": {
    "prompt_tokens": 14298,
    "completion_tokens": 42,
    "total_tokens": 14340
  },
  "citations": [
    {
      "position": 24,
      "references": [
        {
          "file": {
            "status": "Available",
            "id": "cbecaa37-2943-4030-b4d6-ce4350ab774a",
            "name": "Netflix-10-K-01262024.pdf",
            "size": 1073470,
            "metadata": {
              "test-key": "test-value"
            },
            "updated_on": "2025-01-24T16:53:17.148820770Z",
            "created_on": "2025-01-24T16:52:44.851577534Z",
            "percent_done": 1,
            "signed_url": "https://storage.googleapis.com/knowledge-prod-files/bf0dcf22...",
            "error_message": null
          },
          "pages": [
            79
          ],
          "highlight": null
        },
    ...
  ]
}
```

## Extract the response content

In the assistant's response, the message string is contained in the following JSON object:

* `message.content` for the default chat response
* `delta.content` for the streaming chat response
* `message.content` for the JSON response

You can extract the message content and print it to the console:

<Tabs>
  <Tab title="Default response">
    <CodeGroup>
      ```python Python theme={null}
      msg = Message(role="user", content="What is the maximum height of a red pine?")

      response = assistant.chat(messages=[msg])

      print(str(response.message.content))
      ```

      ```javascript JavaScript theme={null}
      const assistant = pc.Assistant(assistantName);
      const chatResp = await assistant.chat({
        messages: [{ role: 'user', content: 'What is the maximum height of a red pine?' }],
      });
      console.log(chatResp.message.content);
      ```

      ```bash curl theme={null}
      | jq '.message.content'
      ```
    </CodeGroup>

    This creates output like the following:

    ```bash  theme={null}
    A red pine, scientifically known as *Pinus resinosa*, is a medium-sized tree that can grow up to 25 meters high and 75 centimeters in diameter. [1, pp. 1]
    ```
  </Tab>

  <Tab title="Streaming response">
    <CodeGroup>
      ```python Python theme={null}
      msg = Message(role="user", content="What is the maximum height of a red pine?")

      response = assistant.chat(messages=[msg], stream=True)

      for data in response:
          if hasattr(data, "delta"):
              print(data.delta.content)
      ```

      ```bash curl theme={null}
      |  sed -u 's/.*"content":"\([^"]*\)".*/\1/'
      ```
    </CodeGroup>

    This creates output like the following:

    ```bash Streaming response theme={null}
    The
     maximum
     height
     of
     a
     red
     pine
     (
    Pin
    us
     resin
    osa
    )
     is
     up
     to
     twenty
    -five
     meters

     [1, pp. 1]
    .
    ```
  </Tab>

  <Tab title="JSON response">
    <CodeGroup>
      ```python Python theme={null}
      import json

      msg = Message(role="user", content="What is the maximum height of a red pine?")

      response = assistant.chat(messages=[msg], json_response=True)

      print(json.loads(response.message.content))
      ```

      ```bash curl theme={null}
      |  sed -u 's/.*"content":"\([^"]*\)".*/\1/'
      ```
    </CodeGroup>

    This creates output like the following:

    ```bash JSON response theme={null}
    {'red pine': 'A red pine, scientifically known as *Pinus resinosa*, is a medium-sized tree that can grow up to 25 meters high and 75 centimeters in diameter.'}
    ```
  </Tab>
</Tabs>

## Choose a model

Pinecone Assistant supports the following models:

* `gpt-4o` (default)
* `gpt-4.1`
* `gpt-5`
* `o4-mini`
* `claude-sonnet-4-5`
* `gemini-2.5-pro`

<Note>
  Anthropic has [deprecated](https://platform.claude.com/docs/en/about-claude/model-deprecations) the Claude 3.5 Sonnet and Claude 3.7 Sonnet models. Assistant automatically routes chat requests that specify `claude-3-5-sonnet` or `claude-3-7-sonnet` to `claude-sonnet-4-5` at the same price.
</Note>

<Tip>
  For chat applications, we recommend using GPT models (`gpt-4o`, `gpt-4.1`, `gpt-5`, or `o4-mini`) as they typically provide faster response times compared to other models.
</Tip>

To choose a non-default model for your assistant, set the `model` parameter in the request:

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone(api_key="YOUR_API_KEY")

  # Get your assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant", 
  )

  # Chat with the assistant.
  chat_context = [Message(role="user", content="What is the maximum height of a red pine?")]
  response = assistant.chat(
      messages=chat_context, 
      model="gpt-4.1"
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const chatResp = await assistant.chat({
    messages: [{ role: 'user', content: 'What is the maximum height of a red pine?' }],
    model: 'gpt-4.1',
  });

  console.log(chatResp);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "What is the maximum height of a red pine?"
      }
    ],
    "model": "gpt-4.1"
  }'
  ```
</CodeGroup>

## Provide conversation history

Models lack memory of previous requests, so any relevant messages from earlier in the conversation must be present in the `messages` object.

In the following example, the `messages` object includes prior messages that are necessary for interpreting the newest message.

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone(api_key="YOUR_API_KEY")

  # Get your assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant", 
  )

  # Chat with the assistant.
  chat_context = [
      Message(content="What is the maximum height of a red pine?", role="user"),
      Message(content="The maximum height of a red pine (Pinus resinosa) is up to 25 meters.", role="assistant"),
      Message(content="What is its maximum diameter?", role="user")
  ]
  response = assistant.chat(messages=chat_context)
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY " \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "What is the maximum height of a red pine?"
      },
      {
        "role": "assistant",
        "content": "The maximum height of a red pine (Pinus resinosa) is up to 25 meters."
      },
      {
        "role": "user",
        "content": "What is its maximum diameter?"
      }
    ]
  }'
  ```
</CodeGroup>

The example returns a response like the following:

```JSON  theme={null}
{
  "finish_reason":"stop",
  "message":{
    "role":"assistant",
    "content":"The maximum diameter of a red pine (Pinus resinosa) is up to 1 meter."
    },
    "id":"0000000000000000236a24a17e55309a",
    "model":"gpt-4o-2024-05-13",
    "usage":{
      "prompt_tokens":21377,
      "completion_tokens":20,
      "total_tokens":21397
      },
      "citations":[...]
}
```

## Filter chat with metadata

You can [filter which documents to use for chat completions](/guides/assistant/files-overview#file-metadata). The following example filters the responses to use only documents that include the metadata `"resource": "encyclopedia"`.

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone(api_key="YOUR_API_KEY")

  # Get your assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant", 
  )

  # Chat with the assistant.
  chat_context = [Message(role="user", content="What is the maximum height of a red pine?")]
  response = assistant.chat(messages=chat_context, stream=True, filter={"resource": "encyclopedia"})
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const chatResp = await assistant.chat({
    messages: [{ role: 'user', content: 'What is the maximum height of a red pine?' }],
    filter: {
      'resource': 'encyclopedia'
    }
  });
  console.log(chatResp);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY "\
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "What is the maximum height of a red pine?"
      }
    ],
    "stream": true,
    "filter": 
      {
      "resource": "encyclopedia"
      }
    }'
  ```
</CodeGroup>

## Control the context size

<Note>
  This is available in API versions `2025-04` and later.
</Note>

To limit the number of [input tokens](/guides/assistant/pricing-and-limits#token-usage) used, you can control the context size by tuning `top_k * snippet_size`. These parameters can be adjusted by setting [`context_options`](/reference/api/latest/assistant/chat_assistant#body-context-options) in the request:

* `snippet_size`: Controls the max size of a snippet (default is 2048 tokens). Note that snippet size can vary and, in rare cases, may be bigger than the set `snippet_size`. Snippet size controls the amount of context the model is given for each chunk of text.
* `top_k`: Controls the max number of context snippets sent to the LLM (default is 16). `top_k` controls the diversity of information sent to the model.

While additional tokens will be used for other parameters (e.g., the system prompt, chat input), adjusting the `top_k` and `snippet_size` can help manage token consumption.

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone(api_key="YOUR_API_KEY")
  assistant = pc.assistant.Assistant(assistant_name="example-assistant")

  msg = Message(role="user", content="Who is the CFO of Netflix?")
  response = assistant.chat(messages=[msg], context_options={snippet_size=2500, top_k=10})

  print(response)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const chatResp = await assistant.chat({
    messages: [{ role: 'user', content: 'Who is the CFO of Netflix?' }],
    contextOptions: { topK: 10, snippetSize: 2500 },
  });

  console.log(chatResp);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "Who is the CFO of Netflix?"
      }
    ],
    "context_options": {
      "top_k":10,
      "snippet_size":2500
      }
  }'
  ```
</CodeGroup>

The example will return up to 10 snippets and each snippet will be up to 2500 tokens in size.

<Tip>
  To better understand the context retrieved using these parameters, you can [retrieve context from an assistant](/reference/api/latest/assistant/context_assistant).
</Tip>

## Set the sampling temperature

<Note>
  This is available in API versions `2025-04` and later.
</Note>

Temperature is a parameter that controls the randomness of a model's predictions during text generation. Lower temperatures (\~0.0) yield more consistent, predictable answers, while higher temperatures increase the model's explanatory power and is generally better for creative tasks.

To control the sampling temperature for a model, set the `temperarture` parameter in the request. If a model does not support a temperature parameter, the parameter is ignored.

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone(api_key="YOUR_API_KEY")
  assistant = pc.assistant.Assistant(assistant_name="example-assistant")

  msg = Message(role="user", content="Who is the CFO of Netflix?")
  response = assistant.chat(
      messages=[msg], 
      temperature=0.8
  )

  print(response)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const chatResp = await assistant.chat({
    messages: [{ role: 'user', content: 'Who is the CFO of Netflix?' }],
    temperature: 0.8,
  });
  console.log(chatResp);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "Who is the CFO of Netflix?"
      }
    ],
    "temperature": 0.8
  }'
  ```
</CodeGroup>

## Include citation highlights in the response

<Note>
  Citation highlights are available in the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/assistant) or API versions `2025-04` and later.
</Note>

When using the [standard chat interface](/reference/api/latest/assistant/chat_assistant), every response includes a `citation` object. The object includes a reference to the document that the assistant used to generate the response. Additionally, you can include highlights, which are the specific parts of the document that the assistant used to generate the response, by setting the `include_highlights` parameter to `true` in the request:

```bash curl theme={null}
PINECONE_API_KEY="YOUR_API_KEY"
ASSISTANT_NAME="example-assistant"

curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME" \
  -H "Api-Key: $PINECONE_API_KEY" \
  -H "Content-Type: application/json" \
  -H "X-Pinecone-Api-Version: 2025-10" \
  -d '{
  "messages": [
    {
      "role": "user",
      "content": "Who is the CFO of Netflix?"
    }
  ],
  "stream": false,
  "model": "gpt-4o",
  "include_highlights": true
}'
```

The example returns response like the following:

```json  theme={null}
{
  "finish_reason":"stop",
  "message":{
    "role":"assistant",
    "content":"The Chief Financial Officer (CFO) of Netflix is Spencer Neumann."
    },
    "id":"00000000000000006685b07087b1ad42",
    "model":"gpt-4o-2024-05-13",
    "usage":{
      "prompt_tokens":12490,
      "completion_tokens":33,
      "total_tokens":12523
      },
      "citations":[{
        "position":63,
        "references":[{
          "file":{
            "status":"Available",
            "id":"cbecaa37-2943-4030-b4d6-ce4350ab774a",
            "name":"Netflix-10-K-01262024.pdf",
            "size":1073470,
            "metadata":{"test-key":"test-value"},
            "updated_on":"2025-01-24T16:53:17.148820770Z",
            "created_on":"2025-01-24T16:52:44.851577534Z",
            "percent_done":1.0,
            "signed_url":"https://storage.googleapis.com/knowledge-prod-files/b...",
            "error_message":null
            },
            "pages":[78],
            "highlight":{
              "type":"text",
              "content":"EXHIBIT 31.3\nCERTIFICATION OF CHIEF FINANCIAL OFFICER\nPURSUANT TO SECTION 302 OF THE SARBANES-OXLEY ACT OF 2002\nI, Spencer Neumann, certify that:"
              }
            },
            {
              "file":{
                "status":"Available",
                "id":"cbecaa37-2943-4030-b4d6-ce4350ab774a",
                "name":"Netflix-10-K-01262024.pdf",
                "size":1073470,
                "metadata":{"test-key":"test-value"},
                "updated_on":"2025-01-24T16:53:17.148820770Z",
                "created_on":"2025-01-24T16:52:44.851577534Z",
                "percent_done":1.0,
                "signed_url":"https://storage.googleapis.com/knowledge-prod-files/bf...",
                "error_message":null
                },
                "pages":[79],
                "highlight":{
                  "type":"text",
                  "content":"operations of\nNetflix, Inc.\nDated: January 26, 2024  By:  /S/ SPENCER NEUMANN\n  Spencer Neumann\n  Chief Financial Officer"
                }
            }
          ]
        }
      ]
}
```

<Note>
  Enabling highlights will increase token usage.
</Note>

---

# Source: https://docs.pinecone.io/reference/api/2025-10/assistant/chat_assistant.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Chat with an assistant

> Chat with an assistant and get back citations in structured form. 

This is the recommended way to chat with an assistant, as it offers more functionality and control over the assistant's responses and references than the OpenAI-compatible chat interface.

For guidance and examples, see [Chat with an assistant](https://docs.pinecone.io/guides/assistant/chat-with-assistant).

<RequestExample>
  ```bash curl | Default theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "What is the inciting incident of Pride and Prejudice?"
      }
    ],
    "stream": false,
    "model": "gpt-4o"
  }'
  ```

  ```bash curl | Streaming theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "What is the inciting incident of Pride and Prejudice?"
      }
    ],
    "stream": true,
    "model": "gpt-4o"
  }'
  ```
</RequestExample>

<ResponseExample>
  ```json Default response theme={null}
  {
    "finish_reason": "stop",
    "message": {
      "role": "assistant",
      "content": "The inciting incident of \"Pride and Prejudice\" occurs when Mrs. Bennet informs Mr. Bennet that Netherfield Park has been let at last, and she is eager to share the news about the new tenant, Mr. Bingley, who is wealthy and single. This sets the stage for the subsequent events of the story, including the introduction of Mr. Bingley and Mr. Darcy to the Bennet family and the ensuing romantic entanglements."
    },
    "id": "00000000000000004ac3add5961aa757",
    "model": "gpt-4o-2024-05-13",
    "usage": {
      "prompt_tokens": 9736,
      "completion_tokens": 105,
      "total_tokens": 9841
    },
    "citations": [
      {
        "position": 406,
        "references": [
          {
            "file": {
              "status": "Available",
              "id": "ae79e447-b89e-4994-994b-3232ca52a654",
              "name": "Pride-and-Prejudice.pdf",
              "size": 2973077,
              "metadata": null,
              "updated_on": "2024-06-14T15:01:57.385425746Z",
              "created_on": "2024-06-14T15:01:02.910452398Z",
              "percent_done": 0,
              "signed_url": "https://storage.googleapis.com/...",
              "error_message": null
            },
            "pages": [
              1
            ]
          }
        ]
      }
    ]
  }

  ```

  ```shell Streaming response theme={null}
  data:{
    "type":"message_start",
    "id":"0000000000000000111b35de85e8a8f9",
    "model":"gpt-4o-2024-05-13",
    "role":"assistant"
  }

  data:
  {
    "type":"content_chunk",
    "id":"0000000000000000111b35de85e8a8f9",
    "model":"gpt-4o-2024-05-13",
    "delta":
    {
      "content":"The"
      }
  }

  ...

  data:
  {
    "type":"citation",
    "id":"0000000000000000111b35de85e8a8f9",
    "model":"gpt-4o-2024-05-13",
    "citation":
    {
      "position":406,
      "references":
      [
        {
          "file":{
            "status":"Available",
            "id":"ae79e447-b89e-4994-994b-3232ca52a654",
            "name":"Pride-and-Prejudice.pdf",
            "size":2973077,
            "metadata":null,
            "updated_on":"2024-06-14T15:01:57.385425746Z", 
            "created_on":"2024-06-14T15:01:02.910452398Z",
            "percent_done":0.0,
            "signed_url":"https://storage.googleapis.com/...",
            "error_message":null
            }, 
        "pages":[1]
        }
      ]
    }
  }

  data:
  {
    "type":"message_end",
    "id":"0000000000000000111b35de85e8a8f9",
    "model":"gpt-4o-2024-05-13",
    "finish_reason":"stop",
    "usage":
    {
      "prompt_tokens":9736,
      "completion_tokens":102,
      "total_tokens":9838
      }
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/assistant_data_2025-10.oas.yaml post /chat/{assistant_name}
openapi: 3.0.3
info:
  title: Pinecone Assistant Data Plane API
  description: >-
    Pinecone Assistant Engine is a context engine to store and retrieve relevant
    knowledge from millions of documents at scale. This API supports
    interactions with assistants.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{assistant_host}
    variables:
      assistant_host:
        default: unknown
        description: host of the created assistant
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Assistants
    description: Actions that manage Assistants
paths:
  /chat/{assistant_name}:
    post:
      tags:
        - Manage Assistants
      summary: Chat with an assistant
      description: >-
        Chat with an assistant and get back citations in structured form. 


        This is the recommended way to chat with an assistant, as it offers more
        functionality and control over the assistant's responses and references
        than the OpenAI-compatible chat interface.


        For guidance and examples, see [Chat with an
        assistant](https://docs.pinecone.io/guides/assistant/chat-with-assistant).
      operationId: chat_assistant
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: assistant_name
          description: The name of the assistant to be described.
          required: true
          schema:
            type: string
          example: test-assistant
          style: simple
      requestBody:
        description: The desired configuration to chat an assistant.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ChatRequest'
        required: true
      responses:
        '200':
          description: Search request successful.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ChatModel'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                files-validation-error:
                  summary: Validation error on ingest.
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Uploaded file can only currently be either a pdf or txt
                        file
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Assistant not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                assistant-not-found:
                  summary: Assistant not found.
                  value:
                    error:
                      code: NOT_FOUND
                      message: Assistant "example-assistant" not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    ChatRequest:
      description: The list of queries / chats to chat an assistant
      type: object
      properties:
        messages:
          type: array
          items:
            $ref: '#/components/schemas/MessageModel'
        stream:
          description: >-
            If false, the assistant will return a single JSON response. If true,
            the assistant will return a stream of responses.
          default: false
          type: boolean
        model:
          description: The large language model to use for answer generation
          default: gpt-4o
          x-enum:
            - gpt-4o
            - gpt-4.1
            - o4-mini
            - claude-3-5-sonnet
            - claude-3-7-sonnet
            - gemini-2.5-pro
          type: string
        temperature:
          description: >-
            Controls the randomness of the model's output: lower values make
            responses more deterministic, while higher values increase
            creativity and variability. If the model does not support a
            temperature parameter, the parameter will be ignored.
          default: 0
          type: number
          format: float
        filter:
          example:
            genre:
              $ne: documentary
          description: >-
            Optionally filter which documents can be retrieved using the
            following metadata fields.
          type: object
        json_response:
          description: >-
            If true, the assistant will be instructed to return a JSON response.
            Cannot be used with streaming.
          default: false
          type: boolean
        include_highlights:
          description: >-
            If true, the assistant will be instructed to return highlights from
            the referenced documents that support its response.
          default: false
          type: boolean
        context_options:
          $ref: '#/components/schemas/ContextOptionsModel'
      required:
        - messages
    ChatModel:
      description: Describes the response format of a chat request from the citation API.
      type: object
      properties:
        id:
          type: string
        finish_reason:
          x-enum:
            - stop
            - length
            - content_filter
            - function_call
          type: string
        message:
          $ref: '#/components/schemas/MessageModel'
        model:
          type: string
        citations:
          type: array
          items:
            $ref: '#/components/schemas/CitationModel'
        usage:
          $ref: '#/components/schemas/UsageModel'
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: Uploaded file can only currently be either a pdf or txt file
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    MessageModel:
      description: Describes the format of a message in a chat.
      type: object
      properties:
        role:
          description: Role of the message such as 'user' or 'assistant'
          type: string
        content:
          description: Content of the message
          type: string
    ContextOptionsModel:
      description: Controls the context snippets sent to the LLM.
      type: object
      properties:
        top_k:
          example: 20
          description: >-
            The maximum number of context snippets to use. Default is 16.
            Maximum is 64.
          type: integer
        snippet_size:
          example: 4096
          description: >-
            The maximum context snippet size. Default is 2048 tokens. Minimum is
            512 tokens. Maximum is 8192 tokens.
          type: integer
        multimodal:
          description: >-
            Whether or not to send image-related context snippets to the LLM. If
            `false`, only text context snippets are sent.
          default: true
          type: boolean
        include_binary_content:
          description: >-
            If image-related context snippets are sent to the LLM, this field
            determines whether or not they should include base64 image data. If
            `false`, only the image caption is sent. Only available when
            `multimodal=true`.
          default: true
          type: boolean
    CitationModel:
      description: Describes a single cited source returned by a chat request.
      type: object
      properties:
        position:
          description: The index position of the citation in the complete text response.
          type: integer
        references:
          type: array
          items:
            $ref: '#/components/schemas/ReferenceModel'
    UsageModel:
      description: Describes the usage of a chat completion.
      type: object
      properties:
        prompt_tokens:
          type: integer
        completion_tokens:
          type: integer
        total_tokens:
          type: integer
    ReferenceModel:
      description: Describes a single reference in a citation.
      type: object
      properties:
        file:
          $ref: '#/components/schemas/AssistantFileModel'
        pages:
          type: array
          items:
            type: integer
        highlight:
          $ref: '#/components/schemas/HighlightModel'
    AssistantFileModel:
      description: The response format for a successful file upload request.
      type: object
      properties:
        name:
          type: string
        id:
          type: string
          format: uuid
        metadata:
          nullable: true
          type: object
        created_on:
          type: string
          format: date-time
        updated_on:
          type: string
          format: date-time
        status:
          description: >-
            The current state of the uploaded file. Possible values:

            - `Processing`: File is being processed (parsed, chunked, embedded)

            - `Available`: Processing completed successfully; file is ready for
            use

            - `Deleting`: Deletion has been initiated but not yet completed

            - `ProcessingFailed`: Processing failed with an error


            Note: Once a file is deleted, the API returns 404 Not Found instead
            of a file object.
          x-enum:
            - Processing
            - Available
            - Deleting
            - ProcessingFailed
          type: string
        percent_done:
          nullable: true
          description: The percentage of the file that has been processed
          type: number
          format: double
        signed_url:
          nullable: true
          example: https://storage.googleapis.com/bucket/file.pdf?...
          description: >-
            A [signed
            URL](https://cloud.google.com/storage/docs/access-control/signed-urls)
            that provides temporary, read-only access to the underlying file.
            Anyone with the link can access the file, so treat it as sensitive
            data. Expires after a short time.
          type: string
        error_message:
          nullable: true
          description: >-
            A message describing any error during file processing. Provided only
            if an error occurs.
          type: string
        multimodal:
          description: Indicates whether the file was processed as multimodal.
          type: boolean
      required:
        - id
        - name
    HighlightModel:
      nullable: true
      description: >-
        Represents a portion of a referenced document that directly supports or
        is relevant to the response.
      type: object
      properties:
        type:
          description: The type of the highlight. Currently it is always text.
          type: string
        content:
          type: string
      required:
        - type
        - content
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: Pinecone API Key

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/assistant/chat_completion_assistant.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Chat through an OpenAI-compatible interface

> Chat with an assistant. This endpoint is based on the OpenAI Chat Completion API, a commonly used and adopted API. 

It is useful if you need inline citations or OpenAI-compatible responses, but has limited functionality compared to the standard chat interface.

For guidance and examples, see [Chat with an assistant](https://docs.pinecone.io/guides/assistant/chat-with-assistant).

<RequestExample>
  ```bash curl | Default theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME/chat/completions" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "What is the maximum height of a red pine?"
      }
    ]
  }'
  ```

  ```bash curl | Streaming theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME/chat/completions" \
    -H "Api-Key: $PINECONE_API_KEY "\
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "messages": [
      {
        "role": "user",
        "content": "What is the maximum height of a red pine?"
      }
    ],
    "stream": true
  }'
  ```
</RequestExample>

<ResponseExample>
  ```JSON Default response theme={null}
  {"chat_completion":
    {
      "id":"chatcmpl-9OtJCcR0SJQdgbCDc9JfRZy8g7VJR",
      "choices":[
        {
          "finish_reason":"stop",
          "index":0,
          "message":{
            "role":"assistant",
            "content":"The maximum height of a red pine (Pinus resinosa) is up to 25 meters."
          }
        }
      ],
      "model":"my_assistant"
    }
  }
  ```

  ```shell Streaming response theme={null}
  {
    'id': '000000000000000009de65aa87adbcf0', 
    'choices': [
        {
        'index': 0, 
        'delta': 
          {
          'role': 'assistant', 
          'content': 'The'
          }, 
        'finish_reason': None
        }
      ], 
    'model': 'gpt-4o-2024-05-13'
  }

  ...

  {
    'id': '00000000000000007a927260910f5839',
    'choices': [
        {
        'index': 0,
        'delta':
          {
            'role': '', 
            'content': 'The'
          }, 
        'finish_reason': None
        }
      ], 
    'model': 'gpt-4o-2024-05-13'
  }

  ...

  {
    'id': '00000000000000007a927260910f5839', 
    'choices': [
      {
        'index': 0, 
        'delta': 
          {
          'role': None, 
          'content': None
          }, 
        'finish_reason': 'stop'
        }
      ], 
    'model': 'gpt-4o-2024-05-13'
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/assistant_data_2025-10.oas.yaml post /chat/{assistant_name}/chat/completions
openapi: 3.0.3
info:
  title: Pinecone Assistant Data Plane API
  description: >-
    Pinecone Assistant Engine is a context engine to store and retrieve relevant
    knowledge from millions of documents at scale. This API supports
    interactions with assistants.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{assistant_host}
    variables:
      assistant_host:
        default: unknown
        description: host of the created assistant
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Assistants
    description: Actions that manage Assistants
paths:
  /chat/{assistant_name}/chat/completions:
    post:
      tags:
        - Manage Assistants
      summary: Chat through an OpenAI-compatible interface
      description: >-
        Chat with an assistant. This endpoint is based on the OpenAI Chat
        Completion API, a commonly used and adopted API. 


        It is useful if you need inline citations or OpenAI-compatible
        responses, but has limited functionality compared to the standard chat
        interface.


        For guidance and examples, see [Chat with an
        assistant](https://docs.pinecone.io/guides/assistant/chat-with-assistant).
      operationId: chat_completion_assistant
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: assistant_name
          description: The name of the assistant to be described.
          required: true
          schema:
            type: string
          example: test-assistant
          style: simple
      requestBody:
        description: The desired configuration to chat an assistant.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/SearchCompletions'
        required: true
      responses:
        '200':
          description: Search request successful.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ChatCompletionModel'
            text/event-stream:
              schema:
                $ref: '#/components/schemas/StreamChatCompletionChunkModel'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                files-validation-error:
                  summary: Validation error on ingest.
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Uploaded file can only currently be either a pdf or txt
                        file
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Assistant not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                assistant-not-found:
                  summary: Assistant not found.
                  value:
                    error:
                      code: NOT_FOUND
                      message: Assistant "example-assistant" not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    SearchCompletions:
      description: The list of queries / chats to chat an assistant
      type: object
      properties:
        messages:
          type: array
          items:
            $ref: '#/components/schemas/MessageModel'
        stream:
          description: >-
            If false, the assistant will return a single JSON response. If true,
            the assistant will return a stream of responses.
          default: false
          type: boolean
        model:
          description: The large language model to use for answer generation
          default: gpt-4o
          x-enum:
            - gpt-4o
            - gpt-4.1
            - o4-mini
            - claude-3-5-sonnet
            - claude-3-7-sonnet
            - gemini-2.5-pro
          type: string
        temperature:
          description: >-
            Controls the randomness of the model's output: lower values make
            responses more deterministic, while higher values increase
            creativity and variability. If the model does not support a
            temperature parameter, the parameter will be ignored.
          default: 0
          type: number
          format: float
        filter:
          example:
            genre:
              $ne: documentary
          description: >-
            Optionally filter which documents can be retrieved using the
            following metadata fields.
          type: object
      required:
        - messages
    ChatCompletionModel:
      description: Describes the response format of a chat request.
      type: object
      properties:
        id:
          type: string
        choices:
          type: array
          items:
            $ref: '#/components/schemas/ChoiceModel'
        model:
          type: string
        usage:
          $ref: '#/components/schemas/UsageModel'
    StreamChatCompletionChunkModel:
      description: Describes the response format of a chat request.
      type: object
      properties:
        id:
          type: string
        choices:
          type: array
          items:
            $ref: '#/components/schemas/ChoiceChunkModel'
        model:
          type: string
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: Uploaded file can only currently be either a pdf or txt file
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    MessageModel:
      description: Describes the format of a message in a chat.
      type: object
      properties:
        role:
          description: Role of the message such as 'user' or 'assistant'
          type: string
        content:
          description: Content of the message
          type: string
    ChoiceModel:
      description: Describes a single choice in a chat completion response.
      type: object
      properties:
        finish_reason:
          x-enum:
            - stop
            - length
            - content_filter
            - function_call
          type: string
        index:
          type: integer
        message:
          $ref: '#/components/schemas/MessageModel'
    UsageModel:
      description: Describes the usage of a chat completion.
      type: object
      properties:
        prompt_tokens:
          type: integer
        completion_tokens:
          type: integer
        total_tokens:
          type: integer
    ChoiceChunkModel:
      description: Describes a single choice in a chat completion response.
      type: object
      properties:
        finish_reason:
          x-enum:
            - stop
            - length
            - content_filter
            - function_call
          type: string
        index:
          type: integer
        delta:
          description: Chat completion message
          type: object
          properties:
            role:
              type: string
            content:
              type: string
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: Pinecone API Key

````

---

# Source: https://docs.pinecone.io/guides/index-data/check-data-freshness.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Check data freshness

> Monitor data freshness in Pinecone using log sequence numbers and vector counts.

Pinecone is eventually consistent, so there can be a slight delay before new or changed records are visible to queries. This page describes two ways of checking the data freshness of a Pinecone index:

* To check if a serverless index queries reflect recent writes to the index, [check the log sequence number](#check-the-log-sequence-number).

* To check whether an index contains recently inserted or deleted vectors, [verify the number of vectors in the index](#verify-vector-counts).

## Check the log sequence number

<Note>
  This method is only available for serverless indexes through the [Database API](https://docs.pinecone.io/reference/api/latest/data-plane/upsert).
</Note>

### Log sequence numbers

When you make a write request to a serverless index namespace, Pinecone assigns a monotonically increasing log sequence number (LSN) to the write operation. The LSN reflects upserts as well as updates and deletes to that namespace. Writes to one namespace do not increase the LSN for other namespaces.

You can use LSNs to verify that specific write operations are reflected in your query responses. If the LSN contained in the query response header is greater than or equal to the LSN of the relevant write operation, then that operation is reflected in the query response. If the LSN contained in the query response header is *greater than* the LSN of the relevant write operation, then subsequent operations are also reflected in the query response.

Follow the steps below to compare the LSNs for a write and a subsequent query.

### 1. Get the LSN for a write operation

Every time you modify records in your namespace, the HTTP response contains the LSN for the upsert. This is contained in a header called `x-pinecone-request-lsn`.

The following example demonstrates how to get the LSN for an `upsert` request using the `curl` option `-i`. This option tells curl to include headers in the displayed response. Use the same method to get the LSN for an `update` or `delete` request.

```shell curl theme={null}
PINECONE_API_KEY="YOUR_API_KEY"
INDEX_HOST="INDEX_HOST"

curl -i "https://$INDEX_HOST/vectors/upsert" \
  -H "Api-Key: $PINECONE_API_KEY" \
  -H "content-type: application/json" \
  -H "X-Pinecone-Api-Version: 2025-10" \
  -d '{
        "vectors": [
          {
            "id": "vec1",
            "values": [0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8],
          }
        ],
        "namespace": "example-namespace"
      }'
```

The preceding request receives a response like the following example:

```shell curl theme={null}
HTTP/2 200 
date: Wed, 21 Aug 2024 15:23:04 GMT
content-type: application/json
content-length: 66
x-pinecone-max-indexed-lsn: 4
x-pinecone-request-latency-ms: 1149
x-pinecone-request-id: 3687967458925971419
x-envoy-upstream-service-time: 1150
grpc-status: 0
server: envoy

{"upsertedCount":1}
```

In the preceding example response, the value of `x-pinecone-max-indexed-lsn` is 4. This means that the index has performed 4 write operations since its creation.

### 2. Get the LSN for a query

Every time you query your index, the HTTP response contains the LSN for the query. This is contained in a header called `x-pinecone-max-indexed-lsn`.

By checking the LSN in your query results, you can confirm that the LSN is greater than or equal to the LSN of the relevant write operation, indicating that the results of that operation are present in the query results.

The following example makes a `query` request to the index:

```shell curl theme={null}
PINECONE_API_KEY="YOUR_API_KEY"
INDEX_HOST="INDEX_HOST"

curl -i "https://$INDEX_HOST/query" \
  -H "Api-Key: $PINECONE_API_KEY" \
  -H 'Content-Type: application/json' \
  -H "X-Pinecone-Api-Version: 2025-10" \
  -d '{
    "vector": [0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3],
    "namespace": "example-namespace",
    "topK": 3,
    "includeValues": true
  }'
```

The preceding request receives a response like the following example:

```shell  theme={null}
HTTP/2 200 
date: Wed, 21 Aug 2024 15:33:36 GMT
content-type: application/json
content-length: 66
x-pinecone-max-indexed-lsn: 5
x-pinecone-request-latency-ms: 40
x-pinecone-request-id: 6683088825552978933
x-envoy-upstream-service-time: 41
grpc-status: 0
server: envoy

{
  "results":[],
  "matches":[
    {
      "id":"vec1",
      "score":0.891132772,
      "values":[0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8],
    }
  ],
  "namespace":"example-namespace",
  "usage":{"readUnits":6}
}
```

In the preceding example response, the value of `x-pinecone-max-indexed-lsn` is 5.

### 3. Compare LSNs for writes and queries

If the LSN of a query is greater than or equal to the LSN for a write operation, then the results of the query reflect the results of the write operation.

In [step 1](#1-get-the-lsn-for-a-write-operation), the LSN contained in the response headers is 4.

In [step 2](#2-get-the-lsn-for-a-query), the LSN contained in the response headers is 5.

5 is greater than or equal to 4; therefore, the results of the query reflect the results of the upsert. However, this does not guarantee that the records upserted are still present or unmodified: the write operation with LSN of 5 may have updated or deleted these records, or upserted additional records.

## Verify record counts

If you insert new records or delete records, the number of records in the index may change. This means that the record count for an index can indicate whether Pinecone has indexed your latest inserts and deletes: if the record count for the index matches the count you expect after inserting or deleting records, the index is probably up-to-date. However, this is not always true. For example, if you delete the same number of records that you insert, the expected record count may remain the same. Also, some write operations, such as updates to an index configuration or vector data values, do not change the number of records in the index.

To verify that your index contains the number of records you expect, [view index stats](/reference/api/latest/data-plane/describeindexstats):

<CodeGroup>
  ```python Python theme={null}
  # pip install "pinecone[grpc]"
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  index.describe_index_stats()
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  const stats = await index.describeIndexStats();
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import io.pinecone.proto.DescribeIndexStatsResponse;

  public class DescribeIndexStatsExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");
          DescribeIndexStatsResponse indexStatsResponse = index.describeIndexStats();
          System.out.println(indexStatsResponse);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      stats, err := idxConnection.DescribeIndexStats(ctx)
      if err != nil {
          log.Fatalf("Failed to describe index \"%v\": %v", idx.Name, err)
      } else {
          fmt.Printf("%+v", *stats)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var indexStatsResponse = await index.DescribeIndexStatsAsync(new DescribeIndexStatsRequest());

  Console.WriteLine(indexStatsResponse);
  ```

  ```shell curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X POST "https://$INDEX_HOST/describe_index_stats" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

The response will look like this:

<CodeGroup>
  ```Python Python theme={null}
  {'dimension': 1024,
   'index_fullness': 0,
   'namespaces': {'example-namespace1': {'vector_count': 4}, 'example-namespace2': {'vector_count': 4}},
   'total_vector_count': 8}
  ```

  ```JavaScript JavaScript theme={null}
  Returns:
  {
    namespaces: { example-namespace1: { recordCount: 4 }, example-namespace2: { recordCount: 4 } },
    dimension: 1024,
    indexFullness: 0,
    totalRecordCount: 8
  }

  // Note: the value of totalRecordCount is the same as total_vector_count.
  ```

  ```java Java theme={null}
  namespaces {
    key: "example-namespace1"
    value {
      vector_count: 4
    }
  }
  namespaces {
    key: "example-namespace2"
    value {
      vector_count: 4
    }
  }
  dimension: 1024
  total_vector_count: 8
  ```

  ```go Go theme={null}
  {
    "dimension": 1024,
    "index_fullness": 0,
    "total_vector_count": 8,
    "namespaces": {
      "example-namespace1": {
        "vector_count": 4
      },
      "example-namespace2": {
        "vector_count": 4
      }
    }
  }
  ```

  ```csharp C# theme={null}
  {
    "namespaces": {
      "example-namespace1": {
        "vectorCount": 4
      },
      "example-namespace2": {
        "vectorCount": 4
      }
    },
    "dimension": 1024,
    "indexFullness": 0,
    "totalVectorCount": 8
  }
  ```

  ```shell curl theme={null}
  {
    "namespaces": {
      "example-namespace1": {
        "vectorCount": 4
      },
      "example-namespace2": {
        "vectorCount": 4
      }
    },
    "dimension": 1024,
    "indexFullness": 0,
    "totalVectorCount": 8
  }
  ```
</CodeGroup>

---

# Source: https://docs.pinecone.io/guides/indexes/pods/choose-a-pod-type-and-size.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Choose a pod type and size

> Select the right pod configuration for your workload

<Warning>
  Customers who sign up for a Standard or Enterprise plan on or after August 18, 2025 cannot create pod-based indexes. Instead, create [serverless indexes](/guides/index-data/create-an-index), and consider using [dedicated read nodes](/guides/index-data/dedicated-read-nodes) for large workloads (millions of records or more, and moderate or high query rates).
</Warning>

When planning your Pinecone deployment, it is important to understand the approximate storage requirements of your vectors to choose the appropriate pod type and number. This page will give guidance on sizing to help you plan accordingly.

As with all guidelines, these considerations are general and may not apply to your specific use case. We caution you to always test your deployment and ensure that the index configuration you are using is appropriate to your requirements.

[Collections](/guides/indexes/pods/understanding-collections) allow you to create new versions of your index with different pod types and sizes. This also allows you to test different configurations. This guide is merely an overview of sizing considerations; test your index configuration before moving to production.

<Tip>
  Users on Standard and Enterprise plans can [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket) for further help with sizing and testing.
</Tip>

## Overview

There are five main considerations when deciding how to configure your Pinecone index:

* Number of vectors
* Dimensionality of your vectors
* Size of metadata on each vector
* Queries per second (QPS) throughput
* Cardinality of indexed metadata

Each of these considerations comes with requirements for index size, pod type, and replication strategy.

### Number of vectors

The most important consideration in sizing is the [number of vectors](/guides/index-data/upsert-data) you plan on working with. As a rule of thumb, a single p1 pod can store approximately 1M vectors, while a s1 pod can store 5M vectors. However, this can be affected by other factors, such as dimensionality and metadata, which are explained below.

### Dimensionality of vectors

The rules of thumb above for how many vectors can be stored in a given pod assumes a typical configuration of 768 [dimensions per vector](/guides/index-data/create-an-index). As your individual use case will dictate the dimensionality of your vectors, the amount of space required to store them may necessarily be larger or smaller.

Each dimension on a single vector consumes 4 bytes of memory and storage per dimension, so if you expect to have 1M vectors with 768 dimensions each, that’s about 3GB of storage without factoring in metadata or other overhead. Using that reference, we can estimate the typical pod size and number needed for a given index. Table 1 below gives some examples of this.

**Table 1: Estimated number of pods per 1M vectors by dimensionality**

| Pod type | Dimensions | Estimated max vectors per pod |
| -------- | ---------: | ----------------------------: |
| **p1**   |        512 |                     1,250,000 |
|          |        768 |                     1,000,000 |
|          |       1024 |                       675,000 |
|          |       1536 |                       500,000 |
| **p2**   |        512 |                     1,250,000 |
|          |        768 |                     1,100,000 |
|          |       1024 |                     1,000,000 |
|          |       1536 |                       550,000 |
| **s1**   |        512 |                     8,000,000 |
|          |        768 |                     5,000,000 |
|          |       1024 |                     4,000,000 |
|          |       1536 |                     2,500,000 |

Pinecone does not support fractional pod deployments, so always round up to the next nearest whole number when choosing your pods.

## Queries per second (QPS)

QPS speeds are governed by a combination of the [pod type](/guides/indexes/pods/understanding-pod-based-indexes#pod-types) of the index, the number of [replicas](/guides/indexes/pods/scale-pod-based-indexes#add-replicas), and the `top_k` value of queries. The pod type is the primary factor driving QPS, as the different pod types are optimized for different approaches.

The [p1 pods](/guides/index-data/indexing-overview/#p1-pods) are performance-optimized pods which provide very low query latencies, but hold fewer vectors per pod than [s1 pods](/guides/index-data/indexing-overview/#s1-pods). They are ideal for applications with low latency requirements (\<100ms). The s1 pods are optimized for storage and provide large storage capacity and lower overall costs with slightly higher query latencies than p1 pods. They are ideal for very large indexes with moderate or relaxed latency requirements.

The [p2 pod type](/guides/index-data/indexing-overview/#p2-pods) provides greater query throughput with lower latency. They support 200 QPS per replica and return queries in less than 10ms. This means that query throughput and latency are better than s1 and p1, especially for low dimension vectors (\<512D).

As a rule, a single p1 pod with 1M vectors of 768 dimensions each and no replicas can handle about 20 QPS. It’s possible to get greater or lesser speeds, depending on the size of your metadata, number of vectors, the dimensionality of your vectors, and the `top_K` value for your search. See Table 2 below for more examples.

**Table 2: QPS by pod type and `top_k` value**\*

| Pod type | top\_k 10 | top\_k 250 | top\_k 1000 |
| -------- | --------- | ---------- | ----------- |
| p1       | 30        | 25         | 20          |
| p2       | 150       | 50         | 20          |
| s1       | 10        | 10         | 10          |

\*The QPS values in Table 2 represent baseline QPS with 1M vectors and 768 dimensions.

[Adding replicas](/guides/indexes/pods/scale-pod-based-indexes#add-replicas) is the simplest way to increase your QPS. Each replica increases the throughput potential by roughly the same QPS, so aiming for 150 QPS using p1 pods means using the primary pod and 5 replicas. Using threading or multiprocessing in your application is also important, as issuing single queries sequentially still subjects you to delays from any underlying latency. The [Pinecone gRPC SDK](/guides/index-data/upsert-data#grpc-python-sdk) can also be used to increase throughput of upserts.

### Metadata cardinality and size

The last consideration when planning your indexes is the cardinality and size of your [metadata](/guides/index-data/upsert-data#inserting-vectors-with-metadata). While the increases are small when talking about a few million vectors, they can have a real impact as you grow to hundreds of millions or billions of vectors.

Indexes with very high cardinality, like those storing a unique user ID on each vector, can have significant memory requirements, resulting in fewer vectors fitting per pod. Also, if the size of the metadata per vector is larger, the index requires more storage. Limiting which metadata fields are indexed using [selective metadata indexing](/guides/indexes/pods/manage-pod-based-indexes#selective-metadata-indexing) can help lower memory usage.

### Pod sizes

You can also start with one of the larger [pod sizes](/guides/index-data/indexing-overview/#pod-size-and-performance), like p1.x2. Each step up in pod size doubles the space available for your vectors. We recommend starting with x1 pods and scaling as you grow. This way, you don’t start with too large a pod size and have nowhere else to go up, meaning you have to migrate to a new index before you’re ready.

### Example applications

The following examples will showcase how to use the sizing guidelines above to choose the appropriate type, size, and number of pods for your index.

#### Example 1: Semantic search of news articles

In our first example, we’ll use the demo app for semantic search from our documentation. In this case, we’re only working with 204,135 vectors. The vectors use 300 dimensions each, well under the general measure of 768 dimensions. Using the rule of thumb above of up to 1M vectors per p1 pod, we can run this app comfortably with a single p1.x1 pod.

#### Example 2: Facial recognition

For this example, suppose you’re building an application to identify customers using facial recognition for a secure banking app. Facial recognition can work with as few as 128 dimensions, but in this case, because the app will be used for access to finances, we want to make sure we’re certain that the person using it is the right one. We plan for 100M customers and use 2048 dimensions per vector.

We know from our rules of thumb above that 1M vectors with 768 dimensions fit nicely in a p1.x1 pod. We can just divide those numbers into the new targets to get the ratios we’ll need for our pod estimate:

```
100M / 1M = 100 base p1 pods
2048 / 768 = 2.667 vector ratio
2.667 * 100 = 267 rounding up
```

So we need 267 p1.x1 pods. We can reduce that by switching to s1 pods instead, sacrificing latency by increasing storage availability. They hold five times the storage of p1.x1, so the math is simple:

```
267 / 5 = 54 rounding up
```

So we estimate that we need 54 s1.x1 pods to store very high dimensional data for the face of each of the bank’s customers.

---

# Source: https://docs.pinecone.io/integrations/cloudera.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Cloudera AI

> Vector embedding, RAG, and semantic search at scale

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

[Cloudera AI](https://www.cloudera.com/) is an enterprise data cloud experience that provides scalable, secure, and agile machine learning and AI workflows. It leverages the power of Python, Apache Spark, R, and a host of other runtimes for distributed data processing, enabling the efficient creation, ingestion, and updating of vector embeddings at scale.

The primary advantage of Cloudera AI lies in its integration with the Cloudera ecosystem, which facilitates seamless data flow and processing across various stages of machine learning and AI pipelines. Cloudera AI offers interactive sessions, collaborative projects, model hosting capabilities, and application hosting features, all within a Python-centric development environment. This multifaceted approach enables users to efficiently develop, train, and deploy machine learning and AI models at scale.

Integrating Pinecone with Cloudera AI elevates the potential of Retrieval-Augmented Generation (RAG) models by providing a robust, scalable vector search platform. Pinecone's strength in handling vector embeddings — characterized by its ultra-low query latency, dynamic index updates, and scalability to billions of vector embeddings — make it the perfect match for the nuanced needs of RAG applications built on Cloudera AI.

Within the Cloudera AI ecosystem, Pinecone acts as a first-class citizen for RAG by efficiently retrieving relevant context from massive datasets, enhancing the generation capabilities of models with relevant, real-time data. This integration enables the development of sophisticated machine learning and AI applications that combine the predictive power of Cloudera AI's hosted models with the dynamic retrieval capabilities of Pinecone, offering unparalleled accuracy and relevance for generated outputs. By leveraging Cloudera AI's project and session management features, developers can prototype, develop, and deploy these complex systems more effectively, making advanced machine learning and AI applications more accessible and practical for enterprise use.

Cloudera's Accelerators for Machine Learning Projects (AMPs) drive efficient deployment of RAG architectures by doing the development work for you. This AMP serves as a prototype for fully integrating Pinecone into a RAG use case and illustrates semantic search with RAG at scale.

<PrimarySecondaryCTA primaryLabel="Get started" primaryHref="https://github.com/cloudera/CML_AMP_Intelligent-QA-Chatbot-with-NiFi-Pinecone-and-Llama2" />

## Additional resources

* [Python script](https://github.com/cloudera/CML_llm-hol/blob/main/2_populate_vector_db/pinecone_vectordb_insert.py) - Example of creating vectors in Pinecone
* [Jupyter notebook](https://github.com/cloudera/CML_llm-hol/blob/main/3_query_vector_db/pinecone_vectordb_query.ipynb) - Example of querying Pinecone collections

---

# Source: https://docs.pinecone.io/integrations/cohere.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Cohere

> Using Cohere and Pinecone to generate and index high-quality vector embeddings

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

The Cohere platform builds natural language processing and generation into your product with a few lines of code. Cohere's large language models (LLMs) can solve a broad spectrum of natural language use cases, including classification, semantic search, paraphrasing, summarization, and content generation.

Use the Cohere Embed API endpoint to generate language embeddings, and then index those embeddings in the Pinecone vector database for fast and scalable vector search.

<PrimarySecondaryCTA secondaryHref={"#setup-guide"} secondaryLabel={"View setup guide"} />

## Setup guide

[View source](https://github.com/pinecone-io/examples/blob/master/integrations/cohere/)

[Open in Colab](https://colab.research.google.com/github/pinecone-io/examples/blob/master/integrations/cohere/semantic%5Fsearch%5Ftrec.ipynb)

In this guide, you will learn how to use the [Cohere Embed API endpoint](https://docs.cohere.ai/reference/embed) to generate language embeddings, and then index those embeddings in the [Pinecone vector database](https://www.pinecone.io) for fast and scalable vector search.

This is a powerful and common combination for building semantic search, question-answering, threat-detection, and other applications that rely on NLP and search over a large corpus of text data.

The basic workflow looks like this:

* Embed and index
  * Use the Cohere Embed API endpoint to generate vector embeddings of your documents (or any text data).
  * Upload those vector embeddings into Pinecone, which can store and index millions/billions of these vector embeddings, and search through them at ultra-low latencies.
* Search
  * Pass your query text or document through the Cohere Embed API endpoint again.
  * Take the resulting vector embedding and send it as a [query](/guides/search/search-overview) to Pinecone.
  * Get back semantically similar documents, even if they don't share any keywords with the query.

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/fd0ba7b-pinecone-cohere-overview.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=60489f242107a8a6856e8a581fdb21a8" alt="Basic workflow of Cohere with Pinecone" data-og-width="1868" width="1868" data-og-height="820" height="820" data-path="images/fd0ba7b-pinecone-cohere-overview.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/fd0ba7b-pinecone-cohere-overview.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=7c70ecf1d04b4fb1b9744726db9169a4 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/fd0ba7b-pinecone-cohere-overview.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=211e2568585c6990a4084e5017709a97 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/fd0ba7b-pinecone-cohere-overview.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=77b3264b98809bfb1b29f88686cfd119 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/fd0ba7b-pinecone-cohere-overview.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=b9fb4b470b45f09ede57f352ad563872 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/fd0ba7b-pinecone-cohere-overview.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6a2ec80742d83e1b735340728c027772 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/fd0ba7b-pinecone-cohere-overview.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6f9ee94d7ff9d7fc78d4074d802b74f1 2500w" />

### Set up the environment

Start by installing the Cohere and Pinecone clients and HuggingFace *Datasets* for downloading the TREC dataset used in this guide:

```shell Shell theme={null}
pip install -U cohere pinecone datasets
```

### Create embeddings

Sign up for an API key at [Cohere](https://dashboard.cohere.com/api-keys) and then use it to initialize your connection.

```Python Python theme={null}
import cohere

co = cohere.Client("<YOUR_API_KEY>")
```

Load the **T**ext **RE**trieval **C**onference (TREC) question classification dataset, which contains 5.5K labeled questions. You will take only the first 1K samples for this walkthrough, but this can be scaled to millions or even billions of samples.

```Python Python theme={null}
from datasets import load_dataset

# load the first 1K rows of the TREC dataset
trec = load_dataset('trec', split='train[:1000]')
```

Each sample in `trec` contains two label features and the *text* feature. Pass the questions from the *text* feature to Cohere to create embeddings.

```Python Python theme={null}
embeds = co.embed(
    texts=trec['text'],
    model='embed-english-v3.0',
    input_type='search_document',
    truncate='END'
).embeddings
```

Check the dimensionality of the returned vectors. You will need to save the embedding dimensionality from this to be used when initializing your Pinecone index later

```Python Python theme={null}
import numpy as np

shape = np.array(embeds).shape
print(shape)

# [Out]:
# (1000, 1024)
```

You can see the `1024` embedding dimensionality produced by Cohere's `embed-english-v3.0` model, and the `1000` samples you built embeddings for.

### Store the Embeddings

Now that you have your embeddings, you can move on to indexing them in the Pinecone vector database. For this, you need a [Pinecone API key](/guides/projects/manage-api-keys).

You first initialize our connection to Pinecone and then create a new index called `cohere-pinecone-trec` for storing the embeddings. When creating the index, you specify that you would like to use the cosine similarity metric to align with Cohere's embeddings, and also pass the embedding dimensionality of `1024`.

```Python Python theme={null}
from pinecone import Pinecone

# initialize connection to pinecone (get API key at app.pinecone.io)
pc = Pinecone(api_key='YOUR_API_KEY')

index_name = 'cohere-pinecone-trec'

# if the index does not exist, we create it
if not pc.has_index(index_name):
    pc.create_index(
        name=index_name,
        dimension=shape[1],
        metric="cosine",
        spec=ServerlessSpec(
            cloud='aws', 
            region='us-east-1'
        ) 
    )

# connect to index
index = pc.Index(index_name)
```

Now you can begin populating the index with your embeddings. Pinecone expects you to provide a list of tuples in the format *(id, vector, metadata)*, where the *metadata* field is an optional extra field where you can store anything you want in a dictionary format. For this example, you will store the original text of the embeddings.

While uploading your data, you will batch everything to avoid pushing too much data in one go.

```Python Python theme={null}
batch_size = 128

ids = [str(i) for i in range(shape[0])]
# create list of metadata dictionaries
meta = [{'text': text} for text in trec['text']]

# create list of (id, vector, metadata) tuples to be upserted
to_upsert = list(zip(ids, embeds, meta))

for i in range(0, shape[0], batch_size):
    i_end = min(i+batch_size, shape[0])
    index.upsert(vectors=to_upsert[i:i_end])

# let's view the index statistics
print(index.describe_index_stats())

# [Out]:
# {'dimension': 1024,
#  'index_fullness': 0.0,
#  'namespaces': {'': {'vector_count': 1000}},
#  'total_vector_count': 1000}
```

You can see from `index.describe_index_stats` that you have a *1024-dimensionality* index populated with *1000* embeddings. Note that serverless indexes scale automatically as needed, so the `index_fullness` metric is relevant only for pod-based indexes.

### Semantic search

Now that you have your indexed vectors, you can perform a few search queries. When searching, you will first embed your query using Cohere, and then search using the returned vector in Pinecone.

```Python Python theme={null}
query = "What caused the 1929 Great Depression?"

# create the query embedding
xq = co.embed(
    texts=[query],
    model='embed-english-v3.0',
    input_type='search_query',
    truncate='END'
).embeddings

print(np.array(xq).shape)

# query, returning the top 5 most similar results
res = index.query(vector=xq, top_k=5, include_metadata=True)
```

The response from Pinecone includes your original text in the `metadata` field. Let's print out the `top_k` most similar questions and their respective similarity scores.

```Python Python theme={null}
for match in res['matches']:
    print(f"{match['score']:.2f}: {match['metadata']['text']}")

# [Out]:
# 0.62: Why did the world enter a global depression in 1929 ?
# 0.49: When was `` the Great Depression '' ?
# 0.38: What crop failure caused the Irish Famine ?
# 0.32: What caused Harry Houdini 's death ?
# 0.31: What causes pneumonia ?
```

Looks good! Let's make it harder and replace *"depression"* with the incorrect term *"recession"*.

```Python Python theme={null}
query = "What was the cause of the major recession in the early 20th century?"

# create the query embedding
xq = co.embed(
    texts=[query],
    model='embed-english-v3.0',
    input_type='search_query',
    truncate='END'
).embeddings

# query, returning the top 5 most similar results
res = index.query(vector=xq, top_k=5, include_metadata=True)

for match in res['matches']:
    print(f"{match['score']:.2f}: {match['metadata']['text']}")

# [Out]:
# 0.43: When was `` the Great Depression '' ?
# 0.40: Why did the world enter a global depression in 1929 ?
# 0.39: When did World War I start ?
# 0.35: What are some of the significant historical events of the 1990s ?
# 0.32: What crop failure caused the Irish Famine ?
```

Let's perform one final search using the definition of depression rather than the word or related words.

```Python Python theme={null}
query = "Why was there a long-term economic downturn in the early 20th century?"

# create the query embedding
xq = co.embed(
    texts=[query],
    model='embed-english-v3.0',
    input_type='search_query',
    truncate='END'
).embeddings

# query, returning the top 10 most similar results
res = index.query(vector=xq, top_k=10, include_metadata=True)

for match in res['matches']:
    print(f"{match['score']:.2f}: {match['metadata']['text']}")

# [Out]:
# 0.40: When was `` the Great Depression '' ?
# 0.39: Why did the world enter a global depression in 1929 ?
# 0.35: When did World War I start ?
# 0.32: What are some of the significant historical events of the 1990s ?
# 0.31: What war did the Wanna-Go-Home Riots occur after ?
# 0.31: What do economists do ?
# 0.29: What historical event happened in Dogtown in 1899 ?
# 0.28: When did the Dow first reach ?
# 0.28: Who earns their money the hard way ?
# 0.28: What were popular songs and types of songs in the 1920s ?
```

It's clear from this example that the semantic search pipeline is clearly able to identify the meaning between each of your queries. Using these embeddings with Pinecone allows you to return the most semantically similar questions from the already indexed TREC dataset.

---

# Source: https://docs.pinecone.io/reference/cli/command-reference.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# CLI command reference

<Note>
  This feature is in [public preview](/release-notes/feature-availability).
</Note>

This document provides a complete reference for all Pinecone CLI commands.

## Command structure

The Pinecone CLI uses a hierarchical command structure. Each command consists of a primary command followed by one or more subcommands and optional flags.

```bash  theme={null}
pc <command> <subcommand> [flags]
pc <command> <subcommand> <subcommand> [flags]
```

For example:

```bash  theme={null}
# Top-level command with flags
pc target -o "organization-name" -p "project-name"

# Command (index) and subcommand (list)
pc index list

# Command (index) and subcommand (create) with flags
pc index create \
  --name my-index \
  --dimension 1536 \
  --metric cosine \
  --cloud aws \
  --region us-east-1

# Command (auth) and nested subcommands (local-keys prune) with flags
pc auth local-keys prune --id proj-abc123 --skip-confirmation
```

## Getting help

The CLI provides help for commands at every level:

```bash  theme={null}
# top-level help
pc --help
pc -h

# command help
pc auth --help
pc index --help
pc project --help

# subcommmand help
pc index create --help
pc project create --help
pc auth configure --help

# nested subcommand help
pc auth local-keys prune --help
```

## Exit codes

All commands return exit code `0` for success and `1` for error.

## Available commands

This section describes all commands offered by the Pinecone CLI.

### Top-level commands

<AccordionGroup>
  <Accordion title="pc login (Authenticate via user login)">
    **Description**

    Authenticate via a web browser. After login, set a [target org and project](/reference/cli/target-context) with `pc target` before accessing data. This command defaults to an initial organization and project to which
    you have access (these values display in the terminal), but you can change them with `pc target`.

    **Usage**

    ```bash  theme={null}
    pc login
    ```

    **Flags**

    None

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Log in via browser
    pc login

    # Then set target context
    pc target -o "my-org" -p "my-project"
    ```

    <Note>
      This is an alias for `pc auth login`. Both commands perform the same operation.
    </Note>
  </Accordion>

  <Accordion title="pc logout (Clear credentials / log out)">
    **Description**

    Clears all authentication data from local storage, including:

    * User login token
    * Service account credentials (client ID and secret)
    * Default (manually specified) API key
    * Locally managed keys (for all projects)
    * Target organization and project context

    **Usage**

    ```bash  theme={null}
    pc logout
    ```

    **Flags**

    None

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Clear all credentials and context
    pc logout
    ```

    <Note>
      This is an alias for `pc auth logout`. Both commands perform the same operation. Does not delete managed API keys from Pinecone's servers. Run `pc auth local-keys prune` before logging out to fully clean up.
    </Note>
  </Accordion>

  <Accordion title="pc target (Set target organization and project)">
    **Description**

    Set the target organization and project for the CLI. Supports interactive organization and project selection or direct specification via flags.  For details, see [CLI target context](/reference/cli/target-context).

    **Usage**

    ```bash  theme={null}
    pc target [flags]
    ```

    **Flags**

    | Long flag           | Short flag | Description                    |
    | :------------------ | :--------- | :----------------------------- |
    | `--clear`           |            | Clear target context           |
    | `--json`            |            | Output in JSON format          |
    | `--org`             | `-o`       | Organization name              |
    | `--organization-id` |            | Organization ID                |
    | `--project`         | `-p`       | Project name                   |
    | `--project-id`      |            | Project ID                     |
    | `--show`            | `-s`       | Display current target context |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Interactive targeting after login
    pc login
    pc target

    # Set specific organization and project
    pc target -o "my-org" -p "my-project"

    # Show current context
    pc target --show

    # Clear all context
    pc target --clear
    ```
  </Accordion>

  <Accordion title="pc version (Show CLI version)">
    **Description**

    Displays version information for the CLI, including the version number, commit SHA, and build date.

    **Usage**

    ```bash  theme={null}
    pc version
    ```

    **Flags**

    None

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Display version information
    pc version
    ```
  </Accordion>

  <Accordion title="pc whoami (Show current user)">
    **Description**

    Displays information about the currently authenticated user. To use this command, you must be authenticated via user login.

    **Usage**

    ```bash  theme={null}
    pc whoami
    ```

    **Flags**

    None

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    pc whoami
    ```

    <Note>
      This is an alias for `pc auth whoami`. Both commands perform the same operation.
    </Note>
  </Accordion>
</AccordionGroup>

### Authentication

<AccordionGroup>
  <Accordion title="pc auth clear (Clear specific credentials)">
    **Description**

    Selectively clears specific authentication data without affecting other credentials. At least one flag is required.

    **Usage**

    ```bash  theme={null}
    pc auth clear [flags]
    ```

    **Flags**

    | Long flag           | Short flag | Description                                         |
    | :------------------ | :--------- | :-------------------------------------------------- |
    | `--api-key`         |            | Clear only the default (manually specified) API key |
    | `--service-account` |            | Clear only service account credentials              |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Clear only the default (manually specified) API key
    pc auth clear --api-key
    pc auth status

    # Clear service account
    pc auth clear --service-account
    ```

    <Note>
      More surgical than `pc auth logout`. Does not clear user login token or managed keys. For those, use `pc auth logout` or `pc auth local-keys prune`.
    </Note>
  </Accordion>

  <Accordion title="pc auth configure (Configure service account or API key)">
    **Description**

    Configures service account credentials or a default (manually specified) API key.

    Service accounts automatically target the organization and prompt for project selection, unless there is only one project. A default API key overrides any previously specified target organization/project context. When setting a service account, this operation clears the user login token, if one exists.
    For details, see [CLI target context](/reference/cli/target-context).

    **Usage**

    ```bash  theme={null}
    pc auth configure [flags]
    ```

    **Flags**

    | Long flag               | Short flag | Description                                          |
    | :---------------------- | :--------- | :--------------------------------------------------- |
    | `--api-key`             |            | Default API key to use for authentication            |
    | `--client-id`           |            | Service account client ID                            |
    | `--client-secret`       |            | Service account client secret                        |
    | `--client-secret-stdin` |            | Read client secret from stdin                        |
    | `--json`                |            | Output in JSON format                                |
    | `--project-id`          | `-p`       | Target project ID (optional, interactive if omitted) |
    | `--prompt-if-missing`   |            | Prompt for missing credentials                       |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Service account setup (auto-targets org and prompts for project)
    pc auth configure --client-id my-id --client-secret my-secret

    # Service account with specific project
    pc auth configure \
      --client-id my-id \
      --client-secret my-secret \
      -p proj-123

    # Default API key (overrides any target context)
    pc auth configure --api-key pcsk_abc123
    ```

    <Note>
      `pc auth configure --api-key "YOUR_API_KEY"` does the same thing as `pc config set-api-key "YOUR_API_KEY"`. To learn about targeting a project after authenticating with a service account, see [CLI target context](/reference/cli/target-context).
    </Note>
  </Accordion>

  <Accordion title="pc auth local-keys list (List managed keys)">
    **Description**

    Displays all [managed API keys](/reference/cli/authentication#managed-keys) stored locally by the CLI, with various details.

    **Usage**

    ```bash  theme={null}
    pc auth local-keys list [flags]
    ```

    **Flags**

    | Long flag  | Short flag | Description                                |
    | :--------- | :--------- | :----------------------------------------- |
    | `--json`   |            | Output in JSON format                      |
    | `--reveal` |            | Show the actual API key values (sensitive) |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # List all locally managed keys
    pc auth local-keys list

    # Show key values
    pc auth local-keys list --reveal

    # After storing a key
    pc api-key create -n "my-key" --store
    pc auth local-keys list
    ```
  </Accordion>

  <Accordion title="pc auth local-keys prune (Delete managed keys)">
    **Description**

    Deletes locally stored [managed API keys](/reference/cli/authentication#managed-keys) from local storage and Pinecone's servers. Filters by origin (`cli`/`user`/`all`) or project ID.

    **Usage**

    ```bash  theme={null}
    pc auth local-keys prune [flags]
    ```

    **Flags**

    | Long flag             | Short flag | Description                                                 |
    | :-------------------- | :--------- | :---------------------------------------------------------- |
    | `--dry-run`           |            | Preview deletions without applying                          |
    | `--id`                |            | Prune keys for specific project ID only                     |
    | `--json`              |            | Output in JSON format                                       |
    | `--origin`            | `-o`       | Filter by origin - `cli`, `user`, or `all` (default: `all`) |
    | `--skip-confirmation` |            | Skip confirmation prompt                                    |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Preview deletions
    pc auth local-keys prune --dry-run

    # Delete CLI-created keys only
    pc auth local-keys prune -o cli --skip-confirmation

    # Delete for specific project
    pc auth local-keys prune --id proj-abc123

    # Before/after check
    pc auth local-keys list
    pc auth local-keys prune -o cli
    pc auth local-keys list
    ```

    <Note>
      This deletes keys from both local storage and Pinecone servers. Use `--dry-run` to preview before committing.
    </Note>
  </Accordion>

  <Accordion title="pc auth login (Authenticate via user login)">
    **Description**

    Authenticate via user login in the web browser. After login, [set a target org and project](/reference/cli/target-context).

    **Usage**

    ```bash  theme={null}
    pc auth login
    pc login  # shorthand
    ```

    **Flags**

    None

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Login and set target
    pc auth login
    pc target -o "my-org" -p "my-project"
    pc index list
    ```

    <Note>
      Tokens expire after 30 minutes, but they refresh automatically for up to 24 hours. After that point, you must re-authenticate. Logging in clears any existing service account credentials. This command does the same thing as `pc login`.
    </Note>
  </Accordion>

  <Accordion title="pc auth logout (Clear authentication credentials)">
    **Description**

    Clears all authentication data from local storage, including:

    * User login token
    * Service account credentials (client ID and secret)
    * Default (manually specified) API key
    * Locally managed keys (for all projects)
    * Target organization and project context

    **Usage**

    ```bash  theme={null}
    pc auth logout
    ```

    **Flags**

    None

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Clear all credentials and context
    pc auth logout
    ```

    <Note>
      This command does the same thing as `pc logout`. Does not delete managed API keys from Pinecone's servers. Run `pc auth local-keys prune` before logging out to fully clean up.
    </Note>
  </Accordion>

  <Accordion title="pc auth status (Show authentication status)">
    **Description**

    Shows details about all configured authentication methods.

    **Usage**

    ```bash  theme={null}
    pc auth status [flags]
    ```

    **Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--json`  |            | Output in JSON format |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Check status after login
    pc auth login
    pc auth status

    # JSON output for scripting
    pc auth status --json
    ```
  </Accordion>

  <Accordion title="pc auth whoami (Show current user information)">
    **Description**

    Displays information about the currently authenticated user. To use this command, you must be authenticated via user login.

    **Usage**

    ```bash  theme={null}
    pc auth whoami
    ```

    **Flags**

    None

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    pc auth whoami
    ```

    <Note>
      This command does the same thing as `pc whoami`.
    </Note>
  </Accordion>
</AccordionGroup>

### Indexes

<AccordionGroup>
  <Accordion title="pc index configure (Update index configuration)">
    **Description**

    Modifies the configuration of an existing index.

    **Usage**

    ```bash  theme={null}
    pc index configure [flags]
    ```

    **Flags**

    | Long flag               | Short flag | Description                                                    |
    | :---------------------- | :--------- | :------------------------------------------------------------- |
    | `--deletion_protection` | `-p`       | Enable or disable deletion protection -`enabled` or `disabled` |
    | `--json`                |            | Output in JSON format                                          |
    | `--name`                | `-n`       | Index name (required)                                          |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Enable deletion protection
    pc index configure -n my-index -p enabled

    # Verify changes
    pc index describe -n my-index
    ```

    <Note>
      Configuration changes may take some time to take effect.
    </Note>
  </Accordion>

  <Accordion title="pc index create (Create a new index)">
    **Description**

    Creates a new serverless index in your Pinecone project, with various configuration optons.

    **Usage**

    ```bash  theme={null}
    pc index create [flags]
    ```

    **Flags**

    | Long flag               | Short flag | Description                                                               |
    | :---------------------- | :--------- | :------------------------------------------------------------------------ |
    | `--cloud`               | `-c`       | Cloud provider - `aws`, `gcp`, or `azure` (required for serverless)       |
    | `--deletion_protection` |            | Deletion protection - `enabled` or `disabled`                             |
    | `--dimension`           | `-d`       | Vector dimension (required for standard indexes, optional for integrated) |
    | `--field_map`           |            | Field mapping for integrated embedding (JSON map)                         |
    | `--json`                |            | Output in JSON format                                                     |
    | `--metric`              | `-m`       | Similarity metric - `cosine`, `euclidean`, or `dotproduct` (required)     |
    | `--model`               |            | Integrated embedding model name                                           |
    | `--name`                | `-n`       | Index name (required)                                                     |
    | `--read_parameters`     |            | Read parameters for embedding model (JSON map)                            |
    | `--region`              | `-r`       | Cloud region                                                              |
    | `--source_collection`   |            | Name of the source collection from which to create the index              |
    | `--tags`                |            | Custom user tags (key=value pairs)                                        |
    | `--vector_type`         | `-v`       | Vector type - `dense` or `sparse` (serverless only)                       |
    | `--write_parameters`    |            | Write parameters for embedding model (JSON map)                           |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Create serverless index
    pc index create -n my-index -d 1536 -m cosine -c aws -r us-east-1

    # With integrated embedding model
    pc index create \
      -n my-index \
      -m cosine \
      -c aws \
      -r us-east-1 \
      --model multilingual-e5-large

    # With deletion protection
    pc index create \
      -n my-index \
      -d 1536 \
      -m cosine \
      -c aws \
      -r us-west-2 \
      --deletion_protection enabled

    # From collection
    pc index create \
      -n my-index \
      -d 1536 \
      -m cosine \
      -c aws \
      -r eu-west-1 \
      --source_collection my-collection
    ```

    <Note>
      For a list of valid regions for a serverless index, see [Create a serverless index](/guides/index-data/create-an-index).
    </Note>
  </Accordion>

  <Accordion title="pc index delete (Delete an index)">
    **Description**

    Permanently deletes an index and all its data. This operation cannot be undone.

    **Usage**

    ```bash  theme={null}
    pc index delete [flags]
    ```

    **Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--name`  | `-n`       | Index name (required) |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Delete an index
    pc index delete -n my-index

    # List before and after
    pc index list
    pc index delete -n test-index
    pc index list
    ```
  </Accordion>

  <Accordion title="pc index describe (Show index details)">
    **Description**

    Displays detailed configuration and status information for a specific index.

    **Usage**

    ```bash  theme={null}
    pc index describe [flags]
    ```

    **Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--json`  |            | Output in JSON format |
    | `--name`  | `-n`       | Index name (required) |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Describe an index
    pc index describe -n my-index

    # JSON output
    pc index describe -n my-index --json

    # Check newly created index
    pc index create -n test-index -d 1536 -m cosine -c aws -r us-east-1
    pc index describe -n test-index
    ```
  </Accordion>

  <Accordion title="pc index list (List all indexes)">
    **Description**

    Displays all indexes in your current target project, including various details.

    **Usage**

    ```bash  theme={null}
    pc index list [flags]
    ```

    **Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--json`  |            | Output in JSON format |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # List all indexes
    pc index list

    # JSON output for scripting
    pc index list --json

    # After creating indexes
    pc index create -n test-1 -d 768 -m cosine -c aws -r us-east-1
    pc index list
    ```
  </Accordion>
</AccordionGroup>

### Projects

<AccordionGroup>
  <Accordion title="pc project create (Create a new project)">
    **Description**

    Creates a new project in your [target organization](/reference/cli/target-context), using the specified configuration.

    **Usage**

    ```bash  theme={null}
    pc project create [flags]
    ```

    **Flags**

    | Long flag            | Short flag | Description                                                    |
    | :------------------- | :--------- | :------------------------------------------------------------- |
    | `--force-encryption` |            | Enable encryption with CMEK                                    |
    | `--json`             |            | Output in JSON format                                          |
    | `--name`             | `-n`       | Project name (required)                                        |
    | `--target`           |            | Automatically target the project in the CLI after it's created |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Basic project creation
    pc project create -n "demo-project"
    ```
  </Accordion>

  <Accordion title="pc project delete (Delete a project)">
    **Description**

    Permanently deletes a project and all its resources. This operation cannot be undone.

    **Usage**

    ```bash  theme={null}
    pc project delete [flags]
    ```

    **Flags**

    | Long flag             | Short flag | Description                                                 |
    | :-------------------- | :--------- | :---------------------------------------------------------- |
    | `--id`                | `-i`       | Project ID (optional, uses target project if not specified) |
    | `--json`              |            | Output in JSON format                                       |
    | `--skip-confirmation` |            | Skip confirmation prompt                                    |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Delete target project
    pc project delete

    # Delete specific project
    pc project delete -i proj-abc123

    # Skip confirmation
    pc project delete -i proj-abc123 --skip-confirmation
    ```

    <Note>
      Must delete all indexes and collections in the project first. If the deleted project is your current target, set a new target after deleting it.
    </Note>
  </Accordion>

  <Accordion title="pc project describe (Show project details)">
    **Description**

    Displays detailed information about a specific project, including various details.

    **Usage**

    ```bash  theme={null}
    pc project describe [flags]
    ```

    **Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--id`    | `-i`       | Project ID (required) |
    | `--json`  |            | Output in JSON format |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Describe a project
    pc project describe -i proj-abc123

    # JSON output
    pc project describe -i proj-abc123 --json

    # Find ID and describe
    pc project list
    pc project describe -i proj-abc123
    ```
  </Accordion>

  <Accordion title="pc project list (List all projects)">
    **Description**

    Displays all projects in your [target organization](/reference/cli/target-context), including various details.

    **Usage**

    ```bash  theme={null}
    pc project list [flags]
    ```

    **Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--json`  |            | Output in JSON format |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # List all projects
    pc project list

    # JSON output
    pc project list --json

    # List after login
    pc auth login
    pc auth target -o "my-org"
    pc project list
    ```
  </Accordion>

  <Accordion title="pc project update (Update project settings)">
    **Description**

    Modifies the configuration of the [target project](/reference/cli/target-context), or a specific project ID.

    **Usage**

    ```bash  theme={null}
    pc project update [flags]
    ```

    **Flags**

    | Long flag            | Short flag | Description                         |
    | :------------------- | :--------- | :---------------------------------- |
    | `--force-encryption` | `-f`       | Enable/disable encryption with CMEK |
    | `--id`               | `-i`       | Project ID (required)               |
    | `--json`             |            | Output in JSON format               |
    | `--name`             | `-n`       | New project name                    |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Update name
    pc project update -i proj-abc123 -n "new-name"
    ```
  </Accordion>
</AccordionGroup>

### Organizations

<AccordionGroup>
  <Accordion title="pc organization delete (Delete an organization)">
    **Description**

    Permanently deletes an organization and all its resources. This operation cannot be undone.

    **Usage**

    ```bash  theme={null}
    pc organization delete [flags]
    ```

    **Flags**

    | Long flag             | Short flag | Description                |
    | :-------------------- | :--------- | :------------------------- |
    | `--id`                | `-i`       | Organization ID (required) |
    | `--json`              |            | Output in JSON format      |
    | `--skip-confirmation` |            | Skip confirmation prompt   |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Delete an organization
    pc organization delete -i org-abc123

    # Skip confirmation
    pc organization delete -i org-abc123 --skip-confirmation
    ```

    <Note>
      This is a highly destructive action. Deletion is permanent. If the deleted organization is your current [target](/reference/cli/target-context), set a new target after deleting.
    </Note>
  </Accordion>

  <Accordion title="pc organization describe (Show organization details)">
    **Description**

    Displays detailed information about a specific organization, including name, ID, creation date, payment status, plan, and support tier.

    **Usage**

    ```bash  theme={null}
    pc organization describe [flags]
    ```

    **Flags**

    | Long flag | Short flag | Description                |
    | :-------- | :--------- | :------------------------- |
    | `--id`    | `-i`       | Organization ID (required) |
    | `--json`  |            | Output in JSON format      |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Describe an organization
    pc organization describe -i org-abc123

    # JSON output
    pc organization describe -i org-abc123 --json

    # Find ID and describe
    pc organization list
    pc organization describe -i org-abc123
    ```
  </Accordion>

  <Accordion title="pc organization list (List organizations)">
    **Description**

    Displays all organizations that the authenticated user has access to, including name, ID, creation date, payment status, plan, and support tier.

    **Usage**

    ```bash  theme={null}
    pc organization list [flags]
    ```

    **Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--json`  |            | Output in JSON format |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # List all organizations
    pc organization list

    # JSON output
    pc organization list --json

    # List after login
    pc auth login
    pc organization list
    ```
  </Accordion>

  <Accordion title="pc organization update (Update organization settings)">
    **Description**

    Modifies the configuration of the [target organization](/reference/cli/target-context), or a specific organization ID.

    **Usage**

    ```bash  theme={null}
    pc organization update [flags]
    ```

    **Flags**

    | Long flag | Short flag | Description                |
    | :-------- | :--------- | :------------------------- |
    | `--id`    | `-i`       | Organization ID (required) |
    | `--json`  |            | Output in JSON format      |
    | `--name`  | `-n`       | New organization name      |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Update name
    pc organization update -i org-abc123 -n "new-name"

    # Verify changes
    pc organization update -i org-abc123 -n "Acme Corp"
    pc organization describe -i org-abc123
    ```
  </Accordion>
</AccordionGroup>

### API keys

<AccordionGroup>
  <Accordion title="pc api-key create (Create a new API key)">
    **Description**

    Creates a new API key for the current [target project](/reference/cli/target-context) or a specific project ID. Optionally stores the key locally for CLI use.

    **Usage**

    ```bash  theme={null}
    pc api-key create [flags]
    ```

    **Flags**

    | Long flag | Short flag | Description                                                                             |
    | :-------- | :--------- | :-------------------------------------------------------------------------------------- |
    | `--id`    | `-i`       | Project ID (optional, uses target project if not specified)                             |
    | `--json`  |            | Output in JSON format                                                                   |
    | `--name`  | `-n`       | Key name (required)                                                                     |
    | `--roles` |            | Roles to assign (default: `ProjectEditor`)                                              |
    | `--store` |            | Store the key locally for CLI use (automatically replaces any existing CLI-managed key) |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Basic key creation
    pc api-key create -n "my-key"

    # Create and store locally
    pc api-key create -n "my-key" --store

    # Create with specific role
    pc api-key create -n "my-key" --store --roles ProjectEditor

    # Create for specific project
    pc api-key create -n "my-key" -i proj-abc123
    ```

    <Note>
      API keys are scoped to a specific organization and project.
    </Note>
  </Accordion>

  <Accordion title="pc api-key delete (Delete an API key)">
    **Description**

    Permanently deletes an API key. Applications using this key immediately lose access.

    **Usage**

    ```bash  theme={null}
    pc api-key delete [flags]
    ```

    **Flags**

    | Long flag             | Short flag | Description              |
    | :-------------------- | :--------- | :----------------------- |
    | `--id`                | `-i`       | API key ID (required)    |
    | `--skip-confirmation` |            | Skip confirmation prompt |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Delete an API key
    pc api-key delete -i key-abc123

    # Skip confirmation
    pc api-key delete -i key-abc123 --skip-confirmation

    # Delete and clean up local storage
    pc api-key delete -i key-abc123
    pc auth local-keys prune --skip-confirmation
    ```

    <Note>
      Deletion is permanent. Applications using this key immediately lose access to Pinecone.
    </Note>
  </Accordion>

  <Accordion title="pc api-key describe (Show API key details)">
    **Description**

    Displays detailed information about a specific API key, including its name, ID, project ID, and roles.

    **Usage**

    ```bash  theme={null}
    pc api-key describe [flags]
    ```

    **Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--id`    | `-i`       | API key ID (required) |
    | `--json`  |            | Output in JSON format |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Describe an API key
    pc api-key describe -i key-abc123

    # JSON output
    pc api-key describe -i key-abc123 --json

    # Find ID and describe
    pc api-key list
    pc api-key describe -i key-abc123
    ```

    <Note>
      Does not display the actual key value.
    </Note>
  </Accordion>

  <Accordion title="pc api-key list (List all API keys)">
    **Description**

    Displays a list of all of the [target project's](/reference/cli/target-context) API keys, as found in  Pinecone (regardless of whether they are stored locally by the CLI). Displays various details about each key, including name, ID, project ID, and roles.

    **Usage**

    ```bash  theme={null}
    pc api-key list [flags]
    ```

    **Flags**

    | Long flag | Short flag | Description                                                 |
    | :-------- | :--------- | :---------------------------------------------------------- |
    | `--id`    | `-i`       | Project ID (optional, uses target project if not specified) |
    | `--json`  |            | Output in JSON format                                       |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # List keys for target project
    pc api-key list

    # List for specific project
    pc api-key list -i proj-abc123

    # JSON output
    pc api-key list --json
    ```

    <Note>
      Does not display key values.
    </Note>
  </Accordion>

  <Accordion title="pc api-key update (Update API key settings)">
    **Description**

    Updates the name and roles of an API key.

    **Usage**

    ```bash  theme={null}
    pc api-key update [flags]
    ```

    **Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--id`    | `-i`       | API key ID (required) |
    | `--json`  | `-j`       | Output in JSON format |
    | `--name`  | `-n`       | New key name          |
    | `--roles` | `-r`       | Roles to assign       |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Update name
    pc api-key update -i key-abc123 -n "new-name"

    # Update roles
    pc api-key update -i key-abc123 -r ProjectEditor

    # Verify changes
    pc api-key update -i key-abc123 -n "production-key"
    pc api-key describe -i key-abc123
    ```

    <Note>
      Cannot change the actual key. If you need a different key, create a new one.
    </Note>
  </Accordion>
</AccordionGroup>

### Config

<AccordionGroup>
  <Accordion title="pc config get-api-key (Show configured API key)">
    **Description**

    Displays the currently configured default (manually specified) API key, if set. By default, the full value of the key is not displayed.

    **Usage**

    ```bash  theme={null}
    pc config get-api-key
    ```

    **Flags**

    | Long flag  | Short flag | Description                               |
    | :--------- | :--------- | :---------------------------------------- |
    | `--reveal` |            | Show the actual API key value (sensitive) |

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Get current API key
    pc config get-api-key

    # Verify after setting
    pc config set-api-key pcsk_abc123
    pc config get-api-key
    ```
  </Accordion>

  <Accordion title="pc config set-api-key (Set default API key)">
    **Description**

    Sets a default API key for the CLI to use for authentication. Provides direct access to control plane operations, but not Admin API operations.

    **Usage**

    ```bash  theme={null}
    pc config set-api-key "YOUR_API_KEY"
    ```

    **Flags**

    None (takes API key as argument)

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Set default API key
    pc config set-api-key pcsk_abc123

    # Use immediately without targeting
    pc config set-api-key pcsk_abc123
    pc index list

    # Verify it's set
    pc auth status
    ```

    <Note>
      `pc config set-api-key "YOUR_API_KEY"` does the same thing as `pc auth configure --api-key "YOUR_API_KEY"`. For control plane operations, a default API key implicitly overrides any previously set [target context](/reference/cli/target-context), because Pinecone API keys are scoped to a specific organization and project.
    </Note>
  </Accordion>

  <Accordion title="pc config set-color (Enable/disable colored output)">
    **Description**

    Enables or disables colored output in CLI responses. Useful for terminal compatibility or log file generation.

    **Usage**

    ```bash  theme={null}
    pc config set-color true
    pc config set-color false
    ```

    **Flags**

    None (takes boolean as argument)

    **Global Flags**

    | Long flag | Short flag | Description           |
    | :-------- | :--------- | :-------------------- |
    | `--help`  | `-h`       | Show help information |
    | `--quiet` | `-q`       | Suppress output       |

    **Example**

    ```bash  theme={null}
    # Enable colored output
    pc config set-color true

    # Disable colored output for CI/CD
    pc config set-color false

    # Test the change
    pc config set-color false
    pc index list
    ```
  </Accordion>
</AccordionGroup>

---

# Source: https://docs.pinecone.io/guides/get-started/concepts.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Concepts

> Understand concepts in Pinecone and how they relate to each other.

<img className="block max-w-full dark:hidden" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=6a1fed2139efe425406492dd141479e5" data-og-width="1560" width="1560" data-og-height="1080" height="1080" data-path="images/objects.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=15a583143f38fe67a9779528372965ea 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=b03b7fe3958699814d8f7600ccb35e31 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=872bdd974921197be14dea5ea341a647 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=96031734ef609a1bf4ce017d47a47fd0 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=73fcff368598a3dba7d66b003fbc1fa9 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=5c9e9fd81f32f83f85d79469de2f95e9 2500w" />

<img className="hidden max-w-full dark:block" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects-dark.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=361fa6648a258883c2265e90a22465bb" data-og-width="1560" width="1560" data-og-height="1080" height="1080" data-path="images/objects-dark.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects-dark.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=49658c275fcaddc3171e318819b5de5a 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects-dark.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=41c0c829761a10e98db5e190abe1c4a8 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects-dark.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=38f01332bd715a340fbcbad69834dccd 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects-dark.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=a879c00cc550e8816d7052c347331086 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects-dark.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=bd8b507daf829f6ff0ba2a75658404fe 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/objects-dark.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=f1b320141a2668e5f32cc2566aa6820c 2500w" />

## Organization

An organization is a group of one or more [projects](#project) that use the same billing. Organizations allow one or more [users](#user) to control billing and permissions for all of the projects belonging to the organization.

For more information, see [Understanding organizations](/guides/organizations/understanding-organizations).

## Project

A project belongs to an [organization](#organization) and contains one or more [indexes](#index). Each project belongs to exactly one organization, but only [users](#user) who belong to the project can access the indexes in that project. [API keys](#api-key) and [Assistants](#assistant) are project-specific.

For more information, see [Understanding projects](/guides/projects/understanding-projects).

## Index

There are two types of [serverless indexes](/guides/index-data/indexing-overview), dense and sparse.

### Dense index

Dense indexes store records that have one [dense vector](#dense-vector) each.

<Note>
  If any records in a dense index also have a [sparse vector](#sparse-vector), the index is a [hybrid index](/guides/search/hybrid-search#use-a-single-hybrid-index).
</Note>

A dense vector is a series of numbers that represent the meaning and relationships of text, images, or other types of data. Each number in a dense vector corresponds to a point in a multidimensional space. Vectors that are closer together in that space are semantically similar.

When you query a dense index, Pinecone retrieves records containing dense vectors that are most semantically similar to the query. This is often called **semantic search**, nearest neighbor search, similarity search, or just vector search.

### Sparse index

Sparse indexes store records that have one [sparse vector](#sparse-vector) each.

Every sparse vector is a series of numbers that represent the words or phrases in a document. Sparse vectors have a very large number of dimensions, where only a small proportion of values are non-zero. The dimensions represent words from a dictionary, and the values represent the importance of these words in the document.

When you search a sparse index, Pinecone retrieves the records with sparse vectors that most exactly match the words or phrases in the query. Query terms are scored independently and then summed, with records that have the most similar vectors scored highest. This is often called **lexical search** or **keyword search**.

## Namespace

A namespace is a partition within a [dense](#dense-index) or [sparse index](#sparse-index). It divides [records](#record) in an index into separate groups.

All [upserts](/guides/index-data/upsert-data), [queries](/guides/search/search-overview), and other [data operations](/reference/api/latest/data-plane) always target one namespace:

<img className="block max-w-full dark:hidden" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=641c2aa9a3238bf70698c583097c1f29" data-og-width="1400" width="1400" data-og-height="880" height="880" data-path="images/quickstart-upsert.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=5b394ed73cf8248f9a9ae8f9d3cdbd2d 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=3bd0b45458ebbcab40605f149b5847d5 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=9c1e1b064228344b3caf4c0a1aa8ab82 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=d40dd481d2e7cc8882d766d6df59fcba 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=bf6be475e7bed3453299f6bbecd6aa54 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=46c9efa0b08b8ac3c907a121289c19f2 2500w" />

<img className="hidden max-w-full dark:block" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=14a3e6c2847455db0821ebbf9bd51df9" data-og-width="1400" width="1400" data-og-height="880" height="880" data-path="images/quickstart-upsert-dark.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=229b97b3ea4581730afab68709201084 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=02fcdee17c689d31769c145fb86f259b 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=dd13a98f1154ffd8c4cc7cd4d37c0d33 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=b13689e423b344828452efcf91c737b4 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=319293ecf19a483936595aaebfb5cb31 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=78ad3d1499ae4d57ff9e9cd0a8e56093 2500w" />

For more information, see [Use namespaces](/guides/index-data/indexing-overview#namespaces).

## Record

A record is a basic unit of data and consists of a [record ID](#record-id), a [dense vector](#dense-vector) or a [sparse vector](#sparse-vector) (depending on the type of index), and optional [metadata](#metadata).

For more information, see [Upsert data](/guides/index-data/upsert-data).

### Record ID

A record ID is a record's unique ID. [Use ID prefixes](/guides/index-data/data-modeling#use-structured-ids) that reflect the type of data you're storing.

### Dense vector

A dense vector, also referred to as a vector embedding or simply a vector, is a series of numbers that represent the meaning and relationships of data. Each number in a dense vector corresponds to a point in a multidimensional space. Vectors that are closer together in that space are semantically similar.

Dense vectors are stored in [dense indexes](#dense-index).

You use a dense embedding model to convert data to dense vectors. The embedding model can be external to Pinecone or [hosted on Pinecone infrastructure](/guides/index-data/create-an-index#embedding-models) and integrated with an index.

For more information about dense vectors, see [What are vector embeddings?](https://www.pinecone.io/learn/vector-embeddings/).

### Sparse vector

Sparse vectors are often used to represent documents or queries in a way that captures keyword information. Each dimension in a sparse vector typically represents a word from a dictionary, and the non-zero values represent the importance of these words in the document.

Sparse vectors have a large number of dimensions, but a small number of those values are non-zero. Because most values are zero, Pinecone stores sparse vectors efficiently by keeping only the non-zero values along with their corresponding indices.

Sparse vectors are stored in [sparse indexes](#sparse-index) and [hybrid indexes](/guides/search/hybrid-search#use-a-single-hybrid-index). To convert data to sparse vectors, use a sparse embedding model. The embedding model can be external to Pinecone or [hosted on Pinecone infrastructure](/guides/index-data/create-an-index#embedding-models) and integrated with an index.

For more information about sparse vectors, see [Sparse retrieval](https://www.pinecone.io/learn/sparse-retrieval/).

### Metadata

Metadata is additional information included in a record to provide more context and enable additional [filtering capabilities](/guides/index-data/indexing-overview#metadata). For example, the original text that was embedded can be stored in the metadata.

## Other concepts

Although not represented in the diagram above, Pinecone also contains the following concepts:

* [API key](#api-key)
* [User](#user)
* [Backup or collection](#backup-or-collection)
* [Pinecone Inference](#pinecone-inference)

### API key

An API key is a unique token that [authenticates](/reference/api/authentication) and authorizes access to the [Pinecone APIs](/reference/api/introduction). API keys are project-specific.

### User

A user is a member of organizations and projects. Users are assigned specific roles at the organization and project levels that determine the user's permissions in the [Pinecone console](https://app.pinecone.io).

For more information, see [Manage organization members](/guides/organizations/manage-organization-members) and [Manage project members](/guides/projects/manage-project-members).

### Backup or collection

A backup is a static copy of a serverless index.

Backups only consume storage. They are non-queryable representations of a set of records. You can create a backup from an index, and you can create a new index from that backup. The new index configuration can differ from the original source index: for example, it can have a different name. However, it must have the same number of dimensions and similarity metric as the source index.

For more information, see [Understanding backups](/guides/manage-data/backups-overview).

### Pinecone Inference

Pinecone Inference is an API service that provides access to [embedding models](/guides/index-data/create-an-index#embedding-models) and [reranking models](/guides/search/rerank-results#reranking-models) hosted on Pinecone's infrastructure.

## Learn more

* [Vector database](https://www.pinecone.io/learn/vector-database/)
* [Pinecone APIs](/reference/api/introduction)
* [Approximate nearest neighbor (ANN) algorithms](https://www.pinecone.io/learn/a-developers-guide-to-ann-algorithms/)
* [Retrieval augmented generation (RAG)](https://www.pinecone.io/learn/retrieval-augmented-generation/)
* [Image search](https://www.pinecone.io/learn/series/image-search/)
* [Tokenization](https://www.pinecone.io/learn/tokenization/)

---

# Source: https://docs.pinecone.io/guides/production/configure-audit-logs.md

# Source: https://docs.pinecone.io/guides/assistant/admin/configure-audit-logs.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Configure audit logs

> Track user and API actions with audit log configuration.

This page describes how to configure audit logs in Pinecone. Audit logs provide a detailed record of user, service account, and API actions that occur on the management and [control plane](/guides/get-started/database-architecture#control-plane) within Pinecone. Pinecone supports Amazon S3 as a destination for audit logs.

<Note>
  To enable and manage audit logs, you must be an [organization owner](/guides/assistant/admin/organizations-overview#organization-roles). This feature is in [public preview](/assistant-release-notes/feature-availability) and available only on [Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

## Enable audit logs

Before you can enable audit logs, you need to create an IAM policy and role in Amazon S3. To start, ensure you have the following:

* A [Pinecone account](https://app.pinecone.io/).
* An [Amazon S3 bucket](https://docs.aws.amazon.com/AmazonS3/latest/userguide/creating-buckets.html).

### 1. Create an IAM policy

In the [AWS IAM console](https://console.aws.amazon.com/iam/home):

1. In the navigation pane, click **Policies**.
2. Click **Create policy**.
3. In **Select a service** section, select **S3**.
4. Select the following actions to allow:
   * `ListBucket`: Permission to list some or all of the objects in an S3 bucket.
   * `PutObject`: Permission to add an object to an S3 bucket.
5. In the **Resources** section, select **Specific**.
6. For the **bucket**, specify the ARN of the bucket you created. For example: `arn:aws:s3:::example-bucket-name`
7. For the **object**, specify an object ARN as the target resource. For example: `arn:aws:s3:::example-bucket-name/*`
8. Click **Next**.
9. Specify the name of your policy. For example:  "Pinecone-S3-Access".
10. Click **Create policy**.

### 2. Set up access using an IAM role

In the [AWS IAM console](https://console.aws.amazon.com/iam/home):

1. In the navigation pane, click **Roles**.

2. Click **Create role**.

3. In the **Trusted entity type** section, select **AWS account**.

4. Select **Another AWS account**.

5. Enter the Pinecone AWS VPC account ID: `713131977538`

6. Click **Next**.

7. Select the policy you created.

8. Click **Next**.

9. Specify the role name. For example: "Pinecone".

10. Click **Create role**.

11. Click the role you created.

12. On the **Summary** page for the role, find the **ARN**.

    For example: `arn:aws:iam::123456789012:role/PineconeAccess`

13. Copy the **ARN**.

    You will need to enter the ARN into Pinecone later.

### 3. Connect Pinecone to Amazon S3

1. Go to [**Settings > Audit logs**](https://app.pinecone.io/organizations/-/settings/logging) in the Pinecone console.
2. Enter the **Role ARN** of the IAM role you created.
3. Enter the name of the Amazon S3 bucket you created.
4. Click **Enable audit logging**.

Once you enable audit logs, Pinecone will start writing logs to the S3 bucket. In your bucket, you will also see a file named `audit-log-access-test`, which is a test file that Pinecone writes to verify that it has the necessary permissions to write logs to the bucket.

## View audit logs

Logs are written to the S3 bucket approximately every 30 minutes. Each log batch will be saved into its own file as a JSON blob, keyed by the time of the log to be written. Only logs since the integration was created and enabled will be saved.

For more information about the log schema and captured events, see [Security overview - Audit logs](/guides/assistant/admin/security-overview#audit-logs).

## Edit audit log integration details

You can edit the details of the audit log integration in the Pinecone console:

1. Go to [**Settings > Audit logs**](https://app.pinecone.io/organizations/-/settings/logging).
2. Enter the new **Role ARN** or **AWS Bucket**.
3. Click **Update settings**.

## Disable audit logs

If you disable audit logs, logs not yet saved will be lost. You can disable audit logs in the Pinecone console:

1. Go to [**Settings > Audit logs**](https://app.pinecone.io/organizations/-/settings/logging).
2. Click the toggle next to **Audit logs are active**.
3. Click **Confirm**.

## Remove audit log integration

If you remove the audit log integration, logs not yet saved will be lost. You can remove the audit log integration in the Pinecone console:

1. Go to [**Settings > Audit logs**](https://app.pinecone.io/organizations/-/settings/logging).
2. At the top of the page, click the **ellipsis (...) menu > Remove integration**.
3. Click **Remove integration**.

---

# Source: https://docs.pinecone.io/guides/production/configure-cmek.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Configure customer-managed encryption keys

> Use customer-managed encryption keys with AWS KMS.

This page describes how to set up and use customer-managed encryption keys (CMEK) to secure data within a Pinecone project. CMEK allows you to encrypt your data using keys that you manage in your cloud provider's key management system (KMS). Pinecone supports CMEK using Amazon Web Services (AWS) KMS.

<Note>
  This feature is in [public preview](/release-notes/feature-availability).
</Note>

## Set up CMEK using AWS KMS

### Before you begin

The following steps assume you have:

* Access to the [AWS console](https://console.aws.amazon.com/console/home).
* A [Pinecone Enterprise plan](https://www.pinecone.io/pricing/).

### 1. Create a role

In the [AWS console](https://console.aws.amazon.com/console/home), create a role that Pinecone can use to access the AWS Key Management System (KMS) key. You can either grant Pinecone access to a key in your account, or if your customers provide their own keys, you can grant access to keys that are outside of your account.

<Tabs>
  <Tab title="Grant access to key in your account">
    1. Open the [Amazon Identity and Access Management (IAM) console](https://console.aws.amazon.com/iam/).

    2. In the navigation pane, click **Roles**.

    3. Click **Create role**.

    4. In the **Trusted entity type** section, select **Custom trust policy**.

    5. In the **Custom trust policy** section, enter one of the following JSON snippets.

       Pick a snippet based on whether you want to allow Pinecone to assume a role from all regions or from explicit regions. Add an optional external ID for additional security. If you use an external ID, you must provide it to Pinecone when [adding a CMEK key](#add-a-key).

       <AccordionGroup>
         <Accordion title="Explicit regions + external ID">
           ```json JSON theme={null}
           {
               "Version": "2012-10-17",
               "Statement": [
                   {
                       "Sid": "AllowPineconeToAssumeIntoRoleFromExplicitRegionswithID",
                       "Effect": "Allow",
                       "Principal": {
                           "AWS": [
                               // Explicit role per Pinecone region. Replace XXXXXXXXXXXX with Pinecone's AWS account number.
                               "arn:aws:iam::XXXXXXXXXXXX:role/pinecone_cmek_access_us-east-1",
                               "arn:aws:iam::XXXXXXXXXXXX:role/pinecone_cmek_access_us-west-2",
                               "arn:aws:iam::XXXXXXXXXXXX:role/pinecone_cmek_access_eu-west-1"
                           ]
                       },
                       "Action": "sts:AssumeRole",
                       "Condition": {
                           "StringEquals": {
                               // Optional. Replace with a UUID v4 for additional security. If you use an external ID, you must provide it to Pinecone when adding an API key.
                               "sts:ExternalId": "XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX"
                           }
                       }
                   }
               ]
           }
           ```
         </Accordion>

         <Accordion title="Explicit regions + no external ID">
           ```json JSON theme={null}
           {
               "Version": "2012-10-17",
               "Statement": [
                   {
                       "Sid": "AllowPineconeToAssumeIntoRoleFromExplicitRegions",
                       "Effect": "Allow",
                       "Principal": {
                           "AWS": [
                               // Explicit role per Pinecone region. Replace XXXXXXXXXXXX with Pinecone's AWS account number.
                               "arn:aws:iam::XXXXXXXXXXXX:role/pinecone_cmek_access_us-east-1",
                               "arn:aws:iam::XXXXXXXXXXXX:role/pinecone_cmek_access_us-west-2",
                               "arn:aws:iam::XXXXXXXXXXXX:role/pinecone_cmek_access_eu-west-1"
                           ]
                       },
                       "Action": "sts:AssumeRole"
                   }
               ]
           }
           ```
         </Accordion>

         <Accordion title="All regions + external ID">
           ```json JSON theme={null}
           {
               "Version": "2012-10-17",
               "Statement": [
                   {
                       "Sid": "AllowPineconeToAssumeIntoRoleFromAllRegions",
                       "Effect": "Allow",
                       "Principal": {
                           "AWS": "*"
                       },
                       "Action": "sts:AssumeRole",
                       "Condition": {
                           "StringEquals": {
                               // Optional. Replace with a UUID v4 for additional security. If you use an external ID, you must provide it to Pinecone when adding an API key.
                               "sts:ExternalId": "XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX"
                           },
                           "StringLike": {
                               // Replace XXXXXXXXXXXX with Pinecone's AWS account number.
                               "aws:PrincipalArn": "arn:aws:iam::XXXXXXXXXXXX:role/pinecone_cmek_access_*"
                           }
                       }
                   }
               ]
           }
           ```
         </Accordion>
       </AccordionGroup>

       <Note>
         Replace `XXXXXXXXXXXX` with Pinecone's AWS account number, which can be found by going to [**Manage > CMEK**](https://app.pinecone.io/organizations/-/projects/-/cmek-encryption) in the Pinecone console and clicking **Add CMEK**.
       </Note>

    6. Click **Next**.

    7. Keep the default permissions as is and click **Next**.

    8. Enter a **Role name** and click **Create role**.

    9. Copy the **Role ARN** (e.g., `arn:aws:iam::XXXXXX:role/YYYYYY`). This will be used to [create a CMEK-enabled project](#3-create-a-cmek-enabled-project).
  </Tab>

  <Tab title="Grant access to keys outside your account">
    1. Open the [Amazon Identity and Access Management (IAM) console](https://console.aws.amazon.com/iam/).

    2. In the navigation pane, click **Roles**.

    3. Click **Create role**.

    4. In the **Trusted entity type** section, select **Custom trust policy**.

    5. In the **Custom trust policy** section, enter the following JSON:

       ```json JSON theme={null}
       {
           "Version": "2012-10-17",
           "Statement": [
               {
                   "Sid": "VisualEditor0",
                   "Effect": "Allow",
                   "Action": [
                       "kms:Decrypt",
                       "kms:Encrypt"
                   ],
                   "Resource": "arn:aws:kms:*:XXXXXX:key/*"
               }
           ]
       }
       ```

       * Replace `XXXXXX` with the account ID of the customer who owns the key.
       * Add a `Statement` array for each customer account ID.

    6. Click **Next**.

    7. Keep the default permissions as is and click **Next**.

    8. Enter a **Role name** and click **Create role**.

    9. Copy the **Role ARN** (e.g., `arn:aws:iam::XXXXXX:role/YYYYYY`). This will be used to [create a CMEK-enabled project](#3-create-a-cmek-enabled-project).
  </Tab>
</Tabs>

### 2. Create an AWS KMS key

In the [AWS console](https://console.aws.amazon.com/console/home), create the KMS key that Pinecone will use to encrypt your data:

1. Open the [Amazon Key Management Service (KMS) console](https://console.aws.amazon.com/kms/home).

2. In the navigation pane, click **Customer managed keys**.

3. Click **Create key**.

4. In the **Key type** section, select **Symmetric**.

5. In the **Key usage** section, select **Encrypt and decrypt**.

6. Under **Advanced options > Key material origin**, select **KMS**.

7. In the **Regionality** section, select **Single-Region key**.

   <Note>
     You can create a multi-regional key to safeguard against data loss in case of regional failure. However, Pinecone only accepts one Key ARN per project. If you set a multi-regional key and need to change the Key ARN to switch region, please [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket) for help.
   </Note>

8. Click **Next**.

9. Enter an **Alias** and click **Next**.

10. Keep the default administrators as is and click **Next**.

11. Select the [role you created](#1-create-a-role) from the **Key users** list and click **Next**.

12. Click **Finish**.

13. Copy the **Key ARN** (e.g., `arn:aws:kms:us-east-1:XXXXXXX:key/YYYYYYY`). This will be used to [create a CMEK-enabled project](#create-a-cmek-enabled-project).

    <Note>
      **AWS KMS automatic key rotation is supported.** Pinecone references the Key ARN, not the underlying key material. As long as the Key ARN remains unchanged and accessible, you can perform key rotations inside AWS KMS without making any changes in Pinecone.
    </Note>

### 3. Create a CMEK-enabled project

Once your [role and key is configured](#set-up-cmek-using-aws-kms), you can create a CMEK-enabled project using the Pinecone console:

1. Go to [**Settings > Organization settings > Projects**](https://app.pinecone.io/organizations/-/settings/projects).

2. Click **+Create project**.

3. Enter a **Name**.

4. Select **Encrypt with Customer Managed Encryption Key**.

5. Click **Create project**.

6. Copy and save the generated API key in a secure place for future use.

   <Warning>
     You will not be able to see the API key again after you close the dialog.
   </Warning>

7. Click **Close**.

## Add a key

To start encrypting your data with a customer-managed key, you need to add the key to the [CMEK-enabled project](#3-create-a-cmek-enabled-project) using the Pinecone console:

1. Go to [**Manage > CMEK**](https://app.pinecone.io/organizations/-/projects/-/cmek-encryption) for the CMEK-enabled project.

2. Click **Add CMEK**.

   <Warning>
     You can only add one key per project, and you cannot change the key in Pinecone once it is set.
   </Warning>

3. Enter a **Key name**.

4. Enter the **Role ARN** for the [role you created](#1-create-a-role).

5. Enter a **Key ARN** for the [key you created](#2-create-a-aws-kms-key).

6. If you [created a role](#1-create-a-role) with an external ID, enter the **External ID**. If not, leave this field blank.

7. Click **Create key**.

## Delete a key

Before a key can be deleted from a project, all indexes in the project must be deleted. Then, you can delete the key using the Pinecone console:

1. Go to the [Manage > CMEK tab](https://app.pinecone.io/organizations/-/projects/-/cmek-encryption) for the project in which the key was created.
2. For the key you want to delete, click the **ellipsis (...) menu > Delete**.
3. Enter the key name to confirm deletion.
4. Click **Delete key**.

## Limitations

* CMEK can be enabled for serverless indexes in AWS regions only.
* [Backups](/guides/manage-data/back-up-an-index) are unavailable for indexes created in a CMEK-enabled project.
* You cannot change a key once it is set.
* You can add only one key per project.

---

# Source: https://docs.pinecone.io/guides/assistant/admin/configure-sso-with-okta.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Configure SSO with Okta

> Enable SSO authentication using Okta integration.

This page describes how to set up Pinecone with Okta as the single sign-on (SSO) provider. These instructions can be adapted for any provider with SAML 2.0 support.

<Note>SSO is available on Standard and Enterprise plans.</Note>

## Before you begin

This page assumes you have the following:

* Access to your organization's [Pinecone console](https://login.pinecone.io) as an [organization owner](/guides/organizations/understanding-organizations#organization-owners).
* Access to your organization's [Okta Admin console](https://login.okta.com/).

## 1. Start SSO setup in Pinecone

First, start setting up SSO in Pinecone. In this step, you'll capture a couple values necessary for configuring Okta in [Step 2](#2-create-an-app-integration-in-okta).

1. In the Pinecone console, go to [**Settings > Manage**](https://app.pinecone.io/organizations/-/settings/manage).
2. In the **Single Sign-On** section, click **Enable SSO**.
3. In the **Setup SSO** dialog, copy the **Entity ID** and the **Assertion Consumer Service (ACS) URL**. You'll need these values in [Step 2](#2-create-an-app-integration-in-okta).
4. Click **Next**.

Keep this window or browser tab open. You'll come back to it in [Step 4](#4-complete-sso-setup-in-pinecone).

## 2. Create an app integration in Okta

In [Okta](https://login.okta.com/), follow these steps to create and configure a Pinecone app integration:

1. If you're not already on the Okta Admin console, navigate there by clicking the **Admin** button.

2. Navigate to **Applications > Applications**.

3. Click **Create App Integration**.

4. Select **SAML 2.0**.

5. Click **Next**.

6. Enter the **General Settings**:

   * **App name**: `Pinecone`
   * **App logo**: (optional)
   * **App visibility**: Set according to your organization's needs.

7. Click **Next**.

8. For **SAML Settings**, enter values you copied in [Step 1](#1-start-sso-setup-in-pinecone):

   * **Single sign-on URL**: Your **Assertion Consumer Service (ACS) URL**
   * **Audience URI (SP Entity ID)**: Your **Entity ID**
   * **Name ID format**: `EmailAddress`
   * **Application username**: `Okta username`
   * **Update application username on**: `Create and update`

9. In the **Attribute Statements** section, create the following attribute:

   * **Name**: `email`
   * **Value**: `user.email`

10. Click **Next**.

11. Click **Finish**.

## 3. Get the sign on URL and certificate from Okta

Next, in Okta, get the URL and certificate for the Pinecone application you just created. You'll use these in [Step 4](#4-complete-sso-setup-in-pinecone).

1. In the Okta Admin console, navigate to **Applications > Pinecone > Sign On**. If you're continuing from the previous step, you should already be on the right page.
2. In the **SAML 2.0** section, expand **More details**.
3. Copy the **Sign on URL**.
4. Download the **Signing Certificate**.

   <Warning>
     Download the certificate, don't copy it. The downloaded version contains necessary `-----BEGIN CERTIFICATE-----` and `-----END CERTIFICATE-----` lines.
   </Warning>

## 4. Complete SSO setup in Pinecone

In the browser tab or window you kept open in [Step 1](#1-start-sso-setup-in-pinecone), complete the SSO setup in Pinecone:

1. In the **SSO Setup** window, enter the following values:

   * **Login URL**: The URL copied in [Step 3](#3-get-the-sign-on-url-and-certificate-from-okta).
   * **Email domain**: Your company's email domain. To target multiple domains, enter each domain separated by a comma.
   * **Certificate**: The contents of the certificate file you copied in [Step 3](#3-get-the-sign-on-url-and-certificate-from-okta).

     <Warning>
       When pasting the certificate, be sure to include the `-----BEGIN CERTIFICATE-----` and `-----END CERTIFICATE-----` lines.
     </Warning>

2. Choose whether or not to **Enforce SSO for all users**.

   * If enabled, all members of your organization must use SSO to log in to Pinecone.
   * If disabled, members can choose to log in with SSO or with their Pinecone credentials.

3. Click **Next**.

4. Select a **Default role** for all users who log in with SSO. You can change user roles later.

   <Warning>
     When users first log in via SSO, they receive the default SSO role regardless of their previous role. Subsequent SSO logins do not change the role. If the default is **User**, existing owners will lose owner access on their first SSO login.

     To prevent losing access to organization management features:

     * **Sole owner**: Temporarily set the default to **Owner**, log in via SSO to retain owner access, then change the default back to **User**. After changing it back, check your organization's user list to verify no one else logged in via SSO while the default was **Owner**—if they did, adjust their roles accordingly.
     * **Multiple owners**: Keep at least one owner signed in via email while others log in via SSO. That owner can restore roles as needed, then log in via SSO last.

     If all owners lose access, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).
   </Warning>

Okta is now ready to be used for single sign-on. Follow the [Okta docs](https://help.okta.com/en-us/content/topics/users-groups-profiles/usgp-main.htm) to learn how to add users and groups.

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/configure_index.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Configure an index

> Configure an existing index. For guidance and examples, see [Manage indexes](https://docs.pinecone.io/guides/manage-data/manage-indexes).

<RequestExample>
  ```shell curl theme={null}
  # EXAMPLE REQUEST 1: Serverless index (on-demand)
  # Enable deletion protection and add tags to an 
  # existing on-demand index.
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_NAME="YOUR_INDEX_NAME"

  curl -X PATCH "https://api.pinecone.io/indexes/$INDEX_NAME" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "deletion_protection": "enabled",
          "tags": {
            "tag1": "value1",
            "tag2": "value2"
          }
        }'

  # EXAMPLE REQUEST 2: Serverless index (dedicated)
  # Add a replica to an existing dedicated index.
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_NAME="YOUR_INDEX_NAME"

  curl -X PATCH "https://api.pinecone.io/indexes/$INDEX_NAME" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "spec": {
            "serverless": {
              "read_capacity": {
                "mode": "Dedicated",
                "dedicated": {
                  "node_type": "b1",
                  "scaling": "Manual",
                  "manual": {
                    "shards": 2,
                    "replicas": 2
                  }
                }
              }
            }
          }
        }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  // EXAMPLE RESPONSE 1: Serverless index (on-demand)
  // Enable deletion protection and add tags to an 
  // existing on-demand index.
  {
    "name": "example-serverless-ondemand-index",
    "vector_type": "dense",
    "metric": "cosine",
    "dimension": 1024,
    "status": {
      "ready": true,
      "state": "Ready"
    },
    "host": "example-serverless-ondemand-index-bhnyigt.svc.aped-4627-b74a.pinecone.io",
    "spec": {
      "serverless": {
        "region": "us-east-1",
        "cloud": "aws",
        "read_capacity": {
          "mode": "OnDemand",
          "status": {
            "state": "Ready",
            "current_shards": null,
            "current_replicas": null
          }
        }
      }
    },
    "deletion_protection": "enabled",
    "tags": {
      "tag1": "value1",
      "tag2": "value2"
    },
    "embed": {
      "model": "llama-text-embed-v2",
      "field_map": {
        "text": "text"
      },
      "dimension": 1024,
      "metric": "cosine",
      "write_parameters": {
        "dimension": 1024,
        "input_type": "passage",
        "truncate": "END"
      },
      "read_parameters": {
        "dimension": 1024,
        "input_type": "query",
        "truncate": "END"
      },
      "vector_type": "dense"
    }
  }

  // EXAMPLE RESPONSE 2: Serverless index (dedicated)
  // Add a replica to an existing dedicated index.
  {
    "name": "example-serverless-dedicated-index",
    "vector_type": "dense",
    "metric": "cosine",
    "dimension": 1536,
    "status": {
      "ready": true,
      "state": "Ready"
    },
    "host": "example-serverless-dedicated-index-bhnyigt.svc.aped-4627-b74a.pinecone.io",
    "spec": {
      "serverless": {
        "region": "us-east-1",
        "cloud": "aws",
        "read_capacity": {
          "mode": "Dedicated",
          "dedicated": {
            "node_type": "b1",
            "scaling": "Manual",
            "manual": {
              "shards": 1,
              "replicas": 2 <-- desired state
            }
          },
          "status": {
            "state": "Scaling",
            "current_shards": 1,
            "current_replicas": 1 <-- current state
          }
        }
      }
    },
    "deletion_protection": "enabled",
    "tags": {
      "tag0": "value0"
    }
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml patch /indexes/{index_name}
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /indexes/{index_name}:
    patch:
      tags:
        - Manage Indexes
      summary: Configure an index
      description: >-
        Configure an existing index. For guidance and examples, see [Manage
        indexes](https://docs.pinecone.io/guides/manage-data/manage-indexes).
      operationId: configure_index
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: index_name
          description: The name of the index to configure.
          required: true
          schema:
            type: string
          example: test-index
          style: simple
      requestBody:
        description: The desired pod size and replica configuration for the index.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ConfigureIndexRequest'
            examples:
              vertical-scaling:
                summary: Vertical scaling with pod size
                value:
                  spec:
                    pod:
                      pod_type: p1.x2
              horizontal-scaling:
                summary: Horizontal scaling with replicas
                value:
                  spec:
                    pod:
                      replicas: 4
              scaling-both:
                summary: Scaling both pod size and number of replicas
                value:
                  spec:
                    pod:
                      pod_type: p1.x2
                      replicas: 4
              disable-deletion-protection:
                summary: Disable deletion protection for the index
                value:
                  delete_protection: disabled
              update-index-tags:
                summary: Update tag0 and delete tag1
                value:
                  tags:
                    tag0: new-val
                    tag1: ''
        required: true
      responses:
        '202':
          description: >-
            The request to configure the index has been accepted. Check the 
            index status to see when the change has been applied.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/IndexModel'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-metric-validation-error:
                  summary: Validation error
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Bad request. The request body included invalid request
                        parameters.
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '402':
          description: >-
            Payment required. Organization is on a paid plan and is delinquent
            on payment.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                payment-required:
                  summary: Payment required
                  value:
                    error:
                      code: PAYMENT_REQUIRED
                      message: >-
                        Request failed. Pay all past due invoices to lift
                        restrictions on your account.
                    status: 402
        '403':
          description: You've exceed your pod quota.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Forbidden
                  value:
                    error:
                      code: FORBIDDEN
                      message: Increase your quota or upgrade to create more indexes.
                    status: 403
        '404':
          description: Index not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-not-found:
                  summary: Index not found
                  value:
                    error:
                      code: NOT_FOUND
                      message: Index example-index not found.
                    status: 404
        '422':
          description: Unprocessable entity. The request body could not be deserialized.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                missing-field:
                  summary: Unprocessable entity
                  value:
                    error:
                      code: UNPROCESSABLE_ENTITY
                      message: >-
                        Failed to deserialize the JSON body into the target
                        type: missing field `metric` at line 1 column 16
                    status: 422
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    ConfigureIndexRequest:
      description: Configuration used to scale an index.
      type: object
      properties:
        spec:
          description: >-
            The spec object defines how the index should be deployed.  Only some
            attributes of an index's spec may be updated.  In general, you can
            modify settings related to scaling and  configuration but you cannot
            change the cloud or region  where the index is hosted.
          oneOf:
            - title: Serverless
              type: object
              properties:
                serverless:
                  description: Updated configuration for serverless indexes
                  type: object
                  properties:
                    read_capacity:
                      $ref: '#/components/schemas/ReadCapacity'
              required:
                - serverless
              additionalProperties: false
            - title: Pod-based
              type: object
              properties:
                pod:
                  description: Updated configuration for pod-based indexes
                  type: object
                  properties:
                    replicas:
                      description: >-
                        The number of replicas. Replicas duplicate your index.
                        They provide higher availability and throughput.
                        Replicas can be scaled up or down as your needs change.
                      default: 1
                      type: integer
                      format: int32
                      minimum: 1
                    pod_type:
                      description: >-
                        The type of pod to use. One of `s1`, `p1`, or `p2`
                        appended with `.` and one of `x1`, `x2`, `x4`, or `x8`.
                      default: p1.x1
                      type: string
              required:
                - pod
              additionalProperties: false
            - title: BYOC
              type: object
              properties:
                byoc:
                  description: Updated configuration for a BYOC index
                  type: object
                  properties:
                    read_capacity:
                      $ref: '#/components/schemas/ReadCapacity'
              required:
                - byoc
              additionalProperties: false
        deletion_protection:
          $ref: '#/components/schemas/DeletionProtection'
        tags:
          $ref: '#/components/schemas/IndexTags'
        embed:
          example:
            field_map:
              text: your-text-field
            model: multilingual-e5-large
            read_parameters:
              input_type: query
              truncate: NONE
            write_parameters:
              input_type: passage
          description: >-
            Configure the integrated inference embedding settings for this
            index.


            You can convert an existing index to an integrated index by
            specifying the embedding model and field_map. The index vector type
            and dimension must match the model vector type and dimension, and
            the index similarity metric must be supported by the model. Refer to
            the [model
            guide](https://docs.pinecone.io/guides/index-data/create-an-index#embedding-models)
            for available models and model details.


            You can later change the embedding configuration to update the field
            map, read parameters, or write parameters. Once set, the model
            cannot be changed.
          type: object
          properties:
            model:
              example: multilingual-e5-large
              description: >-
                The name of the embedding model to use with the index. The index
                dimension and model dimension must match, and the index
                similarity metric must be supported by the model. The index
                embedding model cannot be changed once set.
              type: string
            field_map:
              example:
                text: your-text-field
              description: >-
                Identifies the name of the text field from your document model
                that will be embedded.
              type: object
            read_parameters:
              description: The read parameters for the embedding model.
              type: object
            write_parameters:
              description: The write parameters for the embedding model.
              type: object
    IndexModel:
      description: >-
        The IndexModel describes the configuration and status of a Pinecone
        index.
      type: object
      properties:
        name:
          example: example-index
          description: >
            The name of the index. Resource name must be 1-45 characters long,
            start and end with an alphanumeric character, and consist only of
            lower case alphanumeric characters or '-'.
          type: string
          minLength: 1
          maxLength: 45
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If the 'vector_type' is
            'sparse', the metric must be 'dotproduct'. If the `vector_type` is
            `dense`, the metric defaults to 'cosine'.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        host:
          example: semantic-search-c01b5b5.svc.us-west1-gcp.pinecone.io
          description: The URL address where the index is hosted.
          type: string
        private_host:
          example: semantic-search-c01b5b5.svc.private.us-west1-gcp.pinecone.io
          description: The private endpoint URL of an index.
          type: string
        deletion_protection:
          $ref: '#/components/schemas/DeletionProtection'
        tags:
          $ref: '#/components/schemas/IndexTags'
        embed:
          $ref: '#/components/schemas/ModelIndexEmbed'
        spec:
          example:
            pod:
              environment: us-east-1-aws
              metadata_config:
                indexed:
                  - genre
                  - title
                  - imdb_rating
              pod_type: p1.x1
              pods: 1
              replicas: 1
              shards: 1
          description: The spec object defines how the index should be deployed.
          type: object
          oneOf:
            - title: Serverless
              type: object
              properties:
                serverless:
                  $ref: '#/components/schemas/ServerlessSpecResponse'
              required:
                - serverless
              additionalProperties: false
            - title: Pod-based
              type: object
              properties:
                pod:
                  $ref: '#/components/schemas/PodSpec'
              required:
                - pod
              additionalProperties: false
            - title: BYOC
              type: object
              properties:
                byoc:
                  $ref: '#/components/schemas/ByocSpecResponse'
              required:
                - byoc
              additionalProperties: false
        status:
          example:
            ready: true
            state: ScalingUpPodSize
          description: The current status of the index
          type: object
          properties:
            ready:
              description: Whether the index is ready for use
              type: boolean
            state:
              description: >-
                The state of the index.

                Possible values: `Initializing`, `InitializationFailed`,
                `ScalingUp`, `ScalingDown`, `ScalingUpPodSize`,
                `ScalingDownPodSize`, `Terminating`, `Ready`, or `Disabled`.
              x-enum:
                - Initializing
                - InitializationFailed
                - ScalingUp
                - ScalingDown
                - ScalingUpPodSize
                - ScalingDownPodSize
                - Terminating
                - Ready
                - Disabled
              type: string
          required:
            - ready
            - state
        vector_type:
          description: >-
            The index vector type. You can use 'dense' or 'sparse'. If 'dense',
            the vector dimension must be specified.  If 'sparse', the vector
            dimension should not be specified.
          default: dense
          type: string
      required:
        - name
        - metric
        - status
        - spec
        - host
        - vector_type
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    ReadCapacity:
      description: >-
        By default the index will be created with read capacity  mode
        `OnDemand`. If you prefer to allocate dedicated read  nodes for your
        workload, you must specify mode `Dedicated` and additional
        configurations for `node_type` and `scaling`.
      discriminator:
        propertyName: mode
        mapping:
          OnDemand: '#/components/schemas/ReadCapacityOnDemandSpec'
          Dedicated: '#/components/schemas/ReadCapacityDedicatedSpec'
      oneOf:
        - $ref: '#/components/schemas/ReadCapacityOnDemandSpec'
        - $ref: '#/components/schemas/ReadCapacityDedicatedSpec'
    DeletionProtection:
      description: >-
        Whether [deletion
        protection](http://docs.pinecone.io/guides/manage-data/manage-indexes#configure-deletion-protection)
        is enabled/disabled for the index.

        Possible values: `disabled` or `enabled`.
      default: disabled
      x-enum:
        - disabled
        - enabled
      type: string
    IndexTags:
      example:
        tag0: val0
        tag1: val1
      description: >-
        Custom user tags added to an index. Keys must be 80 characters or less.
        Values must be 120 characters or less. Keys must be alphanumeric, '_',
        or '-'.  Values must be alphanumeric, ';', '@', '_', '-', '.', '+', or '
        '. To unset a key, set the value to be an empty string.
      type: object
      additionalProperties:
        type: string
    ModelIndexEmbed:
      example:
        field_map:
          text: your-text-field
        metric: cosine
        model: multilingual-e5-large
        read_parameters:
          input_type: query
          truncate: NONE
        write_parameters:
          input_type: passage
      description: The embedding model and document fields mapped to embedding inputs.
      type: object
      properties:
        model:
          example: multilingual-e5-large
          description: The name of the embedding model used to create the index.
          type: string
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If not specified, the metric
            will be defaulted according to the model. Cannot be updated once
            set.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        vector_type:
          description: >-
            The index vector type. You can use 'dense' or 'sparse'. If 'dense',
            the vector dimension must be specified.  If 'sparse', the vector
            dimension should not be specified.
          default: dense
          type: string
        field_map:
          example:
            text: your-text-field
          description: >-
            Identifies the name of the text field from your document model that
            is embedded.
          type: object
        read_parameters:
          description: The read parameters for the embedding model.
          type: object
        write_parameters:
          description: The write parameters for the embedding model.
          type: object
      required:
        - model
    ServerlessSpecResponse:
      description: Configuration of a serverless index.
      type: object
      properties:
        cloud:
          example: aws
          description: |-
            The public cloud where you would like your index hosted.
            Possible values: `gcp`, `aws`, or `azure`.
          x-enum:
            - gcp
            - aws
            - azure
          type: string
        region:
          example: us-east-1
          description: The region where you would like your index to be created.
          type: string
        read_capacity:
          $ref: '#/components/schemas/ReadCapacityResponse'
        source_collection:
          example: movie-embeddings
          description: The name of the collection to be used as the source for the index.
          type: string
        schema:
          $ref: '#/components/schemas/MetadataSchema'
      required:
        - cloud
        - region
        - read_capacity
    PodSpec:
      example:
        environment: us-east1-gcp
        metadata_config:
          indexed:
            - genre
            - title
            - imdb_rating
        pod_type: p1.x1
        pods: 1
        replicas: 1
        shards: 1
        source_collection: movie-embeddings
      title: Pod-based
      description: Configuration needed to deploy a pod-based index.
      type: object
      properties:
        environment:
          example: us-east1-gcp
          description: The environment where the index is hosted.
          type: string
        replicas:
          description: >-
            The number of replicas. Replicas duplicate your index. They provide
            higher availability and throughput. Replicas can be scaled up or
            down as your needs change.
          default: 1
          type: integer
          format: int32
          minimum: 1
        shards:
          description: >-
            The number of shards. Shards split your data across multiple pods so
            you can fit more data into an index.
          default: 1
          type: integer
          format: int32
          minimum: 1
        pod_type:
          description: >-
            The type of pod to use. One of `s1`, `p1`, or `p2` appended with `.`
            and one of `x1`, `x2`, `x4`, or `x8`.
          default: p1.x1
          type: string
        pods:
          example: 1
          description: >-
            The number of pods to be used in the index. This should be equal to
            `shards` x `replicas`.'
          default: 1
          type: integer
          minimum: 1
        metadata_config:
          example:
            indexed:
              - genre
              - title
              - imdb_rating
          description: >-
            Configuration for the behavior of Pinecone's internal metadata
            index. By default, all metadata is indexed; when `metadata_config`
            is present, only specified metadata fields are indexed. These
            configurations are only valid for use with pod-based indexes.
          type: object
          properties:
            indexed:
              description: >-
                By default, all metadata is indexed; to change this behavior,
                use this property to specify an array of metadata fields that
                should be indexed.
              type: array
              items:
                type: string
        source_collection:
          example: movie-embeddings
          description: The name of the collection to be used as the source for the index.
          type: string
      required:
        - environment
        - pod_type
    ByocSpecResponse:
      title: BYOC
      description: Configuration of a BYOC index.
      type: object
      properties:
        environment:
          example: aws-us-east-1-b921
          description: The environment where the index is hosted.
          type: string
        read_capacity:
          $ref: '#/components/schemas/ReadCapacityResponse'
        schema:
          $ref: '#/components/schemas/MetadataSchema'
      required:
        - environment
        - read_capacity
    ReadCapacityOnDemandSpec:
      example:
        mode: OnDemand
      title: On-demand
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
      required:
        - mode
      additionalProperties: false
    ReadCapacityDedicatedSpec:
      example:
        dedicated:
          manual:
            replicas: 1
            shards: 1
          node_type: t1
          scaling: Manual
        mode: Dedicated
      title: Dedicated
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
        dedicated:
          $ref: '#/components/schemas/ReadCapacityDedicatedConfig'
      required:
        - mode
        - dedicated
      additionalProperties: true
    ReadCapacityResponse:
      description: Response containing read capacity configuration
      discriminator:
        propertyName: mode
        mapping:
          OnDemand: '#/components/schemas/ReadCapacityOnDemandSpecResponse'
          Dedicated: '#/components/schemas/ReadCapacityDedicatedSpecResponse'
      oneOf:
        - $ref: '#/components/schemas/ReadCapacityOnDemandSpecResponse'
        - $ref: '#/components/schemas/ReadCapacityDedicatedSpecResponse'
    MetadataSchema:
      example:
        fields:
          description:
            filterable: true
          genre:
            filterable: true
          year:
            filterable: true
      description: >-
        Schema for the behavior of Pinecone's internal metadata index. By
        default, all metadata is indexed; when `schema` is present, only fields
        which are present in the `fields` object with a `filterable: true` are
        indexed. Note that `filterable: false` is not currently supported.
      type: object
      properties:
        fields:
          description: >-
            A map of metadata field names to their configuration. The field name
            must be a valid metadata field name. The field name must be unique.
          type: object
          additionalProperties:
            type: object
            properties:
              filterable:
                description: >-
                  Whether the field is filterable. If true, the field is indexed
                  and can be used in filters. Only true values are allowed.
                type: boolean
      required:
        - fields
    ReadCapacityDedicatedConfig:
      description: >-
        Configuration for dedicated read capacity. See  [this
        guide](https://docs.pinecone.io/guides/index-data/dedicated-read-nodes)
        for more details on  how to configure dedicated read capacity.
      type: object
      properties:
        node_type:
          description: >-
            The type of machines to use. Available options: `b1` and `t1`. `t1`
            includes increased processing power and memory.
          type: string
        scaling:
          description: The type of scaling strategy to use.
          type: string
        manual:
          $ref: '#/components/schemas/ScalingConfigManual'
    ReadCapacityOnDemandSpecResponse:
      example:
        mode: OnDemand
        status:
          state: Ready
      title: On-demand
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
        status:
          $ref: '#/components/schemas/ReadCapacityStatus'
      required:
        - mode
        - status
      additionalProperties: false
    ReadCapacityDedicatedSpecResponse:
      example:
        dedicated:
          manual:
            replicas: 2
            shards: 2
          node_type: t1
          scaling: Manual
        mode: Dedicated
        status:
          current_replicas: 2
          current_shards: 2
          state: Ready
      title: Dedicated
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
        dedicated:
          $ref: '#/components/schemas/ReadCapacityDedicatedConfig'
        status:
          $ref: '#/components/schemas/ReadCapacityStatus'
      required:
        - mode
        - dedicated
        - status
      additionalProperties: false
    ScalingConfigManual:
      description: The config to use for manual read capacity scaling.
      type: object
      properties:
        replicas:
          description: >-
            The number of replicas to use. Replicas duplicate the compute
            resources and data of an index, allowing higher query throughput and
            availability. Setting replicas to 0 disables the index but can be
            used to reduce costs while usage is paused.
          type: integer
          format: int32
          minimum: 0
        shards:
          description: >-
            The number of shards to use. Shards determine the storage capacity
            of an index, with each shard providing 250 GB of storage.
          type: integer
          format: int32
          minimum: 1
    ReadCapacityStatus:
      description: >-
        The current status of factors affecting the read capacity of a
        serverless index
      type: object
      properties:
        state:
          description: >-
            The `state` describes the overall status of factors relating to the
            read capacity of an index. 


            Available values:

            - `Ready` is the state most of the time

            - `Scaling` if the number of replicas or shards has been recently
            updated by calling the [configure index
            endpoint](https://docs.pinecone.io/reference/api/2025-10/control-plane/configure_index)

            - `Migrating` if the index is being migrated to a new `node_type`

            - `Error` if there is an error with the read capacity configuration.
            In that case, see `error_message` for more details.
          type: string
        current_replicas:
          description: >-
            The number of replicas. Each replica has dedicated  compute
            resources and data storage. Increasing this number  will increase
            the total throughput of the index.
          type: integer
          format: int32
        current_shards:
          description: >-
            The number of shards. Each shard has dedicated storage.  Increasing
            shards alleiviates index fullness. 
          type: integer
          format: int32
        error_message:
          description: >-
            An optional error message indicating any issues with your read
            capacity configuration
          type: string
      required:
        - state
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/integrations/confluent.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Confluent

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Confluent allows you to connect and process all of your data in real time with a cloud-native and complete data streaming platform available everywhere you need it. Confluent's Kafka Connect is a tool for scalably and reliably streaming data between Apache Kafka and other data systems. It makes it simple to quickly define connectors that move large data sets in and out of Kafka.

Use the Pinecone Sink Connector, a Kafka Connect plugin for Pinecone, to take content from Confluent Cloud, convert it into vector embeddings using large language models (LLMs), and then store these embeddings in a Pinecone vector database.

<PrimarySecondaryCTA
  primaryLabel={"Get started"}
  primaryHref={
  "https://www.confluent.io/hub/confluentinc/kafka-connect-pinecone-sink"
}
  secondaryLabel={"View setup guide"}
  secondaryHref={
  "https://docs.confluent.io/cloud/current/connectors/cc-pinecone-sink.html"
}
/>

---

# Source: https://docs.pinecone.io/guides/production/connect-to-aws-privatelink.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Configure Private Endpoints for AWS PrivateLink

> Secure Pinecone with private VPC endpoints.

This page describes how to create and use [Private Endpoints](/guides/production/security-overview#private-endpoints-for-aws-privatelink) to connect AWS PrivateLink to Pinecone while keeping your VPC private from the public internet.

## Use Private Endpoints to connect to PrivateLink

### Before you begin

The following steps assume you have:

* Access to the [AWS console](https://console.aws.amazon.com/console/home).

* [Created an Amazon VPC](https://docs.aws.amazon.com/vpc/latest/userguide/create-vpc.html#create-vpc-and-other-resources) in the same AWS [region](/guides/index-data/create-an-index#cloud-regions) as the index you want to connect to. You can optionally enable DNS hostnames and resolution, if you want your VPC to automatically discover the DNS CNAME for your PrivateLink and do not want configure a CNAME.

  * To [configure the routing](https://docs.aws.amazon.com/Route53/latest/DeveloperGuide/routing-to-vpc-interface-endpoint.html) yourself, use one of Pinecone's DNS entry for the corresponding region:

  | Index region              | Pinecone DNS entry                     |
  | ------------------------- | -------------------------------------- |
  | `us-east-1` (N. Virginia) | `*.private.aped-4627-b74a.pinecone.io` |
  | `us-west-2` (Oregon)      | `*.private.apw5-4e34-81fa.pinecone.io` |
  | `eu-west-1` (Ireland)     | `*.private.apu-57e2-42f6.pinecone.io`  |

* A [Pinecone Enterprise plan](https://www.pinecone.io/pricing/).

* [Created a serverless index](/guides/index-data/create-an-index#create-a-serverless-index) in the same AWS [region](/guides/index-data/create-an-index#cloud-regions) as your Amazon VPC.

<Note>
  Private Endpoints are configured at the project-level and you can add up to 10 endpoints per project. If you have multiple projects in your organization, Private Endpoints need to be set up separately for each.
</Note>

### 1. Create an Amazon VPC endpoint

In the [AWS console](https://console.aws.amazon.com/console/home):

1. Open the [Amazon VPC console](https://console.aws.amazon.com/vpc/).

2. In the navigation pane, click **Endpoint**.

3. Click **Create endpoint**.

4. For **Service category**, select **Other endpoint services**.

5. In **Service settings**, enter the **Service name**, based on the region your Pinecone index is in:
   | Index region              | Service name                                              |
   | ------------------------- | --------------------------------------------------------- |
   | `us-east-1` (N. Virginia) | `com.amazonaws.vpce.us-east-1.vpce-svc-05ef6f1f0b9130b54` |
   | `us-west-2` (Oregon)      | `com.amazonaws.vpce.us-west-2.vpce-svc-04ecb9a0e0d5aab01` |
   | `eu-west-1` (Ireland)     | `com.amazonaws.vpce.eu-west-1.vpce-svc-03c6b7e17ff02a70f` |

6. Click **Verify service**.

7. Select the **VPC** to host the endpoint.

8. (Optional) In **Additional settings**, **Enable DNS name**.
   The enables you to access our service with the DNS name we configure. An additional CNAME record is needed if you disable this option.

9. Select the **Subnets** and **Subnet ID** for the endpoint.

10. Select the **Security groups** to apply to the endpoint.

11. Click **Create endpoint**.

12. Copy the **VPC endpoint ID** (e.g., `vpce-XXXXXXX`).
    This will be used to [add a Private Endpoint in Pinecone](#2-add-a-private-endpoint-in-pinecone).

### 2. Add a Private Endpoint in Pinecone

To add a Private Endpoint using the [Pinecone console](https://app.pinecone.io/organizations/-/projects):

1. Select your project.
2. Go to **Manage > Network**.
3. Click **Add a connection**.
4. Select your VPC region.
   Only indexes in the selected region in this project will be affected.
5. Click **Next**.
6. Enter the AWS VPC endpoint ID you copied in the [section above](#create-an-amazon-vpc-endpoint).
7. Click **Next**.
8. (optional) To **enable VPC endpoint access only**, turn the toggle on.
   This can also be enabled later. For more information, see [Manage internet access to your project](#optional-manage-internet-access-to-your-project).
9. Click **Finish setup**.

<Note>
  Private Endpoints only affect [data plane](/reference/api/latest/data-plane) access. [Control plane](/reference/api/latest/control-plane) access will continue over the public internet.
</Note>

## Read and write data

Once your private endpoint is configured, you can run data operations against an index as usual, but you must target the index using its private endpoint URL. The only difference in the URL is that `.svc.` is changed to `.svc.private.`.

You can get the private endpoint URL for an index from the Pinecone console or API.

<Tabs>
  <Tab title="Console">
    To get the private endpoint URL for an index from the Pinecone console:

    1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).
    2. Select the project containing the index.
    3. Select the index.
    4. Copy the URL under **PRIVATE ENDPOINT**.
  </Tab>

  <Tab title="API">
    To get the private endpoint URL for an index from the API, use the [`describe_index`](/reference/api/latest/control-plane/describe_index) operation, which returns the private endpoint URL as the `private_host` value:

    <CodeGroup>
      ```JavaScript JavaScript theme={null}
      import { Pinecone } from '@pinecone-database/pinecone';

      const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

      await pc.describeIndex('docs-example');
      ```

      ```go Go theme={null}
      package main

      import (
          "context"
          "encoding/json"
          "fmt"
          "log"

          "github.com/pinecone-io/go-pinecone/v4/pinecone"
      )

      func prettifyStruct(obj interface{}) string {
          bytes, _ := json.MarshalIndent(obj, "", "  ")
          return string(bytes)
      }

      func main() {
          ctx := context.Background()

          pc, err := pinecone.NewClient(pinecone.NewClientParams{
              ApiKey: "YOUR_API_KEY",
          })
          if err != nil {
              log.Fatalf("Failed to create Client: %v", err)
          }

          idx, err := pc.DescribeIndex(ctx, "docs-example")
          if err != nil {
              log.Fatalf("Failed to describe index \"%v\": %v", idx.Name, err)
          } else {
              fmt.Printf("index: %v\n", prettifyStruct(idx))
          }
      }
      ```

      ```bash curl theme={null}
      PINECONE_API_KEY="YOUR_API_KEY"

      curl -i -X GET "https://api.pinecone.io/indexes/docs-example" \
          -H "Api-Key: YOUR_API_KEY" \
          -H "X-Pinecone-Api-Version: 2025-10"
      ```
    </CodeGroup>

    The response includes the private endpoint URL as the `private_host` value:

    <CodeGroup>
      ```json JavaScript {6} theme={null}
      {
        name: 'docs-example',
        dimension: 1536,
        metric: 'cosine',
        host: 'docs-example-jl7boae.svc.aped-4627-b74a.pinecone.io',
        privateHost: 'docs-example-jl7boae.svc.private.aped-4627-b74a.pinecone.io',
        deletionProtection: 'disabled',
        tags: { environment: 'production' },
        embed: undefined,
        spec: {
          byoc: undefined,
          pod: undefined,
          serverless: { cloud: 'aws', region: 'us-east-1' }
        },
        status: { ready: true, state: 'Ready' },
        vectorType: 'dense'
      }
      ```

      ```go Go {5} theme={null}
      index: {
        "name": "docs-example",
        "dimension": 1536,
        "host": "docs-example-jl7boae.svc.aped-4627-b74a.pinecone.io",
        "private_host": "docs-example-jl7boae.svc.private.aped-4627-b74a.pinecone.io",
        "metric": "cosine",
        "deletion_protection": "disabled",
        "spec": {
          "serverless": {
            "cloud": "aws",
            "region": "us-east-1"
          }
        },
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "tags": {
          "environment": "production"
        }
      }
      ```

      ```json curl {12} theme={null}
      {
        "id": "025117b3-e683-423c-b2d1-6d30fbe5027f",
        "vector_type": "dense",
        "name": "docs-example",
        "metric": "cosine",
        "dimension": 1536,
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "host": "docs-example-jl7boae.svc.aped-4627-b74a.pinecone.io",
        "private_host": "docs-example-jl7boae.svc.private.aped-4627-b74a.pinecone.io",
        "spec": {
          "serverless": {
            "region": "us-east-1",
            "cloud": "aws"
          }
        },
        "deletion_protection": "disabled",
        "tags": {
          "environment": "production"
        }
      ```
    </CodeGroup>
  </Tab>
</Tabs>

<Note>
  If you run data operations against an index from outside the Private Endpoint, you will get an `Unauthorized` response.
</Note>

## Manage internet access to your project

Once your Private Endpoint is configured, you can turn off internet access to your project. To enable VPC endpoint access only:

1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).
2. Select your project.
3. Go to **Network > Access**.
4. Turn the **VPC endpoint access only** toggle on.
   This will turn off internet access to the project. This can be turned off at any point.

   <Warning>
     This access control is set at the *project-level* and can unintentionally affect Pinecone indexes that communicate via the internet in the same project. Only indexes communicating through Private Endpoints will continue to work.
   </Warning>

## Manage Private Endpoints

In addition to [creating Private Endpoints](#2-add-a-private-endpoint-in-pinecone), you can also:

* [View Private Endpoints](#view-private-endpoints)
* [Delete a Private Endpoint](#delete-a-private-endpoint)

### View Private Endpoints

To view Private Endpoints using the [Pinecone console](https://app.pinecone.io/organizations/-/projects):

1. Select your project.
2. Go to **Manage > Network**.
   A list of Private Endpoints displays with the associated **VPC ID** and **Cloud** provider.

### Delete a Private Endpoint

To delete a Private Endpoint using the [Pinecone console](https://app.pinecone.io/organizations/-/projects):

1. Select your project.
2. Go to **Manage > Network**.
3. For the Private Endpoint you want to delete, click the *...* (Actions) icon.
4. Click **Delete**.
5. Enter the endpoint name.
6. Click **Delete Endpoint**.

---

# Source: https://docs.pinecone.io/integrations/build-integration/connect-your-users-to-pinecone.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Connect your users to Pinecone

To reduce friction for users using your integration, you can create a [custom object](#custom-object), like a button or link, to trigger a **Connect to Pinecone** popup from your app, website, or [Colab](https://colab.google/) notebook. Within this popup, your users can sign up for or log in to Pinecone, select or create an organization and project to connect to, and generate an API key. The API key is then communicated back to the user to copy or directly sent to the hosting page, app, or notebook.

Alternatively, you can embed our [pre-built widget](#pre-built-widget), which provides the same functionality, but with the ease of a drop-in component.

To start, [create an integration ID](#create-an-integration-id) for your app.

<Note>
  Only [organization owners](/guides/organizations/manage-organization-members) can add or manage integrations.
</Note>

## Create an integration ID

Create a unique `integrationId` to enable usage of the **Connect to Pinecone** [popup](#custom-object) and [widget](#pre-built-widget):

1. On the the [**Integrations**](https://app.pinecone.io/organizations/-/settings/integrations) tab in the Pinecone console, click the **Create Integration** button.

   <Note>The **Integrations** tab does not display unless your organization already has integrations. [Follow this link to create your first integration](https://app.pinecone.io/organizations/-/settings/integrations?create=true).</Note>

2. Fill out the **Create integration** form:
   * **Integration name**: Give your integration a name.

   * **URL Slug**: This is your `integrationID`. Enter a human-readable string that uniquely identifies your integration and that may appear in URLs. Your integration URL slug is public and cannot be changed.

   * **Logo**: Upload a logo for your integration.

   * **Return mechanism**: Select one of the following return methods for the generated API key:
     * **Web Message**: Your application will receive the Pinecone API key via a web message. Select this option if you are using the [@pinecone-database/connect library](/integrations/build-integration/connect-your-users-to-pinecone#javascript). The API key will only be provided to the allowed origin(s) specified below.
     * **Copy/Paste**: The API key will display in the success message, and users will need to copy and paste their Pinecone API keys into your application.

   * **Allowed origin**: If you selected **Web Message** as your **Return mechanism**, list the URL origin(s) where your integration is hosted. The [origin](https://developer.mozilla.org/en-US/docs/Glossary/Origin) is the part of the URL that specifies the protocol, hostname, and port.

3. Click **Create**.

<Note>
  Anyone can create an integration, but [becoming an official Pinecone partner](/integrations/build-integration/integration-ecosystem) can help accelerate your go-to-market and add value to your customers.
</Note>

## Custom object

[Once you have created your `integrationId`](#create-an-integration-id), you can create a custom object, like a button or link, that loads a **Connect to Pinecone** popup that displays as follows:

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-popup.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=f8fd4a551afc7bf73657dc8af27c214f" alt="Connect popup" data-og-width="624" width="624" data-og-height="657" height="657" data-path="images/connect-popup.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-popup.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=52daa926896ec46a4e7d214943a4566b 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-popup.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=40250701b646e305df925ee05377bd8b 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-popup.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=db918eb1651a56bb1f1718aaa9188355 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-popup.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=a59d709c3dbf79357cc08c740503feff 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-popup.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=074b7adcb2c5b4ebc8674df31f3afe9f 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-popup.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=d1a171eb825d863d5c81ee0a736daf39 2500w" />

The `ConnectPopup` function can be called with either the JavaScript library or script. The JavaScript library is the most commonly used method, but the script can be used in instances where you cannot build and use a custom library, like within the constraints of a content management system (CMS).

The function includes the following **required** configuration option:

* `integrationId`: The slug assigned to the integration. If `integrationId` is not passed, the widget will not render.

  <Note>To create a unique `integrationId`, fill out the [Create Integration form](#create-an-integration-id).</Note>

The function returns an object containing the following:

* `open`: A function that opens the popup. Suitable for use as an on-click handler.

Example usage of the library and script:

<CodeGroup>
  ```javascript JavaScriptlibrary theme={null}
  import { ConnectPopup } from '@pinecone-database/connect'

  /*  Define a function called connectWithAPIKey */
  const connectWithAPIKey = () => {
    return new Promise((resolve, reject) => {
      /* Call ConnectPopup function with an object containing options */
      const popup = ConnectPopup({
        onConnect: (key) => {
          resolve(key);
        },
        integrationId: 'myApp'
      }).open();
    });
  };

  /* Handle button click event */
  document.getElementById('connectButton').addEventListener('click', () => {
    connectWithAPIKey()
      .then(apiKey => {
        console.log("API Key:", apiKey);
      })
      .catch(error => {
        console.error("Error:", error);
      });
  });
  ```

  ```html JavaScript script theme={null}
  <head>
  ...
  <script src="https://connect.pinecone.io/embed.js"></script>
  <script>
    const pineconePopup = ConnectPopup({
      onConnect: (key) => console.log(key),
      onCancel: () => console.log("Cancelled"),
      integrationId: 'myApp'
    });
  </script>
  ...
  </head>
  <body>
  ...
  <button onclick="pineconePopup.open()">Connect to Pinecone!</button>
  ...
  </body>
  ```
</CodeGroup>

Once you have created your integration, be sure to [attribute usage to your integration](/integrations/build-integration/attribute-usage-to-your-integration).

## Pre-built widget

The pre-built **Connect** widget displays as follows:

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-widget.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=83c1d4eea967cabb4eef94ace9088943" alt="Connect widget" data-og-width="514" width="514" data-og-height="192" height="192" data-path="images/connect-widget.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-widget.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=d7c1d97bdb81c2542cf83676b8349faa 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-widget.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=d153f812aad16c9d59a202000eddddfd 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-widget.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=b55b5cf20017205db4e78ee6fb963a9e 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-widget.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=a214f92f4f9689d8d53e4d7407182b89 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-widget.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=4cde4082bde874f202bea09125632cb0 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/connect-widget.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=f739d78c95b0ef344f850e25a0f02e60 2500w" />

[Once you have created your `integrationId`](#create-an-integration-id), you can embed the **Connect** widget multiple ways:

* [JavaScript](#javascript) library (`@pinecone-database/connect`) or script: Renders the widget in apps and websites.
* [Colab](#colab) (`pinecone-notebooks`): Renders the widget in Colab notebooks using Python.

Once you have created your integration, be sure to [attribute usage to your integration](/integrations/build-integration/attribute-usage-to-your-integration).

### JavaScript

To embed the **Connect to Pinecone** widget in your app or website using the [`@pinecone-database/connect` library](https://www.npmjs.com/package/@pinecone-database/connect), install the necessary dependencies:

```shell Shell theme={null}
# Install dependencies
npm i -S @pinecone-database/connect
```

You can use the JavaScript library to render the **Connect to Pinecone** widget and obtain the API key with the [`connectToPinecone` function](#connecttopinecone-function). It displays the widget and calls the provided callback function with the Pinecone API key, once the user completes the flow.

The function includes the following **required** configuration options:

* `integrationId`: The slug assigned to the integration. If `integrationId` is not passed, the widget will not render.

  <Note>To create a unique `integrationId`, [fill out the Create Integration form](#create-an-integration-id) with Pinecone.</Note>

* `container`: The HTML element where the **Connect** widget will render.

Example usage:

```JavaScript JavaScript theme={null}
import {connectToPinecone} from '@pinecone-database/connect'

const setupPinecone = (apiKey) => { /* Set up a Pinecone client using the API key */ }

connectToPinecone(
  setupPinecone,
  {
    integrationId: 'myApp',
    container: document.getElementById('connect-widget')
  }
)
```

If you cannot use the JavaScript library, you can directly call the script. For example:

```html HTML theme={null}
<head>
  ...
  <script src="https://connect.pinecone.io/embed.js">
  ...
</head>
<body>

<div id="widgetContainer"></div>
...

<script>
  connectToPinecone(
    (apiKey) => {/* Use the apiKey to create an index*/},
    {
      integrationId: 'myApp',
      container: document.getElementById('widgetContainer'),
    }
  );
</script>

...
</body>
```

### Colab

To embed the **Connect** widget in your Colab notebook, use the [`pinecone-notebooks` Python library](https://pypi.org/project/pinecone-notebooks/#description):

```shell  theme={null}
# Install dependencies using Colab syntax

pip install -qU pinecone-notebooks pinecone[grpc]
```

```python  theme={null}
# Render the Connect widget for the user to authenticate and generate an API key

from pinecone_notebooks.colab import Authenticate

Authenticate()

# The generated API key is available in the PINECONE_API_KEY environment variable

from pinecone.grpc import PineconeGRPC as Pinecone
from pinecone import ServerlessSpec
import os

api_key = os.environ.get('PINECONE_API_KEY')

# Use the API key to initialize the Pinecone client
pc = Pinecone(api_key=api_key)
```

<Tip>
  To see this flow in practice, see our [example notebook](https://colab.research.google.com/drive/1VZ-REFRbleJG4tfJ3waFIrSveqrYQnNx?usp=sharing).
</Tip>

## Manage generated API keys

Your users can [manage the API keys](/guides/projects/manage-api-keys) generated by your integration in the Pinecone console.

---

# Source: https://docs.pinecone.io/troubleshooting/contact-support.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Contact Support

Pinecone Support is available to customers on the **Standard** billing plan.

First-response SLAs only apply to tickets created by users in an organization subscribed to a [support plan](https://www.pinecone.io/pricing/?plans=support). To upgrade your support plan, go to [Manage your support plan](https://app.pinecone.io/organizations/-/settings/support/plans) in the console and select your desired plan. Before subscribing to a support plan, you must [Upgrade your billing plan](/guides/organizations/manage-billing/upgrade-billing-plan) to the **Standard** tier.

Our business hours are Monday to Friday from 8:00 AM to 8:00 PM Eastern Time. We are closed on US federal holidays. Customers subscribed to the **Pro** or **Premium** support plan have 24/7/365 on-call availability, which is triggered by creating a SEV-1 ticket.

All customers subscribed to a support plan can create tickets in the Pinecone console. To create a ticket, go to the [Help center](https://app.pinecone.io/organizations/-/settings/support/ticket), select [Create a ticket](https://app.pinecone.io/organizations/-/settings/support/ticket/create), and fill out the form.

If you are not subscribed to a support plan and need help logging into the console, upgrading your support or billing plan, or have questions about billing, please complete the [public support contact form](https://www.pinecone.io/contact/support/) and we will respond if necessary.

If you have general questions about the platform, pricing, events, and more, [contact our sales team](https://www.pinecone.io/contact/). For inquiries about partnerships, please [contact the Pinecone Partner Program](https://www.pinecone.io/partners/#sales-contact-form-submissions).

---

# Source: https://docs.pinecone.io/integrations/context-data.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Context Data

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Using [Context Data](https://contextdata.ai/), you can easily create end-to-end data flows by connecting to a myriad of data sources (PostgreSQL, MySQL, Amazon S3, Salesforce, etc.), seamlessly embedding and writing the results to Pinecone using Context Data's super simple no-code web interface. These flows can also be configured to be triggered and run on a user-defined schedule.

Additionally, Context Data provides the ability to create transformations like aggregations, joins, and feature engineering using SQL common table expressions before writing to Pinecone.

You also have the ability to "chat" with your Pinecone indexes directly from Context Data's privacy-focused Query Studio.

<PrimarySecondaryCTA primaryLabel="Get started" primaryHref="https://context-data.gitbook.io/context-data-1/adding-target-connections#add-a-pinecone-target-connection" />

---

# Source: https://docs.pinecone.io/guides/assistant/context-snippets-overview.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Context snippets overview

> Retrieve context snippets from your assistant's knowledge base.

You can [retrieve the context snippets](/guides/assistant/retrieve-context-snippets) that Pinecone Assistant uses to generate its responses. This data includes relevant chunks, relevancy scores, and references.

## Use cases

Retrieving context snippets is useful for performing tasks like the following:

* Understanding what relevant data snippets Pinecone Assistant is providing to the LLM for chat generation.
* Using the retrieved snippets with your own LLM.
* Using the retrieved snippets with your own RAG application or agentic workflow.

## SDK support

The Pinecone [Python SDK](/reference/sdks/python/overview) and [Node.js SDK](/reference/sdks/node/overview) provide convenient programmatic access to [retrieve context snippets](/reference/api/latest/assistant/context_assistant).

## Pricing

Context retrieval usage is [measured in tokens](/guides/assistant/pricing-and-limits#token-usage), similar to Pinecone Assistant. See [Pricing](https://www.pinecone.io/pricing/) for up-to-date pricing information.

<Note>
  Pricing updates specific to context retrieval will be made as the feature becomes generally available.
</Note>

---

# Source: https://docs.pinecone.io/reference/api/2025-10/assistant/context_assistant.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Retrieve context from an assistant

> Retrieve context snippets from an assistant to use as part of RAG or any agentic flow.

For guidance and examples, see [Retrieve context snippets](https://docs.pinecone.io/guides/assistant/retrieve-context-snippets).

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME/context" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "accept: application/json" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "query": "Who is the CFO of Netflix?"
  }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
      "snippets":
      [
          {
              "type":"text",
              "content":"EXHIBIT 31.3\nCERTIFICATION OF CHIEF FINANCIAL OFFICER\nPURSUANT TO SECTION 302 OF THE SARBANES-OXLEY ACT OF 2002\nI, Spencer Neumann, certify that: ...",
              "score":0.9960699,
              "reference":
              {
                  "type":"pdf",
                  "file":
                  {
                      "status":"Available","id":"e6034e51-0bb9-4926-84c6-70597dbd07a7",
                      "name":"Netflix-10-K-01262024.pdf", 
                      "size":1073470,
                      "metadata":null,
                      "updated_on":"2024-11-21T22:59:10.426001030Z",
                      "created_on":"2024-11-21T22:58:35.879120257Z", 
                      "percent_done":1.0,
                      "signed_url":"https://storage.googleapis.com...",
                      "error_message":null
                      },
                  "pages":[78]
              }
          },
  {
      "type":"text",
      "content":"EXHIBIT 32.1\n..."
  ...
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/assistant_data_2025-10.oas.yaml post /chat/{assistant_name}/context
openapi: 3.0.3
info:
  title: Pinecone Assistant Data Plane API
  description: >-
    Pinecone Assistant Engine is a context engine to store and retrieve relevant
    knowledge from millions of documents at scale. This API supports
    interactions with assistants.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{assistant_host}
    variables:
      assistant_host:
        default: unknown
        description: host of the created assistant
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Assistants
    description: Actions that manage Assistants
paths:
  /chat/{assistant_name}/context:
    post:
      tags:
        - Manage Assistants
      summary: Retrieve context from an assistant
      description: >-
        Retrieve context snippets from an assistant to use as part of RAG or any
        agentic flow.


        For guidance and examples, see [Retrieve context
        snippets](https://docs.pinecone.io/guides/assistant/retrieve-context-snippets).
      operationId: context_assistant
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: assistant_name
          description: The name of the assistant to be described.
          required: true
          schema:
            type: string
          example: test-assistant
          style: simple
      requestBody:
        description: The desired configuration to retrieve context from an assistant.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ContextRequest'
        required: true
      responses:
        '200':
          description: Context retrieval process successful.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ContextModel'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                files-validation-error:
                  summary: Validation error on ingest.
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Uploaded file can only currently be either a pdf or txt
                        file
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Assistant not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                assistant-not-found:
                  summary: Assistant not found.
                  value:
                    error:
                      code: NOT_FOUND
                      message: Assistant "example-assistant" not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    ContextRequest:
      description: Parameters to retrieve context from an assistant.
      type: object
      properties:
        query:
          description: >-
            The query that is used to generate the context. Exactly one of query
            or messages should be provided.
          type: string
        filter:
          example:
            genre:
              $ne: documentary
          description: >-
            Optionally filter which documents can be retrieved using the
            following metadata fields.
          type: object
        messages:
          description: >-
            The list of messages to use for generating the context. Exactly one
            of query or messages should be provided.
          type: array
          items:
            $ref: '#/components/schemas/MessageModel'
        top_k:
          example: 20
          description: >-
            The maximum number of context snippets to return. Default is 16.
            Maximum is 64.
          type: integer
        snippet_size:
          example: 4096
          description: >-
            The maximum context snippet size. Default is 2048 tokens. Minimum is
            512 tokens. Maximum is 8192 tokens.
          type: integer
        multimodal:
          description: >-
            Whether or not to retrieve image-related context snippets. If
            `false`, only text snippets are returned.
          default: true
          type: boolean
        include_binary_content:
          description: >-
            If image-related context snippets are returned, this field
            determines whether or not they should include base64 image data. If
            `false`, only the image captions are returned. Only available when
            `multimodal=true`.
          default: true
          type: boolean
    ContextModel:
      description: The response format containing the context from an assistant.
      type: object
      properties:
        id:
          type: string
        snippets:
          type: array
          items:
            $ref: '#/components/schemas/SnippetModel'
        usage:
          $ref: '#/components/schemas/UsageModel'
      required:
        - snippets
        - usage
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: Uploaded file can only currently be either a pdf or txt file
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    MessageModel:
      description: Describes the format of a message in a chat.
      type: object
      properties:
        role:
          description: Role of the message such as 'user' or 'assistant'
          type: string
        content:
          description: Content of the message
          type: string
    SnippetModel:
      description: Represents a part of a document that is relevant to the user query.
      discriminator:
        propertyName: type
        mapping:
          text: '#/components/schemas/TextSnippetModel'
          multimodal: '#/components/schemas/MultiModalSnippetModel'
      type: object
      oneOf:
        - $ref: '#/components/schemas/TextSnippetModel'
        - $ref: '#/components/schemas/MultiModalSnippetModel'
    UsageModel:
      description: Describes the usage of a chat completion.
      type: object
      properties:
        prompt_tokens:
          type: integer
        completion_tokens:
          type: integer
        total_tokens:
          type: integer
    TextSnippetModel:
      description: Represents a text context snippet.
      type: object
      properties:
        type:
          description: The type of context snippet.
          type: string
        content:
          type: string
        score:
          type: number
          format: float
        reference:
          $ref: '#/components/schemas/TypedReferenceModel'
      required:
        - type
        - content
        - score
        - reference
    MultiModalSnippetModel:
      description: Represents a multimodal context snippet.
      type: object
      properties:
        type:
          description: The type of context snippet.
          type: string
        content:
          type: array
          items:
            $ref: '#/components/schemas/MultiModalContentBlocksModel'
        score:
          type: number
          format: float
        reference:
          $ref: '#/components/schemas/TypedReferenceModel'
      required:
        - type
        - content
        - score
        - reference
    TypedReferenceModel:
      description: Represents a reference for the information provided.
      discriminator:
        propertyName: type
        mapping:
          text: '#/components/schemas/TextReferenceModel'
          json: '#/components/schemas/JsonReferenceModel'
          markdown: '#/components/schemas/MarkdownReferenceModel'
          pdf: '#/components/schemas/PdfReferenceModel'
          doc_x: '#/components/schemas/DocxReferenceModel'
      type: object
      oneOf:
        - $ref: '#/components/schemas/TextReferenceModel'
        - $ref: '#/components/schemas/JsonReferenceModel'
        - $ref: '#/components/schemas/MarkdownReferenceModel'
        - $ref: '#/components/schemas/PdfReferenceModel'
        - $ref: '#/components/schemas/DocxReferenceModel'
    MultiModalContentBlocksModel:
      description: Represents a block in the multimodal content of a context snippet.
      discriminator:
        propertyName: type
        mapping:
          text: '#/components/schemas/MultiModalContentTextBlockModel'
          image: '#/components/schemas/MultiModalContentImageBlockModel'
      type: object
      oneOf:
        - $ref: '#/components/schemas/MultiModalContentTextBlockModel'
        - $ref: '#/components/schemas/MultiModalContentImageBlockModel'
    TextReferenceModel:
      description: Represents a reference to a part of a text document.
      type: object
      properties:
        type:
          description: The type of reference. Always "text".
          type: string
        file:
          $ref: '#/components/schemas/AssistantFileModel'
      required:
        - type
        - file
    JsonReferenceModel:
      description: Represents a reference to a json document.
      type: object
      properties:
        type:
          description: The type of reference. Always "json".
          type: string
        file:
          $ref: '#/components/schemas/AssistantFileModel'
      required:
        - type
        - file
    MarkdownReferenceModel:
      description: Represents a reference to a part of a markdown document.
      type: object
      properties:
        type:
          description: The type of reference. Always "markdown".
          type: string
        file:
          $ref: '#/components/schemas/AssistantFileModel'
      required:
        - type
        - file
    PdfReferenceModel:
      description: Represents a reference to a part of a PDF document.
      type: object
      properties:
        type:
          description: The type of reference. Always "pdf".
          type: string
        file:
          $ref: '#/components/schemas/AssistantFileModel'
        pages:
          type: array
          items:
            type: integer
      required:
        - type
        - file
        - pages
    DocxReferenceModel:
      description: Represents a reference to a part of a docx document.
      type: object
      properties:
        type:
          description: The type of reference. Always "doc_x".
          type: string
        file:
          $ref: '#/components/schemas/AssistantFileModel'
        pages:
          type: array
          items:
            type: integer
      required:
        - type
        - file
        - pages
    MultiModalContentTextBlockModel:
      description: Represents a text block in a multimodal context snippet.
      type: object
      properties:
        type:
          description: The type of multimodal content block.
          type: string
        text:
          type: string
      required:
        - type
        - text
    MultiModalContentImageBlockModel:
      description: Represents an image block in a multimodal context snippet.
      type: object
      properties:
        type:
          description: The type of multimodal content block.
          type: string
        caption:
          type: string
        image:
          $ref: '#/components/schemas/ImageModel'
      required:
        - type
        - caption
    AssistantFileModel:
      description: The response format for a successful file upload request.
      type: object
      properties:
        name:
          type: string
        id:
          type: string
          format: uuid
        metadata:
          nullable: true
          type: object
        created_on:
          type: string
          format: date-time
        updated_on:
          type: string
          format: date-time
        status:
          description: >-
            The current state of the uploaded file. Possible values:

            - `Processing`: File is being processed (parsed, chunked, embedded)

            - `Available`: Processing completed successfully; file is ready for
            use

            - `Deleting`: Deletion has been initiated but not yet completed

            - `ProcessingFailed`: Processing failed with an error


            Note: Once a file is deleted, the API returns 404 Not Found instead
            of a file object.
          x-enum:
            - Processing
            - Available
            - Deleting
            - ProcessingFailed
          type: string
        percent_done:
          nullable: true
          description: The percentage of the file that has been processed
          type: number
          format: double
        signed_url:
          nullable: true
          example: https://storage.googleapis.com/bucket/file.pdf?...
          description: >-
            A [signed
            URL](https://cloud.google.com/storage/docs/access-control/signed-urls)
            that provides temporary, read-only access to the underlying file.
            Anyone with the link can access the file, so treat it as sensitive
            data. Expires after a short time.
          type: string
        error_message:
          nullable: true
          description: >-
            A message describing any error during file processing. Provided only
            if an error occurs.
          type: string
        multimodal:
          description: Indicates whether the file was processed as multimodal.
          type: boolean
      required:
        - id
        - name
    ImageModel:
      nullable: true
      description: Represents the data for an image.
      type: object
      properties:
        type:
          description: The format of the image data. Currently, this is always "base64".
          type: string
        mime_type:
          description: The MIME type of the image (e.g., "image/jpeg").
          type: string
        data:
          description: The image data. Currently, this is always a base64-encoded string.
          type: string
      required:
        - type
        - mime_type
        - data
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: Pinecone API Key

````

---

# Source: https://docs.pinecone.io/troubleshooting/cors-issues.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# CORS Issues

Cross-Origin Resource Sharing (CORS) is an HTTP-header based security feature that
allows a server to indicate which domains, schemes or ports a browser should accept
content from. When a browser-based app, by default, only loads content from the same
origin as the original request, CORS errors can appear if the responses come from
a different origin. Pinecone's current implementation of CORS can cause this mismatch
and display the following error:

```console console theme={null}
No 'Access-Control-Allow-Origin' header is present on the requested resource.
```

This error occurs in response to cross-origin requests. Most commonly, it occurs when a user is running a local web server with the hostname `localhost`, which Pinecone's Same Origin Policy (SOP) treats as distinct from the IP address of the local machine.

To resolve this issue, host your web server on an external server with a public IP address and DNS name entry.

### About Localhost (running a web server locally)

Localhost is not inherently a problem. However, when running a web server on a local machine (e.g., laptop or desktop computer), using "localhost" as the hostname can cause issues with cross-origin resource sharing (CORS).

The reason for this is that the Same-Origin Policy (SOP) enforced by web browsers treats "localhost" as a different origin than the actual IP address of the machine. For example, if a web application running on "localhost" makes a cross-origin request to a server running on the actual IP address of the machine, the browser will treat it as a cross-origin request and enforce the SOP.

To allow cross-origin requests between "localhost" and the actual IP address of the machine, the server needs to explicitly allow them by including the appropriate CORS headers in its response. However, as mentioned earlier, running a web server on a local machine can present security risks and is generally not recommended for production use.

Therefore, while "localhost" itself is not a problem, using it as the hostname for a web server can cause CORS issues that need to be properly addressed. Additionally, running a web server on a local machine should be done with caution and only for development or testing purposes, rather than for production use.

---

# Source: https://docs.pinecone.io/guides/indexes/pods/create-a-pod-based-index.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create a pod-based index

> Create and configure a pod-based Pinecone index

<Warning>
  Customers who sign up for a Standard or Enterprise plan on or after August 18, 2025 cannot create pod-based indexes. Instead, create [serverless indexes](/guides/index-data/create-an-index), and consider using [dedicated read nodes](/guides/index-data/dedicated-read-nodes) for large workloads (millions of records or more, and moderate or high query rates).
</Warning>

This page shows you how to create a pod-based index. For guidance on serverless indexes, see [Create a serverless index](/guides/index-data/create-an-index).

## Create a pod index

To create a pod index, use the [`create_index`](/reference/api/latest/control-plane/create_index) operation as follows:

* Provide a `name` for the index.
* Specify the `dimension` and `metric` of the vectors you'll store in the index. This should match the dimension and metric supported by your embedding model.
* Set `spec.environment` to the [environment](/guides/index-data/create-an-index#cloud-regions) where the index should be deployed. For Python, you also need to import the `ServerlessSpec` class.
* Set `spec.pod_type` to the [pod type](/guides/indexes/pods/understanding-pod-based-indexes#pod-types) and [size](/guides/index-data/indexing-overview#pod-size-and-performance) that you want.

Other parameters are optional. See the [API reference](/reference/api/latest/control-plane/create_index) for details.

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone, PodSpec

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.create_index(
    name="docs-example",
    dimension=1536,
    metric="cosine",
    spec=PodSpec(
      environment="us-west1-gcp",
      pod_type="p1.x1",
      pods=1
    ),
    deletion_protection="disabled"

  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({
    apiKey: 'YOUR_API_KEY'
  });

  await pc.createIndex({
    name: 'docs-example',
    dimension: 1536,
    metric: 'cosine',
    spec: {
      pod: {
        environment: 'us-west1-gcp',
        podType: 'p1.x1',
        pods: 1
      }
    },
    deletionProtection: 'disabled',
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.IndexModel;
  import org.openapitools.db_control.client.model.DeletionProtection;

  public class CreateIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          pc.createPodsIndex("docs-example", 1536, "us-west1-gcp",
                  "p1.x1", "cosine", DeletionProtection.DISABLED);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      indexName := "docs-example"
    	metric := pinecone.Dotproduct
  	  deletionProtection := pinecone.DeletionProtectionDisabled

      idx, err := pc.CreatePodIndex(ctx, &pinecone.CreatePodIndexRequest{
          Name:               indexName,
          Metric:             &metric,
          Dimension:          1536,
          Environment:        "us-east1-gcp",
          PodType:            "p1.x1",
          DeletionProtection: &deletionProtection,
      })
      if err != nil {
          log.Fatalf("Failed to create pod-based index: %v", idx.Name)
      } else {
          fmt.Printf("Successfully created pod-based index: %v", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var createIndexRequest = await pinecone.CreateIndexAsync(new CreateIndexRequest
  {
      Name = "docs-example",
      Dimension = 1536,
      Metric = MetricType.Cosine,
      Spec = new PodIndexSpec
      {
          Pod = new PodSpec
          {
              Environment = "us-east1-gcp",
              PodType = "p1.x1",
              Pods = 1,
          }
      },
      DeletionProtection = DeletionProtection.Disabled
  });
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s "https://api.pinecone.io/indexes" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
           "name": "docs-example",
           "dimension": 1536,
           "metric": "cosine",
           "spec": {
              "pod": {
                 "environment": "us-west1-gcp",
                 "pod_type": "p1.x1",
                 "pods": 1
              }
           },
           "deletion_protection": "disabled"
        }'
  ```
</CodeGroup>

## Create a pod index from a collection

You can create a pod-based index from a collection. For more details, see [Restore an index](/guides/indexes/pods/restore-a-pod-based-index).

---

# Source: https://docs.pinecone.io/guides/projects/create-a-project.md

# Source: https://docs.pinecone.io/guides/assistant/admin/create-a-project.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create a project

> Create a new Pinecone project in your organization.

This page shows you how to create a project.

If you are an [organization owner or user](/guides/organizations/understanding-organizations#organization-roles), you can create a project in your organization:

<Tabs>
  <Tab title="Pinecone console">
    1. In the Pinecone console, go to [**your profile > Organization settings > Projects**](https://app.pinecone.io/organizations/-/settings/projects).

    2. Click **+ Create Project**.

    3. Enter a **Name**.

       <Note>
         A project name can contain up to 512 characters. For more information, see [Object identifiers](/reference/api/database-limits#identifier-limits).
       </Note>

    4. (Optional) Tags are key-value pairs that you can use to categorize and identify the project. To add a tag, click **+ Add tag** and enter a tag key and value.

    5. (Optional) Select **Encrypt with Customer Managed Encryption Key**. For more information, see [Configure CMEK](/guides/production/configure-cmek).

    6. Click **Create project**.

       To load an index with a [sample dataset](/guides/data/use-sample-datasets), click **Load sample data** and follow the prompts.

    <Note>
      Organizations on the Starter plan are limited to one project. To create additional projects, [upgrade to the Standard or Enterprise plan](/guides/organizations/manage-billing/upgrade-billing-plan).
    </Note>
  </Tab>

  <Tab title="Code">
    <Note>
      An [access token](/guides/organizations/manage-service-accounts#retrieve-an-access-token) must be provided to complete this action through the Admin API. The Admin API is in [public preview](/release-notes/feature-availability).
    </Note>

    <CodeGroup>
      ```bash curl theme={null}
      PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"

      curl "https://api.pinecone.io/admin/projects" \
          -H "X-Pinecone-Api-Version: 2025-10" \
      	-H "Authorization: Bearer $PINECONE_ACCESS_TOKEN" \
      	-d '{
                "name":"example-project"
              }'
      ```

      ```bash CLI theme={null}
      # Target the organization for which you want to 
      # create a project.
      pc target -o "example-org"
      # Create the project and set it as the target 
      # project for the CLI.
      pc project create -n "example-project" --target
      ```
    </CodeGroup>

    The example returns a response like the following:

    <CodeGroup>
      ```json curl theme={null}
      {
        "id": "32c8235a-5220-4a80-a9f1-69c24109e6f2",
        "name": "example-project",
        "max_pods": 0,
        "force_encryption_with_cmek": false,
        "organization_id": "-NM7af6f234168c4e44a",
        "created_at": "2025-03-16T22:46:45.030Z"
      }
      ```

      ```text CLI theme={null}
      [SUCCESS] Project example-cli-project created successfully.

      ATTRIBUTE           VALUE
      Name                example-project
      ID                  32c8235a-5220-4a80-a9f1-69c24109e6f2
      Organization ID     -NM7af6f234168c4e44a
      Created At          2025-10-27 23:27:46.370088 +0000 UTC
      Force Encryption    false
      Max Pods            5

      [SUCCESS] Target project set to example-cli-project
      ```
    </CodeGroup>
  </Tab>
</Tabs>

## Next steps

* [Add users to your project](/guides/projects/manage-project-members#add-members-to-a-project)
* [Create an index](/guides/index-data/create-an-index)

---

# Source: https://docs.pinecone.io/guides/index-data/create-an-index.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create an index

> Create dense or sparse indexes for semantic and lexical search.

* **Dense indexes** store dense vectors, which are numerical representations of the meaning and relationships of text, images, or other types of data. You use dense indexes for [semantic search](/guides/search/semantic-search) or in combination with sparse indexes for [hybrid search](/guides/search/hybrid-search).

* **Sparse indexes** store sparse vectors, which are numerical representations of the words or phrases in a document. You use sparse indexes for [lexical search](/guides/search/lexical-search), or in combination with dense indexes for [hybrid search](/guides/search/hybrid-search).

<Tip>
  You can create an index using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/create-index/serverless).
</Tip>

## Create a dense index

You can create a dense index with [integrated vector embedding](/guides/index-data/indexing-overview#integrated-embedding) or a dense index for storing vectors generated with an external embedding model.

### Integrated embedding

<Note>
  Indexes with integrated embedding do not support [updating](/guides/manage-data/update-data) or [importing](/guides/index-data/import-data) with text.
</Note>

If you want to upsert and search with source text and have Pinecone convert it to dense vectors automatically, [create a dense index with integrated embedding](/reference/api/latest/control-plane/create_for_model) as follows:

* Provide a `name` for the index.
* Set `cloud` and `region` to the [cloud and region](/guides/index-data/create-an-index#cloud-regions) where the index should be deployed.
* Set `embed.model` to one of [Pinecone's hosted embedding models](/guides/index-data/create-an-index#embedding-models).
* Set `embed.field_map` to the name of the field in your source document that contains the data for embedding.

Other parameters are optional. See the [API reference](/reference/api/latest/control-plane/create_for_model) for details.

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  index_name = "integrated-dense-py"

  if not pc.has_index(index_name):
      pc.create_index_for_model(
          name=index_name,
          cloud="aws",
          region="us-east-1",
          embed={
              "model":"llama-text-embed-v2",
              "field_map":{"text": "chunk_text"}
          }
      )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await pc.createIndexForModel({
    name: 'integrated-dense-js',
    cloud: 'aws',
    region: 'us-east-1',
    embed: {
      model: 'llama-text-embed-v2',
      fieldMap: { text: 'chunk_text' },
    },
    waitUntilReady: true,
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.ApiException;
  import org.openapitools.db_control.client.model.CreateIndexForModelRequest;
  import org.openapitools.db_control.client.model.CreateIndexForModelRequestEmbed;
  import org.openapitools.db_control.client.model.DeletionProtection;
  import java.util.HashMap;
  import java.util.Map;

  public class CreateIntegratedIndex {
      public static void main(String[] args) throws ApiException {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          String indexName = "integrated-dense-java";
          String region = "us-east-1";
          HashMap<String, String> fieldMap = new HashMap<>();
          fieldMap.put("text", "chunk_text");
          CreateIndexForModelRequestEmbed embed = new CreateIndexForModelRequestEmbed()
                  .model("llama-text-embed-v2")
                  .fieldMap(fieldMap);
          Map<String, String> tags = new HashMap<>();
          tags.put("environment", "development");
          pc.createIndexForModel(
                  indexName,
                  CreateIndexForModelRequest.CloudEnum.AWS,
                  region,
                  embed,
                  DeletionProtection.DISABLED,
                  tags
          );
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

    	indexName := "integrated-dense-go"
     	deletionProtection := pinecone.DeletionProtectionDisabled

      idx, err := pc.CreateIndexForModel(ctx, &pinecone.CreateIndexForModelRequest{
  		Name:   indexName,
  		Cloud:  pinecone.Aws,
  		Region: "us-east-1",
  		Embed: pinecone.CreateIndexForModelEmbed{
  			Model:    "llama-text-embed-v2",
  			FieldMap: map[string]interface{}{"text": "chunk_text"},
  		},
          DeletionProtection: &deletionProtection,
          Tags:   &pinecone.IndexTags{ "environment": "development" },
  	})
      if err != nil {
          log.Fatalf("Failed to create serverless integrated index: %v", idx.Name)
      } else {
          fmt.Printf("Successfully created serverless integrated index: %v", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var createIndexRequest = await pinecone.CreateIndexForModelAsync(
      new CreateIndexForModelRequest
      {
          Name = "integrated-dense-dotnet",
          Cloud = CreateIndexForModelRequestCloud.Aws,
          Region = "us-east-1",
          Embed = new CreateIndexForModelRequestEmbed
          {
              Model = "llama-text-embed-v2",
              FieldMap = new Dictionary<string, object?>() 
              { 
                  { "text", "chunk_text" } 
              }
          },
          DeletionProtection = DeletionProtection.Disabled,
          Tags = new Dictionary<string, string> 
          { 
              { "environment", "development" }
          }
      }
  );
  ```

  ```json curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -X POST "https://api.pinecone.io/indexes/create-for-model" \
       -H "Content-Type: application/json" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H "X-Pinecone-Api-Version: 2025-10" \
       -d '{
             "name": "integrated-dense-curl",
             "cloud": "aws",
             "region": "us-east-1",
             "embed": {
               "model": "llama-text-embed-v2",
               "field_map": {
                 "text": "chunk_text"
               }
             }
           }'
  ```

  ```bash CLI theme={null}
  # Target the project where you want to create the index.
  pc target -o "example-org" -p "example-project"
  # Create the index.
  pc index create \
    --name "integrated-dense-cli" \
    --metric "cosine" \
    --cloud "aws" \
    --region "us-east-1" \
    --model "llama-text-embed-v2" \
    --field_map "text=chunk_text" \
    --tags "environment=development"
  ```
</CodeGroup>

### Bring your own vectors

If you use an external embedding model to convert your data to dense vectors, [create a dense index](/reference/api/latest/control-plane/create_index) as follows:

* Provide a `name` for the index.
* Set the `vector_type` to `dense`.
* Specify the `dimension` and similarity `metric` of the vectors you'll store in the index. This should match the dimension and metric supported by your embedding model.
* Set `spec.cloud` and `spec.region` to the [cloud and region](/guides/index-data/create-an-index#cloud-regions) where the index should be deployed. For Python, you also need to import the `ServerlessSpec` class.

Other parameters are optional. See the [API reference](/reference/api/latest/control-plane/create_index) for details.

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone
  from pinecone import ServerlessSpec

  pc = Pinecone(api_key="YOUR_API_KEY")

  index_name = "standard-dense-py"

  if not pc.has_index(index_name):
      pc.create_index(
          name=index_name,
          vector_type="dense",
          dimension=1536,
          metric="cosine",
          spec=ServerlessSpec(
              cloud="aws",
              region="us-east-1"
          ),
          deletion_protection="disabled",
          tags={
              "environment": "development"
          }
      )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await pc.createIndex({
    name: 'standard-dense-js',
    vectorType: 'dense',
    dimension: 1536,
    metric: 'cosine',
    spec: {
      serverless: {
        cloud: 'aws',
        region: 'us-east-1'
      }
    },
    deletionProtection: 'disabled',
    tags: { environment: 'development' }, 
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.IndexModel;
  import org.openapitools.db_control.client.model.DeletionProtection;
  import java.util.HashMap;

  public class CreateServerlessIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          String indexName = "standard-dense-java";
          String cloud = "aws";
          String region = "us-east-1";
          String vectorType = "dense";
          Map<String, String> tags = new HashMap<>();
          tags.put("environment", "development");
          pc.createServerlessIndex(
              indexName,
              "cosine", 
              1536, 
              cloud,
              region,
              DeletionProtection.DISABLED, 
              tags, 
              vectorType
          );
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // Serverless index
    	indexName := "standard-dense-go"
  	vectorType := "dense"
      dimension := int32(1536)
      metric := pinecone.Cosine
  	deletionProtection := pinecone.DeletionProtectionDisabled

      idx, err := pc.CreateServerlessIndex(ctx, &pinecone.CreateServerlessIndexRequest{
          Name:               indexName,
          VectorType:         &vectorType,
          Dimension:          &dimension,
          Metric:             &metric,
          Cloud:              pinecone.Aws,
          Region:             "us-east-1",
          DeletionProtection: &deletionProtection,
          Tags:               &pinecone.IndexTags{ "environment": "development" },
      })
      if err != nil {
          log.Fatalf("Failed to create serverless index: %v", err)
      } else {
          fmt.Printf("Successfully created serverless index: %v", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var createIndexRequest = await pinecone.CreateIndexAsync(new CreateIndexRequest
  {
      Name = "standard-dense-dotnet",
      VectorType = VectorType.Dense,
      Dimension = 1536,
      Metric = MetricType.Cosine,
      Spec = new ServerlessIndexSpec
      {
          Serverless = new ServerlessSpec
          {
              Cloud = ServerlessSpecCloud.Aws,
              Region = "us-east-1"
          }
      },
      DeletionProtection = DeletionProtection.Disabled,
      Tags = new Dictionary<string, string> 
      {  
          { "environment", "development" }
      }
  });
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -X POST "https://api.pinecone.io/indexes" \
       -H "Accept: application/json" \
       -H "Content-Type: application/json" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H "X-Pinecone-Api-Version: 2025-10" \
       -d '{
             "name": "standard-dense-curl",
             "vector_type": "dense",
             "dimension": 1536,
             "metric": "cosine",
             "spec": {
               "serverless": {
                 "cloud": "aws",
                 "region": "us-east-1"
                }
              },
              "tags": {
                "environment": "development"
              },
              "deletion_protection": "disabled"
           }'
  ```

  ```bash CLI theme={null}
  # Target the project where you want to create the index.
  pc target -o "example-org" -p "example-project"
  # Create the index.
  pc index create \
    --name "standard-dense-cli" \
    --vector_type "dense" \
    --dimension 1536 \
    --metric "cosine" \
    --cloud "aws" \
    --region "us-east-1" \
    --tags "environment=development" \
    --deletion_protection "disabled"
  ```
</CodeGroup>

## Create a sparse index

You can create a sparse index with [integrated vector embedding](/guides/index-data/indexing-overview#integrated-embedding) or a sparse index for storing vectors generated with an external embedding model.

### Integrated embedding

If you want to upsert and search with source text and have Pinecone convert it to sparse vectors automatically, [create a sparse index with integrated embedding](/reference/api/latest/control-plane/create_for_model) as follows:

* Provide a `name` for the index.
* Set `cloud` and `region` to the [cloud and region](/guides/index-data/create-an-index#cloud-regions) where the index should be deployed.
* Set `embed.model` to one of [Pinecone's hosted sparse embedding models](/guides/index-data/create-an-index#embedding-models).
* Set `embed.field_map` to the name of the field in your source document that contains the text for embedding.
* If needed, `embed.read_parameters` and `embed.write_parameters` can be used to override the default model embedding behavior.

Other parameters are optional. See the [API reference](/reference/api/latest/control-plane/create_for_model) for details.

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  index_name = "integrated-sparse-py"

  if not pc.has_index(index_name):
      pc.create_index_for_model(
          name=index_name,
          cloud="aws",
          region="us-east-1",
          embed={
              "model":"pinecone-sparse-english-v0",
              "field_map":{"text": "chunk_text"}
          }
      )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await pc.createIndexForModel({
    name: 'integrated-sparse-js',
    cloud: 'aws',
    region: 'us-east-1',
    embed: {
      model: 'pinecone-sparse-english-v0',
      fieldMap: { text: 'chunk_text' },
    },
    waitUntilReady: true,
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.ApiException;
  import org.openapitools.db_control.client.model.CreateIndexForModelRequest;
  import org.openapitools.db_control.client.model.CreateIndexForModelRequestEmbed;
  import org.openapitools.db_control.client.model.DeletionProtection;
  import java.util.HashMap;
  import java.util.Map;

  public class CreateIntegratedIndex {
      public static void main(String[] args) throws ApiException {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          String indexName = "integrated-sparse-java";
          String region = "us-east-1";
          HashMap<String, String> fieldMap = new HashMap<>();
          fieldMap.put("text", "chunk_text");
          CreateIndexForModelRequestEmbed embed = new CreateIndexForModelRequestEmbed()
                  .model("pinecone-sparse-english-v0")
                  .fieldMap(fieldMap);
          Map<String, String> tags = new HashMap<>();
          tags.put("environment", "development");
          pc.createIndexForModel(
                  indexName,
                  CreateIndexForModelRequest.CloudEnum.AWS,
                  region,
                  embed,
                  DeletionProtection.DISABLED,
                  tags
          );
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

    	indexName := "integrated-sparse-go"
  	deletionProtection := pinecone.DeletionProtectionDisabled

      idx, err := pc.CreateIndexForModel(ctx, &pinecone.CreateIndexForModelRequest{
  		Name:   indexName,
  		Cloud:  pinecone.Aws,
  		Region: "us-east-1",
  		Embed: pinecone.CreateIndexForModelEmbed{
  			Model:    "pinecone-sparse-english-v0",
  			FieldMap: map[string]interface{}{"text": "chunk_text"},
  		},
          DeletionProtection: &deletionProtection,
          Tags:   &pinecone.IndexTags{ "environment": "development" },

  	})
      if err != nil {
          log.Fatalf("Failed to create serverless integrated index: %v", idx.Name)
      } else {
          fmt.Printf("Successfully created serverless integrated index: %v", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var createIndexRequest = await pinecone.CreateIndexForModelAsync(
      new CreateIndexForModelRequest
      {
          Name = "integrated-sparse-dotnet",
          Cloud = CreateIndexForModelRequestCloud.Aws,
          Region = "us-east-1",
          Embed = new CreateIndexForModelRequestEmbed
          {
              Model = "pinecone-sparse-english-v0",
              FieldMap = new Dictionary<string, object?>() 
              { 
                  { "text", "chunk_text" } 
              }
          },
          DeletionProtection = DeletionProtection.Disabled,
          Tags = new Dictionary<string, string> 
          { 
              { "environment", "development" }
          }
      }
  );
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -X POST "https://api.pinecone.io/indexes/create-for-model" \
       -H "Content-Type: application/json" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H "X-Pinecone-Api-Version: 2025-10" \
       -d '{
             "name": "integrated-sparse-curl",
             "cloud": "aws",
             "region": "us-east-1",
             "embed": {
               "model": "pinecone-sparse-english-v0",
               "field_map": {
                 "text": "chunk_text"
               }
             }
           }'
  ```

  ```bash CLI theme={null}
  # Target the project where you want to create the index.
  pc target -o "example-org" -p "example-project"
  # Create the index.
  pc index create \
    --name "integrated-sparse-cli" \
    --cloud "aws" \
    --region "us-east-1" \
    --model "pinecone-sparse-english-v0" \
    --field_map "text=chunk_text" \
    --tags "environment=development"
  ```
</CodeGroup>

### Bring your own vectors

If you use an external embedding model to convert your data to sparse vectors, [create a sparse index](/reference/api/latest/control-plane/create_index) as follows:

* Provide a `name` for the index.
* Set the `vector_type` to `sparse`.
* Set the distance `metric` to `dotproduct`. Sparse indexes do not support other [distance metrics](/guides/index-data/indexing-overview#distance-metrics).
* Set `spec.cloud` and `spec.region` to the cloud and region where the index should be deployed.

Other parameters are optional. See the [API reference](/reference/api/latest/control-plane/create_index) for details.

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone, ServerlessSpec

  pc = Pinecone(api_key="YOUR_API_KEY")

  index_name = "standard-sparse-py"

  if not pc.has_index(index_name):
      pc.create_index(
          name=index_name,
          vector_type="sparse",
          metric="dotproduct",
          spec=ServerlessSpec(cloud="aws", region="us-east-1")
      )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await pc.createIndex({
    name: 'standard-sparse-js',
    vectorType: 'sparse',
    metric: 'dotproduct',
    spec: {
      serverless: {
        cloud: 'aws',
        region: 'us-east-1'
      },
    },
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.DeletionProtection;

  import java.util.*;

  public class SparseIndex {
      public static void main(String[] args) throws InterruptedException {
          // Instantiate Pinecone class
          Pinecone pinecone = new Pinecone.Builder("YOUR_API_KEY").build();

          // Create sparse Index
          String indexName = "standard-sparse-java";
          String cloud = "aws";
          String region = "us-east-1";
          String vectorType = "sparse";
          Map<String, String> tags = new HashMap<>();
          tags.put("env", "test");
          pinecone.createSparseServelessIndex(indexName,
                  cloud,
                  region,
                  DeletionProtection.DISABLED,
                  tags,
                  vectorType);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
  	ctx := context.Background()

  	pc, err := pinecone.NewClient(pinecone.NewClientParams{
  		ApiKey: "YOUR_API_KEY",
  	})
  	if err != nil {
  		log.Fatalf("Failed to create Client: %v", err)
  	}

  	indexName := "standard-sparse-go"
  	vectorType := "sparse"
  	metric := pinecone.Dotproduct
  	deletionProtection := pinecone.DeletionProtectionDisabled

  	idx, err := pc.CreateServerlessIndex(ctx, &pinecone.CreateServerlessIndexRequest{
  		Name:               indexName,
  		Metric:             &metric,
  		VectorType:         &vectorType,
  		Cloud:              pinecone.Aws,
  		Region:             "us-east-1",
  		DeletionProtection: &deletionProtection,
  	})
  	if err != nil {
  		log.Fatalf("Failed to create serverless index: %v", err)
  	} else {
  		fmt.Printf("Successfully created serverless index: %v", idx.Name)
  	}
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var createIndexRequest = await pinecone.CreateIndexAsync(new CreateIndexRequest
  {
      Name = "standard-sparse-dotnet",
      VectorType = VectorType.Sparse,
      Metric = MetricType.Dotproduct,
      Spec = new ServerlessIndexSpec
      {
          Serverless = new ServerlessSpec
          {
              Cloud = ServerlessSpecCloud.Aws,
              Region = "us-east-1"
          }
      }
  });
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -X POST "https://api.pinecone.io/indexes" \
       -H "Accept: application/json" \
       -H "Content-Type: application/json" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H "X-Pinecone-Api-Version: 2025-10" \
       -d '{
              "name": "standard-sparse-curl",
              "vector_type": "sparse",
              "metric": "dotproduct",
              "spec": {
                 "serverless": {
                    "cloud": "aws",
                    "region": "us-east-1"
                 }
              }
           }'
  ```

  ```bash CLI theme={null}
  # Target the project where you want to create the index.
  pc target -o "example-org" -p "example-project"
  # Create the index.
  pc index create \
    --name "standard-sparse-cli" \
    --vector_type "sparse" \
    --metric "dotproduct" \
    --cloud "aws" \
    --region "us-east-1" \
    --tags "environment=development"
  ```
</CodeGroup>

## Create an index from a backup

You can create a dense or sparse index from a backup. For more details, see [Restore an index](/guides/manage-data/restore-an-index).

## Metadata indexing

<Warning>
  This feature is in [early access](/release-notes/feature-availability) and available only on the `2025-10` version of the API. The CLI does not yet support this feature.
</Warning>

Pinecone indexes all metadata fields by default. However, large amounts of metadata can cause slower [index building](/guides/get-started/database-architecture#index-builder) as well as slower [query execution](/guides/get-started/database-architecture#query-executors), particularly when data is not cached in a query executor's memory and local SSD and must be fetched from object storage.

To prevent performance issues due to excessive metadata, you can limit metadata indexing to the fields that you plan to use for [query filtering](/guides/search/filter-by-metadata).

### Set metadata indexing

You can set metadata indexing during index creation or [namespace creation](/reference/api/2025-10/data-plane/createnamespace):

* Index-level metadata indexing rules apply to all namespaces that don't have explicit metadata indexing rules.
* Namespace-level metadata indexing rules overrides index-level metadata indexing rules.

For example, let's say you want to store records that represent chunks of a document, with each record containing many metadata fields. Since you plan to use only a few of the metadata fields to filter queries, you would specify the metadata fields to index as follows.

<Warning>
  Metadata indexing cannot be changed after index or namespace creation.
</Warning>

<CodeGroup>
  ```shell Index-level metadata indexing theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://api.pinecone.io/indexes" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "name": "example-index-metadata",
          "vector_type": "dense",
          "dimension": 1536,
          "metric": "cosine",
          "spec": {
            "serverless": {
              "cloud": "aws",
              "region": "us-east-1",
              "schema": {
                "fields": {
                  "document_id": {
                    "filterable": true
                  },
                  "document_title": {
                    "filterable": true
                  },
                  "chunk_number": {
                    "filterable": true
                  },
                  "document_url": {
                    "filterable": true
                  },
                  "created_at": {
                    "filterable": true
                  }
                }
              }
            }
          },
          "deletion_protection": "disabled"
        }'
  ```

  ```shell Namespace-level metadata indexing theme={null}
  # To learn how to get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/namespaces" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "name": "example-namespace",
          "schema": {
            "fields": {
              "document_id": {
                "filterable": true
              },
              "document_title": {
                "filterable": true
              },
              "chunk_number": {
                "filterable": true
              },
              "document_url": {
                "filterable": true
              },
              "created_at": {
                "filterable": true
              }
            }
          }
        }'
  ```
</CodeGroup>

### Check metadata indexing

To check which metadata fields are indexed, you can describe the index or namespace:

<CodeGroup>
  ```shell Describe index theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -X GET "https://api.pinecone.io/indexes/example-index-metadata" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H "X-Pinecone-Api-Version: 2025-10"
  ```

  ```shell Describe namespace theme={null}
  # To learn how to get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X GET "https://$INDEX_HOST/namespaces/example-namespace" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

The response includes the `schema` object with the names of the metadata fields explicitly indexed during index or namespace creation.

<Note>
  The response does not include unindexed metadata fields or metadata fields indexed by default.
</Note>

<CodeGroup>
  ```json Describe index theme={null}
  {
    "id": "751ab850-6e61-4f92-bd23-fa129803d207",
    "vector_type": "dense",
    "name": "example-index",
    "metric": "cosine",
    "dimension": 1536,
    "status": {
      "ready": false,
      "state": "Initializing"
    },
    "host": "example-index-fa77d8e.svc.aped-4627-b74a.pinecone.io",
    "spec": {
      "serverless": {
        "region": "us-east-1",
        "cloud": "aws",
        "read_capacity": {
          "mode": "OnDemand",
          "status": "Ready"
        },
        "schema": {
          "fields": {
            "document_id": {
              "filterable": true
            },
            "document_title": {
              "filterable": true
            },
            "created_at": {
              "filterable": true
            },
            "chunk_number": {
              "filterable": true
            },
            "document_url": {
              "filterable": true
            }
          }
        }
      }
    },
    "deletion_protection": "disabled",
    "tags": null
  }

  ```

  ```json Describe namespace theme={null}
  {
    "name": "example-namespace",
    "record_count": "20000",
    "schema": {
      "fields": {
        "document_title": {
          "filterable": true
        },
        "document_url": {
          "filterable": true
        },
        "chunk_number": {
          "filterable": true
        },
        "document_id": {
          "filterable": true
        },
        "created_at": {
          "filterable": true
        }
      }
    }
  }
  ```
</CodeGroup>

## Index options

### Cloud regions

When creating an index, you must choose the cloud and region where you want the index to be hosted. The following table lists the available public clouds and regions and the plans that support them:

| Cloud   | Region                       | [Supported plans](https://www.pinecone.io/pricing/) | [Availability phase](/release-notes/feature-availability) |
| ------- | ---------------------------- | --------------------------------------------------- | --------------------------------------------------------- |
| `aws`   | `us-east-1` (Virginia)       | Starter, Standard, Enterprise                       | General availability                                      |
| `aws`   | `us-west-2` (Oregon)         | Standard, Enterprise                                | General availability                                      |
| `aws`   | `eu-west-1` (Ireland)        | Standard, Enterprise                                | General availability                                      |
| `gcp`   | `us-central1` (Iowa)         | Standard, Enterprise                                | General availability                                      |
| `gcp`   | `europe-west4` (Netherlands) | Standard, Enterprise                                | General availability                                      |
| `azure` | `eastus2` (Virginia)         | Standard, Enterprise                                | General availability                                      |

The cloud and region cannot be changed after a serverless index is created.

<Note>
  On the free Starter plan, you can create serverless indexes in the `us-east-1` region of AWS only. To create indexes in other regions, [upgrade your plan](/guides/organizations/manage-billing/upgrade-billing-plan).
</Note>

### Similarity metrics

When creating a dense index, you can choose from the following similarity metrics. For the most accurate results, choose the similarity metric used to train the embedding model for your vectors. For more information, see [Vector Similarity Explained](https://www.pinecone.io/learn/vector-similarity/).

<Note>[Sparse indexes](#sparse-indexes) must use the `dotproduct` metric.</Note>

<AccordionGroup>
  <Accordion title="Euclidean">
    Querying indexes with this metric returns a similarity score equal to the squared Euclidean distance between the result and query vectors.

    This metric calculates the square of the distance between two data points in a plane. It is one of the most commonly used distance metrics. For an example, see our [IT threat detection example](https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/it-threat-detection.ipynb).

    When you use `metric='euclidean'`, the most similar results are those with the **lowest similarity score**.
  </Accordion>

  <Accordion title="Cosine">
    This is often used to find similarities between different documents. The advantage is that the scores are normalized to \[-1,1] range. For an example, see our [generative question answering example](https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/gen-qa-openai.ipynb).
  </Accordion>

  <Accordion title="Dotproduct">
    This is used to multiply two vectors. You can use it to tell us how similar the two vectors are. The more positive the answer is, the closer the two vectors are in terms of their directions. For an example, see our [semantic search example](https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/semantic-search.ipynb).
  </Accordion>
</AccordionGroup>

### Embedding models

[Dense vectors](/guides/get-started/concepts#dense-vector) and [sparse vectors](/guides/get-started/concepts#sparse-vector) are the basic units of data in Pinecone and what Pinecone was specially designed to store and work with. Dense vectors represents the semantics of data such as text, images, and audio recordings, while sparse vectors represent documents or queries in a way that captures keyword information.

To transform data into vector format, you use an embedding model. Pinecone hosts several embedding models so it's easy to manage your vector storage and search process on a single platform. You can use a hosted model to embed your data as an integrated part of upserting and querying, or you can use a hosted model to embed your data as a standalone operation.

The following embedding models are hosted by Pinecone.

<Note>
  To understand how cost is calculated for embedding, see [Embedding cost](/guides/manage-cost/understanding-cost#embedding). To get model details via the API, see [List models](/reference/api/latest/inference/list_models) and [Describe a model](/reference/api/latest/inference/describe_model).
</Note>

#### multilingual-e5-large

[`multilingual-e5-large`](/models/multilingual-e5-large) is an efficient dense embedding model trained on a mixture of multilingual datasets. It works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs).

**Details**

* Vector type: Dense
* Modality: Text
* Dimension: 1024
* Recommended similarity metric: Cosine
* Max sequence length: 507 tokens
* Max batch size: 96 sequences

For rate limits, see [Embedding tokens per minute](/reference/api/database-limits#embedding-tokens-per-minute-per-model) and [Embedding tokens per month](/reference/api/database-limits#embedding-tokens-per-month-per-model).

**Parameters**

The `multilingual-e5-large` model supports the following parameters:

| Parameter    | Type   | Required/Optional | Description                                                                                                                                                                                                                                    | Default |
| :----------- | :----- | :---------------- | :--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | :------ |
| `input_type` | string | Required          | The type of input data. Accepted values: `query` or `passage`.                                                                                                                                                                                 |         |
| `truncate`   | string | Optional          | How to handle inputs longer than those supported by the model. Accepted values: `END` or `NONE`.<br /><br />`END` truncates the input sequence at the input token limit. `NONE` returns an error when the input exceeds the input token limit. | `END`   |

#### llama-text-embed-v2

[`llama-text-embed-v2`](/models/llama-text-embed-v2) is a high-performance dense embedding model optimized for text retrieval and ranking tasks. It is trained on a diverse range of text corpora and provides strong performance on longer passages and structured documents.

**Details**

* Vector type: Dense
* Modality: Text
* Dimension: 1024 (default), 2048, 768, 512, 384
* Recommended similarity metric: Cosine
* Max sequence length: 2048 tokens
* Max batch size: 96 sequences

For rate limits, see [Embedding tokens per minute](/reference/api/database-limits#embedding-tokens-per-minute-per-model) and [Embedding tokens per month](/reference/api/database-limits#embedding-tokens-per-month-per-model).

**Parameters**

The `llama-text-embed-v2` model supports the following parameters:

| Parameter    | Type    | Required/Optional | Description                                                                                                                                                                                                                                    | Default |
| :----------- | :------ | :---------------- | :--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | :------ |
| `input_type` | string  | Required          | The type of input data. Accepted values: `query` or `passage`.                                                                                                                                                                                 |         |
| `truncate`   | string  | Optional          | How to handle inputs longer than those supported by the model. Accepted values: `END` or `NONE`.<br /><br />`END` truncates the input sequence at the input token limit. `NONE` returns an error when the input exceeds the input token limit. | `END`   |
| `dimension`  | integer | Optional          | Dimension of the vector to return.                                                                                                                                                                                                             | 1024    |

#### pinecone-sparse-english-v0

[`pinecone-sparse-english-v0`](/models/pinecone-sparse-english-v0) is a sparse embedding model for converting text to [sparse vectors](/guides/get-started/concepts#sparse-vector) for keyword or hybrid semantic/keyword search. Built on the innovations of the [DeepImpact architecture](https://arxiv.org/pdf/2104.12016), the model directly estimates the lexical importance of tokens by leveraging their context, unlike traditional retrieval models like BM25, which rely solely on term frequency.

**Details**

* Vector type: Sparse
* Modality: Text
* Recommended similarity metric: Dotproduct
* Max sequence length: 512 or 2048
* Max batch size: 96 sequences

For rate limits, see [Embedding tokens per minute](/reference/api/database-limits#embedding-tokens-per-minute-per-model) and [Embedding tokens per month](/reference/api/database-limits#embedding-tokens-per-month-per-model).

**Parameters**

The `pinecone-sparse-english-v0` model supports the following parameters:

| Parameter                 | Type    | Required/Optional | Description                                                                                                                                                                                                                                                                    | Default |
| :------------------------ | :------ | :---------------- | :----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | :------ |
| `input_type`              | string  | Required          | The type of input data. Accepted values: `query` or `passage`.                                                                                                                                                                                                                 |         |
| `max_tokens_per_sequence` | integer | Optional          | Maximum number of tokens to embed. Accepted values: `512` or `2048`.                                                                                                                                                                                                           | `512`   |
| `truncate`                | string  | Optional          | How to handle inputs longer than those supported by the model. Accepted values: `END` or `NONE`.<br /><br />`END` truncates the input sequence at the the `max_tokens_per_sequence` limit. `NONE` returns an error when the input exceeds the `max_tokens_per_sequence` limit. | `END`   |
| `return_tokens`           | boolean | Optional          | Whether to return the string tokens.                                                                                                                                                                                                                                           | `false` |

---

# Source: https://docs.pinecone.io/troubleshooting/create-and-manage-vectors-with-metadata.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create and manage vectors with metadata

Performing deletes by metadata filtering can be a very expensive process for any database. By using a hierarchical naming convention for vector IDs, you can avoid this process and perform deletes by ID. This is more efficient and will reduce the impact on the compute resources, minimize query latency, and maintain a more consistent user experience.

## 1. Upsert

* Generate a hierarchical naming convention for vector IDs.
  * One recommended pattern may be `parentId-chunkId` where parentId is the ID of the document and `chunkId` is an integer starting with 0 to the total number of chunks
  * While capturing embeddings and preparing upserts for Pinecone, capture the total number of chunks for each `parentId`.
  * Append the `chunkCount` to the metadata field of the `parentId-0` vector, or you may append them to all chunks if desired. This should be an integer and cardinality will naturally be low.
  * Upsert the vectors with the `parentId-chunkId` as the ID.
  * Reverse lookups can be created where you find a chunk and want to find the parent document or sibling chunks.

## 2. Delete by ID (to avoid delete by metadata filter)

* Identify the `parentId`
  * This could be an internal process to identify documents that have been modified or deleted.
  * Or, this could be a end-user initiated process to delete a document based on a query that finds a sibling chunk or `parentId`.

* Once the `parentId` is identified, use the [`fetch`](/reference/api/2024-10/data-plane/fetch) endpoint to retrieve the `chunkCount` from the metadata field by sending the `parentId-0` vector ID.

* Build a list of IDs using the pattern of `parentId` and `chunkCount`.

* Batch these together and send them to the [`delete`](/reference/api/2024-10/data-plane/delete) endpoint using the IDs of the vectors.

  ```shell curl theme={null}
  INDEX_NAME="docs-example"
  PROJECT_ID="example-project-id"

  curl "https://$INDEX_NAME-$PROJECT_ID.svc.environment.pinecone.io/vectors/delete" \
    -H "accept: application/json" \
    -H "content-type: application/json"\
    -H "X-Pinecone-Api-Version: 2024-07" \
    -d '
    {
      "deleteAll": "false",
      "ids": [
        "someParentDoc-0",
        "someParentDoc-1",
        "someParentDoc-2"
      ]
    }'
  ```

* You may then [upsert](/reference/api/2024-10/data-plane/upsert) the new version of the document with the new vectors and metadata or if it is a delete-only process, you are finished.

## 3. Updates

* [Updates](/reference/api/2024-10/data-plane/update) are intended to apply small changes to a record whether that means updating the vector, or more commonly, the metadata.
* In cases where you are chunking data, you are more likely going to need to delete and re-upsert using the steps above.
* If you are only performing very small changes to a small number of vectors, the update process is ideal.
* If you are updating a large number of vectors, you may want to consider batching and slowing down the updates to avoid rate limiting or affecting query latency and response times.

---

# Source: https://docs.pinecone.io/guides/assistant/create-assistant.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create an assistant

> Create and deploy a Pinecone Assistant for your knowledge base.

This page shows you how to create an [assistant](/guides/assistant/overview).

You can [create an assistant](/reference/api/latest/assistant/create_assistant), as in the following example:

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  assistant = pc.assistant.create_assistant(
      assistant_name="example-assistant", 
      instructions="Use American English for spelling and grammar.", # Description or directive for the assistant to apply to all responses.
      metadata={"team": "customer-support", "version": "1.0"}, # Optional metadata (max 16KB) for organizing assistants.
      region="us", # Region to deploy assistant. Options: "us" (default) or "eu".
      timeout=30 # Maximum seconds to wait for assistant status to become "Ready" before timing out.
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistant = await pc.createAssistant({
    name: 'example-assistant',
    instructions: 'Use American English for spelling and grammar.',
    metadata: { team: 'customer-support', version: '1.0' }, // Optional metadata (max 16KB) for organizing assistants.
    region: 'us'
  });
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://api.pinecone.io/assistant/assistants" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
    "name": "example-assistant",
    "instructions": "Use American English for spelling and grammar.",
    "metadata": {"team": "customer-support", "version": "1.0"},
    "region":"us"
  }'
  ```
</CodeGroup>

<Note>
  Instructions (maximum size 16 KB) are included in every chat API call. Longer instructions increase input token costs for each request and consume more of the LLM's context window, reducing available space for retrieved context and conversation history.
</Note>

<Tip>
  You can create an assistant using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/assistant/-/files).
</Tip>

---

# Source: https://docs.pinecone.io/reference/api/2025-10/admin/create_api_key.md

# Source: https://docs.pinecone.io/reference/api/2025-10/admin-assistant/create_api_key.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create an API key

> Create a new API key for a project. Developers can use the API key to authenticate requests to Pinecone's Data Plane and Control Plane APIs.


<RequestExample>
  ```bash curl theme={null}
  PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"
  PINECONE_PROJECT_ID="YOUR_PROJECT_ID"

  curl "https://api.pinecone.io/admin/projects/$PINECONE_PROJECT_ID/api-keys" \
  	-H "X-Pinecone-Api-Version: 2025-10" \
  	-H "Authorization: Bearer $PINECONE_ACCESS_TOKEN" \
  	-d '{
  		"name": "Example API Key",
  		"roles": ["ProjectEditor"]
  	}'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "key": {
      "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
      "name": "Example API key",
      "project_id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
      "roles": [
        "ProjectEditor"
      ]
    },
    "value": "string"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/admin_2025-10.oas.yaml post /admin/projects/{project_id}/api-keys
openapi: 3.0.3
info:
  title: Pinecone Admin API
  description: |
    Provides an API for managing a Pinecone organization and its resources.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - BearerAuth: []
tags:
  - name: API Keys
    description: Actions that manage API Keys.
  - name: Organizations
    description: Actions that manage organizations.
  - name: Projects
    description: Actions that manage projects.
paths:
  /admin/projects/{project_id}/api-keys:
    post:
      tags:
        - API Keys
      summary: Create an API key
      description: >
        Create a new API key for a project. Developers can use the API key to
        authenticate requests to Pinecone's Data Plane and Control Plane APIs.
      operationId: create_api_key
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: project_id
          description: Project ID
          required: true
          schema:
            type: string
            format: uuid
          style: simple
      requestBody:
        description: The details of the new API key.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateAPIKeyRequest'
        required: true
      responses:
        '201':
          description: API key created successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/APIKeyWithSecret'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-metric-validation-error:
                  summary: Validation error
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Bad request. The request body included invalid request
                        parameters.
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '403':
          description: Not enough available quota to complete this operation.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
        4XX:
          description: Unexpected error on request.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
components:
  schemas:
    CreateAPIKeyRequest:
      type: object
      properties:
        name:
          example: devkey
          description: |
            The name of the API key. The name must be 1-80 characters long.
          type: string
          minLength: 1
          maxLength: 80
        roles:
          description: >
            The roles to create the API key with. Default is
            `["ProjectEditor"]`.
          type: array
          items:
            example: ProjectEditor
            description: >-
              A role that can be assigned to an API key.

              Possible values: `ProjectEditor`, `ProjectViewer`,
              `ControlPlaneEditor`, `ControlPlaneViewer`, `DataPlaneEditor`, or
              `DataPlaneViewer`.
            x-enum:
              - ProjectEditor
              - ProjectViewer
              - ControlPlaneEditor
              - ControlPlaneViewer
              - DataPlaneEditor
              - DataPlaneViewer
            type: string
      required:
        - name
    APIKeyWithSecret:
      description: >
        The details of an API key, including the secret. Only returned on API
        key creation.
      type: object
      properties:
        key:
          $ref: '#/components/schemas/APIKey'
        value:
          description: >
            The value to use as an API key. New keys will have the format
            `"pckey_<public-label>_<unique-key>"`. The entire string should be
            used when authenticating.
          type: string
      required:
        - key
        - value
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, or `UNPROCESSABLE_ENTITY`.        
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    APIKey:
      description: The details of an API key, without the secret.
      type: object
      properties:
        id:
          description: The unique ID of the API key.
          type: string
          format: uuid
        name:
          description: The name of the API key.
          type: string
        project_id:
          description: The ID of the project containing the API key.
          type: string
          format: uuid
        roles:
          description: The roles assigned to the API key.
          type: array
          items:
            example: ProjectEditor
            description: >-
              A role that can be assigned to an API key.

              Possible values: `ProjectEditor`, `ProjectViewer`,
              `ControlPlaneEditor`, `ControlPlaneViewer`, `DataPlaneEditor`, or
              `DataPlaneViewer`.
            x-enum:
              - ProjectEditor
              - ProjectViewer
              - ControlPlaneEditor
              - ControlPlaneViewer
              - DataPlaneEditor
              - DataPlaneViewer
            type: string
      required:
        - id
        - name
        - project_id
        - roles
  securitySchemes:
    BearerAuth:
      type: http
      scheme: bearer
      description: >
        An [access
        token](https://docs.pinecone.io/guides/organizations/manage-service-accounts#retrieve-an-access-token)
        must be provided in the `Authorization` header using the `Bearer`
        scheme.

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/assistant/create_assistant.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create an assistant

> Create an assistant. This is where you specify the underlying training model, which cloud provider you would like to deploy with, and more.

For guidance and examples, see [Create an assistant](https://docs.pinecone.io/guides/assistant/create-assistant)

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://api.pinecone.io/assistant/assistants" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "name": "example-assistant",
    "instructions": "Use American English for spelling and grammar.",
    "metadata": {"team": "customer-support", "version": "1.0"},
    "region":"us"
  }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "name": "example-assistant",
    "instructions": "Use American English for spelling and grammar.",
    "metadata": {"team": "customer-support", "version": "1.0"},
    "status": "Initializing",
    "host": "https://prod-1-data.ke.pinecone.io",
    "created_at": "2025-10-01T12:30:00Z",
    "updated_at": "2025-10-01T12:30:00Z"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/assistant_control_2025-10.oas.yaml post /assistants
openapi: 3.0.3
info:
  title: Pinecone Assistant Control Plane API
  description: >-
    Pinecone Assistant Engine is a context engine to store and retrieve relevant
    knowledge  from millions of documents at scale. This API supports creating
    and managing assistants. 
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io/assistant
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Assistants
    description: Actions that manage Assistants
paths:
  /assistants:
    post:
      tags:
        - Manage Assistants
      summary: Create an assistant
      description: >-
        Create an assistant. This is where you specify the underlying training
        model, which cloud provider you would like to deploy with, and more.


        For guidance and examples, see [Create an
        assistant](https://docs.pinecone.io/guides/assistant/create-assistant)
      operationId: create_assistant
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        description: The desired configuration to create an assistant.
        content:
          application/json:
            schema:
              description: The configuration needed to create an assistant.
              type: object
              properties:
                name:
                  example: example-assistant
                  description: >
                    The name of the assistant. Resource name must be 1-63
                    characters long, start and end with an alphanumeric
                    character, and consist only of lower case alphanumeric
                    characters or '-'.
                  type: string
                  minLength: 1
                  maxLength: 63
                instructions:
                  nullable: true
                  description: >-
                    Description or directive for the assistant to apply to all
                    responses. Maximum 16 KB.
                  type: string
                metadata:
                  description: >-
                    Optional metadata associated with the assistant. Metadata is
                    a JSON object that can store custom organizational data,
                    tags, and attributes. Maximum size is 16KB.
                  type: object
                region:
                  description: >-
                    The region to deploy the assistant in. Our current options
                    are either us or eu. Defaults to us.
                  x-enum:
                    - us
                    - eu
                  type: string
              required:
                - name
        required: true
      responses:
        '200':
          description: Create request successful.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Assistant'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                upload-validation-error:
                  summary: Validation error on ingest
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Uploaded file can only currently be either a pdf or txt
                        file
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '429':
          description: Assistant of given name already exists.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    Assistant:
      description: Describes the configuration and status of a Pinecone Assistant.
      type: object
      properties:
        name:
          example: example-assistant
          description: >
            The name of the assistant. Resource name must be 1-63 characters
            long, start and end with an alphanumeric character, and consist only
            of lower case alphanumeric characters or '-'.
          type: string
          minLength: 1
          maxLength: 63
        instructions:
          nullable: true
          description: >-
            Description or directive for the assistant to apply to all
            responses.
          type: string
        metadata:
          description: >-
            Optional metadata associated with the assistant. Metadata is a JSON
            object that can store custom organizational data, tags, and
            attributes.
          nullable: true
          type: object
        status:
          x-enum:
            - Initializing
            - Failed
            - Ready
            - Terminating
            - InitializationFailed
          type: string
        host:
          description: The host where the assistant is deployed.
          type: string
        created_at:
          type: string
          format: date-time
        updated_at:
          type: string
          format: date-time
      required:
        - name
        - status
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: Uploaded file can only currently be either a pdf or txt file
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: Pinecone API Key

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/create_backup.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create a backup of an index

> Create a backup of an index.


<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_NAME="docs-example"

  curl "https://api.pinecone.io/indexes/$INDEX_NAME/backups" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H 'accept: application/json' \
      -H 'Content-Type: application/json' \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
        "name": "example-backup", 
        "description": "Monthly backup of production index"
        }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "backup_id":"8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
    "source_index_id":"f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
    "source_index_name":"docs-example",
    "tags":{},
    "name":"example-backup",
    "description":"Monthly backup of production index",
    "status":"Ready",
    "cloud":"aws",
    "region":"us-east-1",
    "dimension":1024,
    "record_count":96,
    "namespace_count":3,
    "size_bytes":1069169,
    "created_at":"2025-05-14T16:37:25.625540Z"
    }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml post /indexes/{index_name}/backups
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /indexes/{index_name}/backups:
    post:
      tags:
        - Manage Indexes
      summary: Create a backup of an index
      description: |
        Create a backup of an index.
      operationId: create_backup
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: index_name
          description: Name of the index to backup
          required: true
          schema:
            type: string
          style: simple
      requestBody:
        description: The desired configuration for the backup.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateBackupRequest'
            examples:
              backup-index-no-name:
                summary: Creating a backup of an index with no name
                value: {}
              backup-index-with-name:
                summary: Creating a backup of an index with a name
                value:
                  name: backup-index
              backup-index-with-name-and-description:
                summary: Creating a backup of an index with a name and description
                value:
                  description: Backup of the index
                  name: backup-index
        required: true
      responses:
        '201':
          description: The backup has been successfully created.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/BackupModel'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-metric-validation-error:
                  summary: Validation error
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Bad request. The request body included invalid request
                        parameters.
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '402':
          description: >-
            Payment required. Organization is on a paid plan and is delinquent
            on payment.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                payment-required:
                  summary: Payment required
                  value:
                    error:
                      code: PAYMENT_REQUIRED
                      message: >-
                        Request failed. Pay all past due invoices to lift
                        restrictions on your account.
                    status: 402
        '403':
          description: You've exceed your backup quota.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Forbidden
                  value:
                    error:
                      code: FORBIDDEN
                      message: Increase your quota or upgrade to create more backups.
                    status: 403
        '422':
          description: Unprocessable entity. The request body could not be deserialized.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                missing-field:
                  summary: Unprocessable entity
                  value:
                    error:
                      code: UNPROCESSABLE_ENTITY
                      message: >-
                        Failed to deserialize the JSON body into the target
                        type: missing field `metric` at line 1 column 16
                    status: 422
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    CreateBackupRequest:
      description: The configuration needed to create a backup of an index.
      type: object
      properties:
        name:
          description: The name of the backup.
          type: string
        description:
          description: A description of the backup.
          type: string
    BackupModel:
      description: >-
        The BackupModel describes the configuration and status of a Pinecone
        backup.
      type: object
      properties:
        backup_id:
          example: 670e8400-e29b-41d4-a716-446655440001
          description: Unique identifier for the backup.
          type: string
        source_index_name:
          example: my-index
          description: Name of the index from which the backup was taken.
          type: string
        source_index_id:
          example: 670e8400-e29b-41d4-a716-446655440000
          description: ID of the index.
          type: string
        name:
          example: backup-2025-02-04
          description: Optional user-defined name for the backup.
          type: string
        description:
          example: Backup before bulk update.
          description: Optional description providing context for the backup.
          type: string
        status:
          example: Ready
          description: Current status of the backup (e.g., Initializing, Ready, Failed).
          type: string
        cloud:
          example: aws
          description: Cloud provider where the backup is stored.
          type: string
        region:
          example: us-east-1
          description: Cloud region where the backup is stored.
          type: string
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If the 'vector_type' is
            'sparse', the metric must be 'dotproduct'. If the `vector_type` is
            `dense`, the metric defaults to 'cosine'.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        schema:
          $ref: '#/components/schemas/MetadataSchema'
        record_count:
          example: 120000
          description: Total number of records in the backup.
          type: integer
        namespace_count:
          example: 3
          description: Number of namespaces in the backup.
          type: integer
        size_bytes:
          example: 10000000
          description: Size of the backup in bytes.
          type: integer
        tags:
          $ref: '#/components/schemas/IndexTags'
        created_at:
          description: Timestamp when the backup was created.
          type: string
      required:
        - backup_id
        - source_index_name
        - source_index_id
        - status
        - cloud
        - region
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    MetadataSchema:
      example:
        fields:
          description:
            filterable: true
          genre:
            filterable: true
          year:
            filterable: true
      description: >-
        Schema for the behavior of Pinecone's internal metadata index. By
        default, all metadata is indexed; when `schema` is present, only fields
        which are present in the `fields` object with a `filterable: true` are
        indexed. Note that `filterable: false` is not currently supported.
      type: object
      properties:
        fields:
          description: >-
            A map of metadata field names to their configuration. The field name
            must be a valid metadata field name. The field name must be unique.
          type: object
          additionalProperties:
            type: object
            properties:
              filterable:
                description: >-
                  Whether the field is filterable. If true, the field is indexed
                  and can be used in filters. Only true values are allowed.
                type: boolean
      required:
        - fields
    IndexTags:
      example:
        tag0: val0
        tag1: val1
      description: >-
        Custom user tags added to an index. Keys must be 80 characters or less.
        Values must be 120 characters or less. Keys must be alphanumeric, '_',
        or '-'.  Values must be alphanumeric, ';', '@', '_', '-', '.', '+', or '
        '. To unset a key, set the value to be an empty string.
      type: object
      additionalProperties:
        type: string
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/create_collection.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create a collection

> Create a Pinecone collection.
  
Serverless indexes do not support collections.


<RequestExample>
  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s "https://api.pinecone.io/collections" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "name": "example-collection",
          "source": "docs-example"
    }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
      "name": "example-collection",
      "status": "Initializing",
      "environment": "us-east-1-aws",
      "dimension": 1536
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml post /collections
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /collections:
    post:
      tags:
        - Manage Indexes
      summary: Create a collection
      description: |
        Create a Pinecone collection.
          
        Serverless indexes do not support collections.
      operationId: create_collection
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        description: The desired configuration for the collection.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateCollectionRequest'
            examples:
              creating-collection:
                summary: Creating a collection
                value:
                  name: example-collection
                  source: example-source-index
        required: true
      responses:
        '201':
          description: The collection has been successfully created.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/CollectionModel'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-metric-validation-error:
                  summary: Validation error
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Bad request. The request body included invalid request
                        parameters.
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '402':
          description: >-
            Payment required. Organization is on a paid plan and is delinquent
            on payment.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                payment-required:
                  summary: Payment required
                  value:
                    error:
                      code: PAYMENT_REQUIRED
                      message: >-
                        Request failed. Pay all past due invoices to lift
                        restrictions on your account.
                    status: 402
        '403':
          description: You've exceed your collections quota.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Forbidden
                  value:
                    error:
                      code: FORBIDDEN
                      message: >-
                        Collection exceeds quota. Maximum allowed on your
                        account is 1. Currently have 1.
                    status: 403
        '409':
          description: Collection of given name already exists.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                collection-name-already-exists:
                  summary: Collection name needs to be unique.
                  value:
                    error:
                      code: ALREADY_EXISTS
                      message: Resource already exists.
                    status: 409
        '422':
          description: Unprocessable entity. The request body could not be deserialized.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                missing-field:
                  summary: Unprocessable entity
                  value:
                    error:
                      code: UNPROCESSABLE_ENTITY
                      message: >-
                        Failed to deserialize the JSON body into the target
                        type: missing field `metric` at line 1 column 16
                    status: 422
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    CreateCollectionRequest:
      description: The configuration needed to create a Pinecone collection.
      type: object
      properties:
        name:
          description: >
            The name of the collection to be created. Resource name must be 1-45
            characters long, start and end with an alphanumeric character, and
            consist only of lower case alphanumeric characters or '-'.
          type: string
          minLength: 1
          maxLength: 45
        source:
          example: example-source-index
          description: The name of the index to be used as the source for the collection.
          type: string
      required:
        - name
        - source
    CollectionModel:
      description: >-
        The CollectionModel describes the configuration and status of a Pinecone
        collection.
      type: object
      properties:
        name:
          example: example-collection
          description: The name of the collection.
          type: string
        size:
          example: 10000000
          description: The size of the collection in bytes.
          type: integer
          format: int64
        status:
          example: Initializing
          description: |-
            The status of the collection.
            Possible values: `Initializing`, `Ready`, or `Terminating`.
          x-enum:
            - Initializing
            - Ready
            - Terminating
          type: string
        dimension:
          example: 1536
          description: >-
            The dimension of the vectors stored in each record held in the
            collection.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        vector_count:
          example: 120000
          description: The number of records stored in the collection.
          type: integer
          format: int32
        environment:
          example: us-east1-gcp
          description: The environment where the collection is hosted.
          type: string
      required:
        - name
        - status
        - environment
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/create_for_model.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create an index with integrated embedding

> Create an index with integrated embedding.
With this type of index, you provide source text, and  Pinecone uses a [hosted embedding model](https://docs.pinecone.io/guides/index-data/create-an-index#embedding-models)  to convert the text automatically during [upsert](https://docs.pinecone.io/reference/api/2025-10/data-plane/upsert_records)  and [search](https://docs.pinecone.io/reference/api/2025-10/data-plane/search_records).  
For guidance and examples, see [Create an index](https://docs.pinecone.io/guides/index-data/create-an-index#integrated-embedding).

<RequestExample>
  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl https://api.pinecone.io/indexes/create-for-model \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "name": "integrated-dense-curl",
          "cloud": "aws",
          "region": "us-east-1",
          "embed": {
            "model": "llama-text-embed-v2",
            "metric": "cosine",
            "field_map": {
              "text": "chunk_text"
            },
            "write_parameters": {
              "input_type": "passage",
              "truncate": "END"
            },
            "read_parameters": {
              "input_type": "query",
              "truncate": "END"
            }
          }
        }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "id": "9dabb7cb-ec0a-4e2e-b79e-c7c997e592ce",
    "name": "integrated-dense-curl",
    "metric": "cosine",
    "dimension": 1024,
    "status": {
      "ready": false,
      "state": "Initializing"
    },
    "host": "integrated-dense-curl-govk0nt.svc.aped-4627-b74a.pinecone.io",
    "spec": {
      "serverless": {
        "region": "us-east-1",
        "cloud": "aws"
      }
    },
    "deletion_protection": "disabled",
    "tags": null,
    "embed": {
      "model": "llama-text-embed-v2",
      "field_map": {
        "text": "chunk_text"
      },
      "dimension": 1024,
      "metric": "cosine",
      "write_parameters": {
        "input_type": "passage",
        "truncate": "END"
      },
      "read_parameters": {
        "input_type": "query",
        "truncate": "END"
      }
    }
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml post /indexes/create-for-model
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /indexes/create-for-model:
    post:
      tags:
        - Manage Indexes
      summary: Create an index with integrated embedding
      description: >-
        Create an index with integrated embedding.

        With this type of index, you provide source text, and  Pinecone uses a
        [hosted embedding
        model](https://docs.pinecone.io/guides/index-data/create-an-index#embedding-models) 
        to convert the text automatically during
        [upsert](https://docs.pinecone.io/reference/api/2025-10/data-plane/upsert_records) 
        and
        [search](https://docs.pinecone.io/reference/api/2025-10/data-plane/search_records).  

        For guidance and examples, see [Create an
        index](https://docs.pinecone.io/guides/index-data/create-an-index#integrated-embedding).
      operationId: create_index_for_model
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        description: >-
          The desired configuration for the index and associated embedding
          model.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateIndexForModelRequest'
            examples:
              index-for-model:
                summary: Creating a serverless index for the specified model
                value:
                  cloud: gcp
                  deletion_protection: enabled
                  embed:
                    field_map:
                      text: your-text-field
                    metric: cosine
                    model: multilingual-e5-large
                  name: multilingual-e5-large-index
                  region: us-east1
        required: true
      responses:
        '201':
          description: The index has successfully been created for the embedding model.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/IndexModel'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-metric-validation-error:
                  summary: Validation error
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Bad request. The request body included invalid request
                        parameters.
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Unknown cloud or region when creating a serverless index.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                serverless-spec-cloud-not-found:
                  summary: Cannot create serverless index with invalid spec.
                  value:
                    error:
                      code: NOT_FOUND
                      message: 'Resource cloud: aws region: us-west1 not found.'
                    status: 404
        '409':
          description: Index of given name already exists.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-name-already-exists:
                  summary: Index name needs to be unique
                  value:
                    error:
                      code: ALREADY_EXISTS
                      message: Resource already exists.
                    status: 409
        '422':
          description: Unprocessable entity. The request body could not be deserialized.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                missing-field:
                  summary: Unprocessable entity
                  value:
                    error:
                      code: UNPROCESSABLE_ENTITY
                      message: >-
                        Failed to deserialize the JSON body into the target
                        type: missing field `metric` at line 1 column 16
                    status: 422
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    CreateIndexForModelRequest:
      description: The desired configuration for the index and associated embedding model.
      type: object
      properties:
        name:
          example: example-index
          description: >
            The name of the index. Resource name must be 1-45 characters long,
            start and end with an alphanumeric character, and consist only of
            lower case alphanumeric characters or '-'.
          type: string
          minLength: 1
          maxLength: 45
        cloud:
          example: aws
          description: |-
            The public cloud where you would like your index hosted.
            Possible values: `gcp`, `aws`, or `azure`.
          x-enum:
            - gcp
            - aws
            - azure
          type: string
        region:
          example: us-east-1
          description: The region where you would like your index to be created.
          type: string
        deletion_protection:
          $ref: '#/components/schemas/DeletionProtection'
        tags:
          $ref: '#/components/schemas/IndexTags'
        schema:
          $ref: '#/components/schemas/MetadataSchema'
        read_capacity:
          $ref: '#/components/schemas/ReadCapacity'
        embed:
          example:
            field_map:
              text: your-text-field
            metric: cosine
            model: multilingual-e5-large
            read_parameters:
              input_type: query
              truncate: NONE
            write_parameters:
              input_type: passage
          description: >-
            Specify the integrated inference embedding configuration for the
            index.


            Once set the model cannot be changed, but you can later update the
            embedding configuration for an integrated inference index including
            field map, read parameters, or write parameters.


            Refer to the [model
            guide](https://docs.pinecone.io/guides/index-data/create-an-index#embedding-models)
            for available models and model details.
          type: object
          properties:
            model:
              example: multilingual-e5-large
              description: The name of the embedding model to use for the index.
              type: string
            metric:
              description: >-
                The distance metric to be used for similarity search. You can
                use 'euclidean', 'cosine', or 'dotproduct'. If not specified,
                the metric will be defaulted according to the model. Cannot be
                updated once set.

                Possible values: `cosine`, `euclidean`, or `dotproduct`.
              x-enum:
                - cosine
                - euclidean
                - dotproduct
              type: string
            field_map:
              example:
                text: your-text-field
              description: >-
                Identifies the name of the text field from your document model
                that will be embedded.
              type: object
            dimension:
              description: The dimension of embedding vectors produced for the index.
              type: integer
            read_parameters:
              description: The read parameters for the embedding model.
              type: object
            write_parameters:
              description: The write parameters for the embedding model.
              type: object
          required:
            - model
            - field_map
      required:
        - name
        - cloud
        - region
        - embed
    IndexModel:
      description: >-
        The IndexModel describes the configuration and status of a Pinecone
        index.
      type: object
      properties:
        name:
          example: example-index
          description: >
            The name of the index. Resource name must be 1-45 characters long,
            start and end with an alphanumeric character, and consist only of
            lower case alphanumeric characters or '-'.
          type: string
          minLength: 1
          maxLength: 45
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If the 'vector_type' is
            'sparse', the metric must be 'dotproduct'. If the `vector_type` is
            `dense`, the metric defaults to 'cosine'.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        host:
          example: semantic-search-c01b5b5.svc.us-west1-gcp.pinecone.io
          description: The URL address where the index is hosted.
          type: string
        private_host:
          example: semantic-search-c01b5b5.svc.private.us-west1-gcp.pinecone.io
          description: The private endpoint URL of an index.
          type: string
        deletion_protection:
          $ref: '#/components/schemas/DeletionProtection'
        tags:
          $ref: '#/components/schemas/IndexTags'
        embed:
          $ref: '#/components/schemas/ModelIndexEmbed'
        spec:
          example:
            pod:
              environment: us-east-1-aws
              metadata_config:
                indexed:
                  - genre
                  - title
                  - imdb_rating
              pod_type: p1.x1
              pods: 1
              replicas: 1
              shards: 1
          description: The spec object defines how the index should be deployed.
          type: object
          oneOf:
            - title: Serverless
              type: object
              properties:
                serverless:
                  $ref: '#/components/schemas/ServerlessSpecResponse'
              required:
                - serverless
              additionalProperties: false
            - title: Pod-based
              type: object
              properties:
                pod:
                  $ref: '#/components/schemas/PodSpec'
              required:
                - pod
              additionalProperties: false
            - title: BYOC
              type: object
              properties:
                byoc:
                  $ref: '#/components/schemas/ByocSpecResponse'
              required:
                - byoc
              additionalProperties: false
        status:
          example:
            ready: true
            state: ScalingUpPodSize
          description: The current status of the index
          type: object
          properties:
            ready:
              description: Whether the index is ready for use
              type: boolean
            state:
              description: >-
                The state of the index.

                Possible values: `Initializing`, `InitializationFailed`,
                `ScalingUp`, `ScalingDown`, `ScalingUpPodSize`,
                `ScalingDownPodSize`, `Terminating`, `Ready`, or `Disabled`.
              x-enum:
                - Initializing
                - InitializationFailed
                - ScalingUp
                - ScalingDown
                - ScalingUpPodSize
                - ScalingDownPodSize
                - Terminating
                - Ready
                - Disabled
              type: string
          required:
            - ready
            - state
        vector_type:
          description: >-
            The index vector type. You can use 'dense' or 'sparse'. If 'dense',
            the vector dimension must be specified.  If 'sparse', the vector
            dimension should not be specified.
          default: dense
          type: string
      required:
        - name
        - metric
        - status
        - spec
        - host
        - vector_type
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    DeletionProtection:
      description: >-
        Whether [deletion
        protection](http://docs.pinecone.io/guides/manage-data/manage-indexes#configure-deletion-protection)
        is enabled/disabled for the index.

        Possible values: `disabled` or `enabled`.
      default: disabled
      x-enum:
        - disabled
        - enabled
      type: string
    IndexTags:
      example:
        tag0: val0
        tag1: val1
      description: >-
        Custom user tags added to an index. Keys must be 80 characters or less.
        Values must be 120 characters or less. Keys must be alphanumeric, '_',
        or '-'.  Values must be alphanumeric, ';', '@', '_', '-', '.', '+', or '
        '. To unset a key, set the value to be an empty string.
      type: object
      additionalProperties:
        type: string
    MetadataSchema:
      example:
        fields:
          description:
            filterable: true
          genre:
            filterable: true
          year:
            filterable: true
      description: >-
        Schema for the behavior of Pinecone's internal metadata index. By
        default, all metadata is indexed; when `schema` is present, only fields
        which are present in the `fields` object with a `filterable: true` are
        indexed. Note that `filterable: false` is not currently supported.
      type: object
      properties:
        fields:
          description: >-
            A map of metadata field names to their configuration. The field name
            must be a valid metadata field name. The field name must be unique.
          type: object
          additionalProperties:
            type: object
            properties:
              filterable:
                description: >-
                  Whether the field is filterable. If true, the field is indexed
                  and can be used in filters. Only true values are allowed.
                type: boolean
      required:
        - fields
    ReadCapacity:
      description: >-
        By default the index will be created with read capacity  mode
        `OnDemand`. If you prefer to allocate dedicated read  nodes for your
        workload, you must specify mode `Dedicated` and additional
        configurations for `node_type` and `scaling`.
      discriminator:
        propertyName: mode
        mapping:
          OnDemand: '#/components/schemas/ReadCapacityOnDemandSpec'
          Dedicated: '#/components/schemas/ReadCapacityDedicatedSpec'
      oneOf:
        - $ref: '#/components/schemas/ReadCapacityOnDemandSpec'
        - $ref: '#/components/schemas/ReadCapacityDedicatedSpec'
    ModelIndexEmbed:
      example:
        field_map:
          text: your-text-field
        metric: cosine
        model: multilingual-e5-large
        read_parameters:
          input_type: query
          truncate: NONE
        write_parameters:
          input_type: passage
      description: The embedding model and document fields mapped to embedding inputs.
      type: object
      properties:
        model:
          example: multilingual-e5-large
          description: The name of the embedding model used to create the index.
          type: string
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If not specified, the metric
            will be defaulted according to the model. Cannot be updated once
            set.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        vector_type:
          description: >-
            The index vector type. You can use 'dense' or 'sparse'. If 'dense',
            the vector dimension must be specified.  If 'sparse', the vector
            dimension should not be specified.
          default: dense
          type: string
        field_map:
          example:
            text: your-text-field
          description: >-
            Identifies the name of the text field from your document model that
            is embedded.
          type: object
        read_parameters:
          description: The read parameters for the embedding model.
          type: object
        write_parameters:
          description: The write parameters for the embedding model.
          type: object
      required:
        - model
    ServerlessSpecResponse:
      description: Configuration of a serverless index.
      type: object
      properties:
        cloud:
          example: aws
          description: |-
            The public cloud where you would like your index hosted.
            Possible values: `gcp`, `aws`, or `azure`.
          x-enum:
            - gcp
            - aws
            - azure
          type: string
        region:
          example: us-east-1
          description: The region where you would like your index to be created.
          type: string
        read_capacity:
          $ref: '#/components/schemas/ReadCapacityResponse'
        source_collection:
          example: movie-embeddings
          description: The name of the collection to be used as the source for the index.
          type: string
        schema:
          $ref: '#/components/schemas/MetadataSchema'
      required:
        - cloud
        - region
        - read_capacity
    PodSpec:
      example:
        environment: us-east1-gcp
        metadata_config:
          indexed:
            - genre
            - title
            - imdb_rating
        pod_type: p1.x1
        pods: 1
        replicas: 1
        shards: 1
        source_collection: movie-embeddings
      title: Pod-based
      description: Configuration needed to deploy a pod-based index.
      type: object
      properties:
        environment:
          example: us-east1-gcp
          description: The environment where the index is hosted.
          type: string
        replicas:
          description: >-
            The number of replicas. Replicas duplicate your index. They provide
            higher availability and throughput. Replicas can be scaled up or
            down as your needs change.
          default: 1
          type: integer
          format: int32
          minimum: 1
        shards:
          description: >-
            The number of shards. Shards split your data across multiple pods so
            you can fit more data into an index.
          default: 1
          type: integer
          format: int32
          minimum: 1
        pod_type:
          description: >-
            The type of pod to use. One of `s1`, `p1`, or `p2` appended with `.`
            and one of `x1`, `x2`, `x4`, or `x8`.
          default: p1.x1
          type: string
        pods:
          example: 1
          description: >-
            The number of pods to be used in the index. This should be equal to
            `shards` x `replicas`.'
          default: 1
          type: integer
          minimum: 1
        metadata_config:
          example:
            indexed:
              - genre
              - title
              - imdb_rating
          description: >-
            Configuration for the behavior of Pinecone's internal metadata
            index. By default, all metadata is indexed; when `metadata_config`
            is present, only specified metadata fields are indexed. These
            configurations are only valid for use with pod-based indexes.
          type: object
          properties:
            indexed:
              description: >-
                By default, all metadata is indexed; to change this behavior,
                use this property to specify an array of metadata fields that
                should be indexed.
              type: array
              items:
                type: string
        source_collection:
          example: movie-embeddings
          description: The name of the collection to be used as the source for the index.
          type: string
      required:
        - environment
        - pod_type
    ByocSpecResponse:
      title: BYOC
      description: Configuration of a BYOC index.
      type: object
      properties:
        environment:
          example: aws-us-east-1-b921
          description: The environment where the index is hosted.
          type: string
        read_capacity:
          $ref: '#/components/schemas/ReadCapacityResponse'
        schema:
          $ref: '#/components/schemas/MetadataSchema'
      required:
        - environment
        - read_capacity
    ReadCapacityOnDemandSpec:
      example:
        mode: OnDemand
      title: On-demand
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
      required:
        - mode
      additionalProperties: false
    ReadCapacityDedicatedSpec:
      example:
        dedicated:
          manual:
            replicas: 1
            shards: 1
          node_type: t1
          scaling: Manual
        mode: Dedicated
      title: Dedicated
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
        dedicated:
          $ref: '#/components/schemas/ReadCapacityDedicatedConfig'
      required:
        - mode
        - dedicated
      additionalProperties: true
    ReadCapacityResponse:
      description: Response containing read capacity configuration
      discriminator:
        propertyName: mode
        mapping:
          OnDemand: '#/components/schemas/ReadCapacityOnDemandSpecResponse'
          Dedicated: '#/components/schemas/ReadCapacityDedicatedSpecResponse'
      oneOf:
        - $ref: '#/components/schemas/ReadCapacityOnDemandSpecResponse'
        - $ref: '#/components/schemas/ReadCapacityDedicatedSpecResponse'
    ReadCapacityDedicatedConfig:
      description: >-
        Configuration for dedicated read capacity. See  [this
        guide](https://docs.pinecone.io/guides/index-data/dedicated-read-nodes)
        for more details on  how to configure dedicated read capacity.
      type: object
      properties:
        node_type:
          description: >-
            The type of machines to use. Available options: `b1` and `t1`. `t1`
            includes increased processing power and memory.
          type: string
        scaling:
          description: The type of scaling strategy to use.
          type: string
        manual:
          $ref: '#/components/schemas/ScalingConfigManual'
    ReadCapacityOnDemandSpecResponse:
      example:
        mode: OnDemand
        status:
          state: Ready
      title: On-demand
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
        status:
          $ref: '#/components/schemas/ReadCapacityStatus'
      required:
        - mode
        - status
      additionalProperties: false
    ReadCapacityDedicatedSpecResponse:
      example:
        dedicated:
          manual:
            replicas: 2
            shards: 2
          node_type: t1
          scaling: Manual
        mode: Dedicated
        status:
          current_replicas: 2
          current_shards: 2
          state: Ready
      title: Dedicated
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
        dedicated:
          $ref: '#/components/schemas/ReadCapacityDedicatedConfig'
        status:
          $ref: '#/components/schemas/ReadCapacityStatus'
      required:
        - mode
        - dedicated
        - status
      additionalProperties: false
    ScalingConfigManual:
      description: The config to use for manual read capacity scaling.
      type: object
      properties:
        replicas:
          description: >-
            The number of replicas to use. Replicas duplicate the compute
            resources and data of an index, allowing higher query throughput and
            availability. Setting replicas to 0 disables the index but can be
            used to reduce costs while usage is paused.
          type: integer
          format: int32
          minimum: 0
        shards:
          description: >-
            The number of shards to use. Shards determine the storage capacity
            of an index, with each shard providing 250 GB of storage.
          type: integer
          format: int32
          minimum: 1
    ReadCapacityStatus:
      description: >-
        The current status of factors affecting the read capacity of a
        serverless index
      type: object
      properties:
        state:
          description: >-
            The `state` describes the overall status of factors relating to the
            read capacity of an index. 


            Available values:

            - `Ready` is the state most of the time

            - `Scaling` if the number of replicas or shards has been recently
            updated by calling the [configure index
            endpoint](https://docs.pinecone.io/reference/api/2025-10/control-plane/configure_index)

            - `Migrating` if the index is being migrated to a new `node_type`

            - `Error` if there is an error with the read capacity configuration.
            In that case, see `error_message` for more details.
          type: string
        current_replicas:
          description: >-
            The number of replicas. Each replica has dedicated  compute
            resources and data storage. Increasing this number  will increase
            the total throughput of the index.
          type: integer
          format: int32
        current_shards:
          description: >-
            The number of shards. Each shard has dedicated storage.  Increasing
            shards alleiviates index fullness. 
          type: integer
          format: int32
        error_message:
          description: >-
            An optional error message indicating any issues with your read
            capacity configuration
          type: string
      required:
        - state
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/create_index.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create an index

> Create a Pinecone index. This is where you specify the measure of similarity, the dimension of vectors to be stored in the index, which cloud provider you would like to deploy with, and more.
  
For guidance and examples, see [Create an index](https://docs.pinecone.io/guides/index-data/create-an-index).


<RequestExample>
  ```shell curl theme={null}
  # EXAMPLE REQUEST 1: Serverless index (on-demand)
  PINECONE_API_KEY="YOUR_API_KEY"
  curl -s "https://api.pinecone.io/indexes" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "name": "example-serverless-index",
          "vector_type": "dense",
          "dimension": 1536,
          "metric": "cosine",
          "spec": {
              "serverless": {
                  "cloud": "aws",
                  "region": "us-east-1"
              }
          },
          "tags": {
              "tag0": "value0"
          },
          "deletion_protection": "disabled"
        }'

  # EXAMPLE REQUEST 2: Serverless index (dedicated)
  PINECONE_API_KEY="YOUR_API_KEY"
  curl "https://api.pinecone.io/indexes" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
  		"name": "example-serverless-dedicated-index",
  		"dimension": 1536,
  		"metric": "cosine",
  		"deletion_protection": "enabled",
  		"tags": {
  			"tag0": "value0"
  		},
  		"vector_type": "dense",
  		"spec": {
  			"serverless": {
  				"cloud": "aws",
  				"region": "us-east-1",
  				"read_capacity": {
  					"mode": "Dedicated",
  					"dedicated": {
  						"node_type": "b1",
  						"scaling": "Manual",
  						"manual": {
  							"shards": 2,
  							"replicas": 1
  						}
  					}
  				}
  			}
  		}
  	}'

  # EXAMPLE REQUEST 3: BYOC index
  PINECONE_API_KEY="YOUR_API_KEY"
  curl -s "https://api.pinecone.io/indexes" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "name": "example-byoc-index",
          "vector_type": "dense",
          "dimension": 1536,
          "metric": "cosine",
          "spec": {
              "byoc": {
                  "environment": "aws-us-east-1-b921"
              }
          },
          "tags": {
              "tag0": "value0"
          },
          "deletion_protection": "disabled"
        }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  // EXAMPLE RESPONSE 1: Serverless index (on-demand)
  {
    "name": "example-serverless-ondemand-index",
    "vector_type": "dense",
    "metric": "cosine",
    "dimension": 1536,
    "status": {
      "ready": false,
      "state": "Initializing"
    },
    "host": "example-serverless-ondemand-index-bhnyigt.svc.aped-4627-b74a.pinecone.io",
    "spec": {
      "serverless": {
        "region": "us-east-1",
        "cloud": "aws",
        "read_capacity": {
          "mode": "OnDemand",
          "status": {
            "state": "Ready",
            "current_shards": null,
            "current_replicas": null
          }
        }
      }
    },
    "deletion_protection": "disabled",
    "tags": {
      "tag0": "value0"
    }
  }

  // EXAMPLE RESPONSE 2: Serverless index (dedicated)
  {
  	"name": "example-serverless-dedicated-index",
  	"vector_type": "dense",
  	"metric": "cosine",
  	"dimension": 1536,
  	"status": {
  		"ready": false,
  		"state": "Initializing"
  	},
  	"host": "example-serverless-dedicated-index-bhnyigt.svc.aped-4627-b74a.pinecone.io",
  	"spec": {
  		"serverless": {
  			"region": "us-east-1",
  			"cloud": "aws",
  			"read_capacity": {
  				"mode": "Dedicated",
  				"dedicated": {
  					"node_type": "b1",
  					"scaling": "Manual",
  					"manual": {
  						"shards": 2,
  						"replicas": 1
  					}
  				},
  				"status": {
  					"state": "Migrating",
  					"current_shards": null,
  					"current_replicas": null
  				}
  			}
  		}
  	},
  	"deletion_protection": "enabled",
  	"tags": {
  		"tag0": "value0"
  	}
  }

  // EXAMPLE RESPONSE 3: BYOC index
  {
      "name": "example-byoc-index",
      "vector_type": "dense",
      "metric": "cosine",
      "dimension": 1536,
      "status": {
          "ready": true,
          "state": "Ready"
      },
      "host": "example-byoc-index-govk0nt.svc.private.aped-4627-b74a.pinecone.io",
      "spec": {
          "byoc": {
              "environment": "aws-us-east-1-b921"
          }
      },
      "deletion_protection": "disabled",
      "tags": {
          "tag0": "value0"
      }
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml post /indexes
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /indexes:
    post:
      tags:
        - Manage Indexes
      summary: Create an index
      description: >
        Create a Pinecone index. This is where you specify the measure of
        similarity, the dimension of vectors to be stored in the index, which
        cloud provider you would like to deploy with, and more.
          
        For guidance and examples, see [Create an
        index](https://docs.pinecone.io/guides/index-data/create-an-index).
      operationId: create_index
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        description: The desired configuration for the index.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateIndexRequest'
            examples:
              serverless-index:
                summary: Creating a serverless index
                value:
                  deletion_protection: enabled
                  dimension: 1536
                  metric: cosine
                  name: movie-recommendations
                  spec:
                    serverless:
                      cloud: gcp
                      read_capacity:
                        mode: OnDemand
                      region: us-east1
                      source_collection: movie-embeddings
              serverless-sparse-index:
                summary: Creating a sparse serverless index
                value:
                  deletion_protection: enabled
                  metric: dotproduct
                  name: sparse-index
                  spec:
                    serverless:
                      cloud: gcp
                      read_capacity:
                        mode: OnDemand
                      region: us-east1
                  vector_type: sparse
              serverless-dedicated-index:
                summary: Creating an index with dedicated capacity
                value:
                  deletion_protection: enabled
                  metric: cosine
                  name: dedicated-index
                  spec:
                    serverless:
                      cloud: gcp
                      read_capacity:
                        dedicated:
                          manual:
                            replicas: 3
                            shards: 2
                          node_type: b1
                          scaling: Manual
                        mode: Dedicated
                      region: us-east1
              pod-index:
                summary: Creating a pod-based index
                value:
                  deletion_protection: enabled
                  dimension: 1536
                  metric: cosine
                  name: movie-recommendations
                  spec:
                    pod:
                      environment: us-east-1-aws
                      metadata_config:
                        indexed:
                          - genre
                          - title
                          - imdb_rating
                      pod_type: p1.x1
                      pods: 1
                      replicas: 1
                      shards: 1
                      source_collection: movie-embeddings
        required: true
      responses:
        '201':
          description: The index has been successfully created.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/IndexModel'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-metric-validation-error:
                  summary: Validation error
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Bad request. The request body included invalid request
                        parameters.
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '402':
          description: >-
            Payment required. Organization is on a paid plan and is delinquent
            on payment.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                payment-required:
                  summary: Payment required
                  value:
                    error:
                      code: PAYMENT_REQUIRED
                      message: >-
                        Request failed. Pay all past due invoices to lift
                        restrictions on your account.
                    status: 402
        '403':
          description: You've exceed your pod quota.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Forbidden
                  value:
                    error:
                      code: FORBIDDEN
                      message: Increase your quota or upgrade to create more indexes.
                    status: 403
        '404':
          description: Unknown cloud or region when creating a serverless index.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                serverless-spec-cloud-not-found:
                  summary: Cannot create serverless index with invalid spec.
                  value:
                    error:
                      code: NOT_FOUND
                      message: 'Resource cloud: aws region: us-west1 not found.'
                    status: 404
        '409':
          description: Index of given name already exists.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-name-already-exists:
                  summary: Index name needs to be unique.
                  value:
                    error:
                      code: ALREADY_EXISTS
                      message: Resource already exists.
                    status: 409
        '422':
          description: Unprocessable entity. The request body could not be deserialized.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                missing-field:
                  summary: Unprocessable entity
                  value:
                    error:
                      code: UNPROCESSABLE_ENTITY
                      message: >-
                        Failed to deserialize the JSON body into the target
                        type: missing field `metric` at line 1 column 16
                    status: 422
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    CreateIndexRequest:
      description: The configuration needed to create a Pinecone index.
      type: object
      properties:
        name:
          example: example-index
          description: >
            The name of the index. Resource name must be 1-45 characters long,
            start and end with an alphanumeric character, and consist only of
            lower case alphanumeric characters or '-'.
          type: string
          minLength: 1
          maxLength: 45
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If the 'vector_type' is
            'sparse', the metric must be 'dotproduct'. If the `vector_type` is
            `dense`, the metric defaults to 'cosine'.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        deletion_protection:
          $ref: '#/components/schemas/DeletionProtection'
        tags:
          $ref: '#/components/schemas/IndexTags'
        spec:
          $ref: '#/components/schemas/IndexSpec'
        vector_type:
          description: >-
            The index vector type. You can use 'dense' or 'sparse'. If 'dense',
            the vector dimension must be specified.  If 'sparse', the vector
            dimension should not be specified.
          default: dense
          type: string
      required:
        - name
        - spec
    IndexModel:
      description: >-
        The IndexModel describes the configuration and status of a Pinecone
        index.
      type: object
      properties:
        name:
          example: example-index
          description: >
            The name of the index. Resource name must be 1-45 characters long,
            start and end with an alphanumeric character, and consist only of
            lower case alphanumeric characters or '-'.
          type: string
          minLength: 1
          maxLength: 45
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If the 'vector_type' is
            'sparse', the metric must be 'dotproduct'. If the `vector_type` is
            `dense`, the metric defaults to 'cosine'.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        host:
          example: semantic-search-c01b5b5.svc.us-west1-gcp.pinecone.io
          description: The URL address where the index is hosted.
          type: string
        private_host:
          example: semantic-search-c01b5b5.svc.private.us-west1-gcp.pinecone.io
          description: The private endpoint URL of an index.
          type: string
        deletion_protection:
          $ref: '#/components/schemas/DeletionProtection'
        tags:
          $ref: '#/components/schemas/IndexTags'
        embed:
          $ref: '#/components/schemas/ModelIndexEmbed'
        spec:
          example:
            pod:
              environment: us-east-1-aws
              metadata_config:
                indexed:
                  - genre
                  - title
                  - imdb_rating
              pod_type: p1.x1
              pods: 1
              replicas: 1
              shards: 1
          description: The spec object defines how the index should be deployed.
          type: object
          oneOf:
            - title: Serverless
              type: object
              properties:
                serverless:
                  $ref: '#/components/schemas/ServerlessSpecResponse'
              required:
                - serverless
              additionalProperties: false
            - title: Pod-based
              type: object
              properties:
                pod:
                  $ref: '#/components/schemas/PodSpec'
              required:
                - pod
              additionalProperties: false
            - title: BYOC
              type: object
              properties:
                byoc:
                  $ref: '#/components/schemas/ByocSpecResponse'
              required:
                - byoc
              additionalProperties: false
        status:
          example:
            ready: true
            state: ScalingUpPodSize
          description: The current status of the index
          type: object
          properties:
            ready:
              description: Whether the index is ready for use
              type: boolean
            state:
              description: >-
                The state of the index.

                Possible values: `Initializing`, `InitializationFailed`,
                `ScalingUp`, `ScalingDown`, `ScalingUpPodSize`,
                `ScalingDownPodSize`, `Terminating`, `Ready`, or `Disabled`.
              x-enum:
                - Initializing
                - InitializationFailed
                - ScalingUp
                - ScalingDown
                - ScalingUpPodSize
                - ScalingDownPodSize
                - Terminating
                - Ready
                - Disabled
              type: string
          required:
            - ready
            - state
        vector_type:
          description: >-
            The index vector type. You can use 'dense' or 'sparse'. If 'dense',
            the vector dimension must be specified.  If 'sparse', the vector
            dimension should not be specified.
          default: dense
          type: string
      required:
        - name
        - metric
        - status
        - spec
        - host
        - vector_type
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    DeletionProtection:
      description: >-
        Whether [deletion
        protection](http://docs.pinecone.io/guides/manage-data/manage-indexes#configure-deletion-protection)
        is enabled/disabled for the index.

        Possible values: `disabled` or `enabled`.
      default: disabled
      x-enum:
        - disabled
        - enabled
      type: string
    IndexTags:
      example:
        tag0: val0
        tag1: val1
      description: >-
        Custom user tags added to an index. Keys must be 80 characters or less.
        Values must be 120 characters or less. Keys must be alphanumeric, '_',
        or '-'.  Values must be alphanumeric, ';', '@', '_', '-', '.', '+', or '
        '. To unset a key, set the value to be an empty string.
      type: object
      additionalProperties:
        type: string
    IndexSpec:
      description: >
        The spec object defines how the index should be deployed.


        For serverless indexes, you define only the [cloud and
        region](http://docs.pinecone.io/guides/index-data/create-an-index#cloud-regions)
        where the index should be hosted. For pod-based indexes, you define the
        [environment](http://docs.pinecone.io/guides/indexes/pods/understanding-pod-based-indexes#pod-environments)
        where the index should be hosted, the [pod type and
        size](http://docs.pinecone.io/guides/indexes/pods/understanding-pod-based-indexes#pod-types)
        to use, and other index characteristics.
      type: object
      oneOf:
        - title: Serverless
          type: object
          properties:
            serverless:
              $ref: '#/components/schemas/ServerlessSpec'
          required:
            - serverless
          additionalProperties: false
        - title: Pod-based
          type: object
          properties:
            pod:
              $ref: '#/components/schemas/PodSpec'
          required:
            - pod
          additionalProperties: false
        - title: BYOC
          type: object
          properties:
            byoc:
              $ref: '#/components/schemas/ByocSpec'
          required:
            - byoc
          additionalProperties: false
    ModelIndexEmbed:
      example:
        field_map:
          text: your-text-field
        metric: cosine
        model: multilingual-e5-large
        read_parameters:
          input_type: query
          truncate: NONE
        write_parameters:
          input_type: passage
      description: The embedding model and document fields mapped to embedding inputs.
      type: object
      properties:
        model:
          example: multilingual-e5-large
          description: The name of the embedding model used to create the index.
          type: string
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If not specified, the metric
            will be defaulted according to the model. Cannot be updated once
            set.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        vector_type:
          description: >-
            The index vector type. You can use 'dense' or 'sparse'. If 'dense',
            the vector dimension must be specified.  If 'sparse', the vector
            dimension should not be specified.
          default: dense
          type: string
        field_map:
          example:
            text: your-text-field
          description: >-
            Identifies the name of the text field from your document model that
            is embedded.
          type: object
        read_parameters:
          description: The read parameters for the embedding model.
          type: object
        write_parameters:
          description: The write parameters for the embedding model.
          type: object
      required:
        - model
    ServerlessSpecResponse:
      description: Configuration of a serverless index.
      type: object
      properties:
        cloud:
          example: aws
          description: |-
            The public cloud where you would like your index hosted.
            Possible values: `gcp`, `aws`, or `azure`.
          x-enum:
            - gcp
            - aws
            - azure
          type: string
        region:
          example: us-east-1
          description: The region where you would like your index to be created.
          type: string
        read_capacity:
          $ref: '#/components/schemas/ReadCapacityResponse'
        source_collection:
          example: movie-embeddings
          description: The name of the collection to be used as the source for the index.
          type: string
        schema:
          $ref: '#/components/schemas/MetadataSchema'
      required:
        - cloud
        - region
        - read_capacity
    PodSpec:
      example:
        environment: us-east1-gcp
        metadata_config:
          indexed:
            - genre
            - title
            - imdb_rating
        pod_type: p1.x1
        pods: 1
        replicas: 1
        shards: 1
        source_collection: movie-embeddings
      title: Pod-based
      description: Configuration needed to deploy a pod-based index.
      type: object
      properties:
        environment:
          example: us-east1-gcp
          description: The environment where the index is hosted.
          type: string
        replicas:
          description: >-
            The number of replicas. Replicas duplicate your index. They provide
            higher availability and throughput. Replicas can be scaled up or
            down as your needs change.
          default: 1
          type: integer
          format: int32
          minimum: 1
        shards:
          description: >-
            The number of shards. Shards split your data across multiple pods so
            you can fit more data into an index.
          default: 1
          type: integer
          format: int32
          minimum: 1
        pod_type:
          description: >-
            The type of pod to use. One of `s1`, `p1`, or `p2` appended with `.`
            and one of `x1`, `x2`, `x4`, or `x8`.
          default: p1.x1
          type: string
        pods:
          example: 1
          description: >-
            The number of pods to be used in the index. This should be equal to
            `shards` x `replicas`.'
          default: 1
          type: integer
          minimum: 1
        metadata_config:
          example:
            indexed:
              - genre
              - title
              - imdb_rating
          description: >-
            Configuration for the behavior of Pinecone's internal metadata
            index. By default, all metadata is indexed; when `metadata_config`
            is present, only specified metadata fields are indexed. These
            configurations are only valid for use with pod-based indexes.
          type: object
          properties:
            indexed:
              description: >-
                By default, all metadata is indexed; to change this behavior,
                use this property to specify an array of metadata fields that
                should be indexed.
              type: array
              items:
                type: string
        source_collection:
          example: movie-embeddings
          description: The name of the collection to be used as the source for the index.
          type: string
      required:
        - environment
        - pod_type
    ByocSpecResponse:
      title: BYOC
      description: Configuration of a BYOC index.
      type: object
      properties:
        environment:
          example: aws-us-east-1-b921
          description: The environment where the index is hosted.
          type: string
        read_capacity:
          $ref: '#/components/schemas/ReadCapacityResponse'
        schema:
          $ref: '#/components/schemas/MetadataSchema'
      required:
        - environment
        - read_capacity
    ServerlessSpec:
      title: Serverless
      description: Configuration needed to deploy a serverless index.
      type: object
      properties:
        cloud:
          example: aws
          description: |-
            The public cloud where you would like your index hosted.
            Possible values: `gcp`, `aws`, or `azure`.
          x-enum:
            - gcp
            - aws
            - azure
          type: string
        region:
          example: us-east-1
          description: The region where you would like your index to be created.
          type: string
        read_capacity:
          $ref: '#/components/schemas/ReadCapacity'
        source_collection:
          example: movie-embeddings
          description: The name of the collection to be used as the source for the index.
          type: string
        schema:
          $ref: '#/components/schemas/MetadataSchema'
      required:
        - cloud
        - region
    ByocSpec:
      example:
        environment: aws-us-east-1-b921
      title: BYOC
      description: Configuration needed to deploy an index in a BYOC environment.
      type: object
      properties:
        environment:
          example: aws-us-east-1-b921
          description: The environment where the index is hosted.
          type: string
        read_capacity:
          $ref: '#/components/schemas/ReadCapacity'
        schema:
          $ref: '#/components/schemas/MetadataSchema'
      required:
        - environment
    ReadCapacityResponse:
      description: Response containing read capacity configuration
      discriminator:
        propertyName: mode
        mapping:
          OnDemand: '#/components/schemas/ReadCapacityOnDemandSpecResponse'
          Dedicated: '#/components/schemas/ReadCapacityDedicatedSpecResponse'
      oneOf:
        - $ref: '#/components/schemas/ReadCapacityOnDemandSpecResponse'
        - $ref: '#/components/schemas/ReadCapacityDedicatedSpecResponse'
    MetadataSchema:
      example:
        fields:
          description:
            filterable: true
          genre:
            filterable: true
          year:
            filterable: true
      description: >-
        Schema for the behavior of Pinecone's internal metadata index. By
        default, all metadata is indexed; when `schema` is present, only fields
        which are present in the `fields` object with a `filterable: true` are
        indexed. Note that `filterable: false` is not currently supported.
      type: object
      properties:
        fields:
          description: >-
            A map of metadata field names to their configuration. The field name
            must be a valid metadata field name. The field name must be unique.
          type: object
          additionalProperties:
            type: object
            properties:
              filterable:
                description: >-
                  Whether the field is filterable. If true, the field is indexed
                  and can be used in filters. Only true values are allowed.
                type: boolean
      required:
        - fields
    ReadCapacity:
      description: >-
        By default the index will be created with read capacity  mode
        `OnDemand`. If you prefer to allocate dedicated read  nodes for your
        workload, you must specify mode `Dedicated` and additional
        configurations for `node_type` and `scaling`.
      discriminator:
        propertyName: mode
        mapping:
          OnDemand: '#/components/schemas/ReadCapacityOnDemandSpec'
          Dedicated: '#/components/schemas/ReadCapacityDedicatedSpec'
      oneOf:
        - $ref: '#/components/schemas/ReadCapacityOnDemandSpec'
        - $ref: '#/components/schemas/ReadCapacityDedicatedSpec'
    ReadCapacityOnDemandSpecResponse:
      example:
        mode: OnDemand
        status:
          state: Ready
      title: On-demand
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
        status:
          $ref: '#/components/schemas/ReadCapacityStatus'
      required:
        - mode
        - status
      additionalProperties: false
    ReadCapacityDedicatedSpecResponse:
      example:
        dedicated:
          manual:
            replicas: 2
            shards: 2
          node_type: t1
          scaling: Manual
        mode: Dedicated
        status:
          current_replicas: 2
          current_shards: 2
          state: Ready
      title: Dedicated
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
        dedicated:
          $ref: '#/components/schemas/ReadCapacityDedicatedConfig'
        status:
          $ref: '#/components/schemas/ReadCapacityStatus'
      required:
        - mode
        - dedicated
        - status
      additionalProperties: false
    ReadCapacityOnDemandSpec:
      example:
        mode: OnDemand
      title: On-demand
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
      required:
        - mode
      additionalProperties: false
    ReadCapacityDedicatedSpec:
      example:
        dedicated:
          manual:
            replicas: 1
            shards: 1
          node_type: t1
          scaling: Manual
        mode: Dedicated
      title: Dedicated
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
        dedicated:
          $ref: '#/components/schemas/ReadCapacityDedicatedConfig'
      required:
        - mode
        - dedicated
      additionalProperties: true
    ReadCapacityStatus:
      description: >-
        The current status of factors affecting the read capacity of a
        serverless index
      type: object
      properties:
        state:
          description: >-
            The `state` describes the overall status of factors relating to the
            read capacity of an index. 


            Available values:

            - `Ready` is the state most of the time

            - `Scaling` if the number of replicas or shards has been recently
            updated by calling the [configure index
            endpoint](https://docs.pinecone.io/reference/api/2025-10/control-plane/configure_index)

            - `Migrating` if the index is being migrated to a new `node_type`

            - `Error` if there is an error with the read capacity configuration.
            In that case, see `error_message` for more details.
          type: string
        current_replicas:
          description: >-
            The number of replicas. Each replica has dedicated  compute
            resources and data storage. Increasing this number  will increase
            the total throughput of the index.
          type: integer
          format: int32
        current_shards:
          description: >-
            The number of shards. Each shard has dedicated storage.  Increasing
            shards alleiviates index fullness. 
          type: integer
          format: int32
        error_message:
          description: >-
            An optional error message indicating any issues with your read
            capacity configuration
          type: string
      required:
        - state
    ReadCapacityDedicatedConfig:
      description: >-
        Configuration for dedicated read capacity. See  [this
        guide](https://docs.pinecone.io/guides/index-data/dedicated-read-nodes)
        for more details on  how to configure dedicated read capacity.
      type: object
      properties:
        node_type:
          description: >-
            The type of machines to use. Available options: `b1` and `t1`. `t1`
            includes increased processing power and memory.
          type: string
        scaling:
          description: The type of scaling strategy to use.
          type: string
        manual:
          $ref: '#/components/schemas/ScalingConfigManual'
    ScalingConfigManual:
      description: The config to use for manual read capacity scaling.
      type: object
      properties:
        replicas:
          description: >-
            The number of replicas to use. Replicas duplicate the compute
            resources and data of an index, allowing higher query throughput and
            availability. Setting replicas to 0 disables the index but can be
            used to reduce costs while usage is paused.
          type: integer
          format: int32
          minimum: 0
        shards:
          description: >-
            The number of shards to use. Shards determine the storage capacity
            of an index, with each shard providing 250 GB of storage.
          type: integer
          format: int32
          minimum: 1
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/create_index_from_backup.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create an index from a backup

> Create an index from a backup.

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  BACKUP_ID="a65ff585-d987-4da5-a622-72e19a6ed5f4"

  curl "https://api.pinecone.io/backups/$BACKUP_ID/create-index" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -H 'Content-Type: application/json' \
    -d '{
          "name": "restored-index",
          "tags": {
            "tag0": "val0",
            "tag1": "val1"
          },
          "deletion_protection": "enabled"
        }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
      "restore_job_id":"e9ba8ff8-7948-4cfa-ba43-34227f6d30d4",
      "index_id":"025117b3-e683-423c-b2d1-6d30fbe5027f"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml post /backups/{backup_id}/create-index
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /backups/{backup_id}/create-index:
    post:
      tags:
        - Manage Indexes
      summary: Create an index from a backup
      description: Create an index from a backup.
      operationId: create_index_from_backup_operation
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: backup_id
          description: The ID of the backup to create an index from.
          required: true
          schema:
            type: string
          example: 670e8400-e29b-41d4-a716-446655440000
          style: simple
      requestBody:
        description: The desired configuration for the index created from a backup.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateIndexFromBackupRequest'
        required: true
      responses:
        '202':
          description: The request to create the index has been accepted.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/CreateIndexFromBackupResponse'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-metric-validation-error:
                  summary: Validation error
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Bad request. The request body included invalid request
                        parameters.
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '402':
          description: >-
            Payment required. Organization is on a paid plan and is delinquent
            on payment.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                payment-required:
                  summary: Payment required
                  value:
                    error:
                      code: PAYMENT_REQUIRED
                      message: >-
                        Request failed. Pay all past due invoices to lift
                        restrictions on your account.
                    status: 402
        '403':
          description: You've exceed your pod quota.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Forbidden
                  value:
                    error:
                      code: FORBIDDEN
                      message: Increase your quota or upgrade to create more indexes.
                    status: 403
        '404':
          description: Backup not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                backup-not-found:
                  summary: Backup not found
                  value:
                    error:
                      code: NOT_FOUND
                      message: Backup bkp_123abc not found.
                    status: 404
        '409':
          description: Index of given name already exists.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '422':
          description: Unprocessable entity. The request body could not be deserialized.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                missing-field:
                  summary: Unprocessable entity
                  value:
                    error:
                      code: UNPROCESSABLE_ENTITY
                      message: >-
                        Failed to deserialize the JSON body into the target
                        type: missing field `metric` at line 1 column 16
                    status: 422
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    CreateIndexFromBackupRequest:
      description: The configuration needed to create a Pinecone index from a backup.
      type: object
      properties:
        name:
          example: example-index
          description: >
            The name of the index. Resource name must be 1-45 characters long,
            start and end with an alphanumeric character, and consist only of
            lower case alphanumeric characters or '-'.
          type: string
          minLength: 1
          maxLength: 45
        tags:
          $ref: '#/components/schemas/IndexTags'
        deletion_protection:
          $ref: '#/components/schemas/DeletionProtection'
      required:
        - name
    CreateIndexFromBackupResponse:
      description: The response for creating an index from a backup.
      type: object
      properties:
        restore_job_id:
          example: 670e8400-e29b-41d4-a716-446655440000
          description: The ID of the restore job that was created.
          type: string
        index_id:
          example: 123e4567-e89b-12d3-a456-426614174000
          description: The ID of the index that was created from the backup.
          type: string
      required:
        - restore_job_id
        - index_id
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    IndexTags:
      example:
        tag0: val0
        tag1: val1
      description: >-
        Custom user tags added to an index. Keys must be 80 characters or less.
        Values must be 120 characters or less. Keys must be alphanumeric, '_',
        or '-'.  Values must be alphanumeric, ';', '@', '_', '-', '.', '+', or '
        '. To unset a key, set the value to be an empty string.
      type: object
      additionalProperties:
        type: string
    DeletionProtection:
      description: >-
        Whether [deletion
        protection](http://docs.pinecone.io/guides/manage-data/manage-indexes#configure-deletion-protection)
        is enabled/disabled for the index.

        Possible values: `disabled` or `enabled`.
      default: disabled
      x-enum:
        - disabled
        - enabled
      type: string
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/admin/create_project.md

# Source: https://docs.pinecone.io/reference/api/2025-10/admin-assistant/create_project.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create a new project

> Creates a new project.

<RequestExample>
  ```bash curl theme={null}
  PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"

  curl "https://api.pinecone.io/admin/projects" \
      -H "X-Pinecone-Api-Version: 2025-10" \
  	-H "Authorization: Bearer $PINECONE_ACCESS_TOKEN" \
  	-d '{
          "name":"example-project"
          }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
    "name": "example-project",
    "max_pods": 0,
    "force_encryption_with_cmek": false,
    "organization_id": "string",
    "created_at": "2025-03-16T22:46:45.030Z"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/admin_2025-10.oas.yaml post /admin/projects
openapi: 3.0.3
info:
  title: Pinecone Admin API
  description: |
    Provides an API for managing a Pinecone organization and its resources.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - BearerAuth: []
tags:
  - name: API Keys
    description: Actions that manage API Keys.
  - name: Organizations
    description: Actions that manage organizations.
  - name: Projects
    description: Actions that manage projects.
paths:
  /admin/projects:
    post:
      tags:
        - Projects
      summary: Create a new project
      description: Creates a new project.
      operationId: create_project
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        description: The details of the new project.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateProjectRequest'
        required: true
      responses:
        '200':
          description: The project was successfully created.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Project'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-metric-validation-error:
                  summary: Validation error
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Bad request. The request body included invalid request
                        parameters.
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '403':
          description: Forbidden
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '404':
          description: Not found
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
        4XX:
          description: Unexpected error on request.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
components:
  schemas:
    CreateProjectRequest:
      type: object
      properties:
        name:
          example: chatbot-prod
          description: The name of the new project.
          type: string
          minLength: 1
          maxLength: 512
        max_pods:
          description: >
            The maximum number of Pods that can be created in the project.
            Default is `0` (serverless only).
          type: integer
        force_encryption_with_cmek:
          description: >
            Whether to force encryption with a customer-managed encryption key
            (CMEK). Default is `false`.
          type: boolean
      required:
        - name
    Project:
      description: The details of a project.
      type: object
      properties:
        id:
          description: The unique ID of the project.
          type: string
          format: uuid
        name:
          description: The name of the project.
          type: string
          minLength: 1
          maxLength: 512
        max_pods:
          description: The maximum number of Pods that can be created in the project.
          type: integer
        force_encryption_with_cmek:
          description: >-
            Whether to force encryption with a customer-managed encryption key
            (CMEK).
          type: boolean
        organization_id:
          description: The unique ID of the organization that the project belongs to.
          type: string
        created_at:
          description: The date and time when the project was created.
          type: string
          format: date-time
      required:
        - id
        - name
        - max_pods
        - force_encryption_with_cmek
        - organization_id
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, or `UNPROCESSABLE_ENTITY`.        
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    BearerAuth:
      type: http
      scheme: bearer
      description: >
        An [access
        token](https://docs.pinecone.io/guides/organizations/manage-service-accounts#retrieve-an-access-token)
        must be provided in the `Authorization` header using the `Bearer`
        scheme.

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/createnamespace.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create a namespace

> Create a namespace in a serverless index.

For guidance and examples, see [Manage namespaces](https://docs.pinecone.io/guides/manage-data/manage-namespaces).

**Note:** This operation is not supported for pod-based indexes.

<RequestExample>
  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/namespaces" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "name": "example-namespace",
          "schema": {
  		  "fields": { 
              "document_id": {"filterable": true},
              "document_title": {"filterable": true},
              "chunk_number": {"filterable": true},
              "document_url": {"filterable": true},
              "created_at": {"filterable": true}
            }
          }
        }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
      "name": "example-namespace",
      "record_count": "0",
      "schema": {
          "fields": {
              "document_title": {
                  "filterable": true
              },
              "document_url": {
                  "filterable": true
              },
              "chunk_number": {
                  "filterable": true
              },
              "document_id": {
                  "filterable": true
              },
              "created_at": {
                  "filterable": true
              }
          }
      }
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml post /namespaces
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /namespaces:
    post:
      tags:
        - Namespace Operations
      summary: Create a namespace
      description: >-
        Create a namespace in a serverless index.


        For guidance and examples, see [Manage
        namespaces](https://docs.pinecone.io/guides/manage-data/manage-namespaces).


        **Note:** This operation is not supported for pod-based indexes.
      operationId: createNamespace
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateNamespaceRequest'
        required: true
      responses:
        '200':
          description: A successful response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/NamespaceDescription'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        '409':
          description: Namespace of the given name already exists on the index.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    CreateNamespaceRequest:
      description: A request for creating a namespace with the specified namespace name.
      type: object
      properties:
        name:
          example: example-namespace
          description: The name of the namespace.
          type: string
        schema:
          example:
            fields:
              description:
                filterable: true
              genre:
                filterable: true
              year:
                filterable: true
          description: >-
            Schema for the behavior of Pinecone's internal metadata index. By
            default, all metadata is indexed; when `schema` is present, only
            fields which are present in the `fields` object with a `filterable:
            true` are indexed. Note that `filterable: false` is not currently
            supported.
          type: object
          properties:
            fields:
              description: >-
                A map of metadata field names to their configuration. The field
                name must be a valid metadata field name. The field name must be
                unique.
              type: object
              additionalProperties:
                type: object
                properties:
                  filterable:
                    description: >-
                      Whether the field is filterable. If true, the field is
                      indexed and can be used in filters. Only true values are
                      allowed.
                    type: boolean
          required:
            - fields
      required:
        - name
    NamespaceDescription:
      description: A description of a namespace, including the name and record count.
      type: object
      properties:
        name:
          example: example-namespace
          description: The name of the namespace.
          type: string
        record_count:
          example: 20000
          description: The total amount of records within the namespace.
          type: integer
          format: int64
        schema:
          example:
            fields:
              description:
                filterable: true
              genre:
                filterable: true
              year:
                filterable: true
          description: >-
            Schema for the behavior of Pinecone's internal metadata index. By
            default, all metadata is indexed; when `schema` is present, only
            fields which are present in the `fields` object with a `filterable:
            true` are indexed. Note that `filterable: false` is not currently
            supported.
          type: object
          properties:
            fields:
              description: >-
                A map of metadata field names to their configuration. The field
                name must be a valid metadata field name. The field name must be
                unique.
              type: object
              additionalProperties:
                type: object
                properties:
                  filterable:
                    description: >-
                      Whether the field is filterable. If true, the field is
                      indexed and can be used in filters. Only true values are
                      allowed.
                    type: boolean
          required:
            - fields
        indexed_fields:
          description: A list of all indexed metatadata fields in the namespace
          type: object
          properties:
            fields:
              type: array
              items:
                type: string
              example:
                - genre
                - year
                - author
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/troubleshooting/custom-data-processing-agreements.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Custom data processing agreements

If you need a data processing agreement (DPA) with Pinecone you can get started by filling out the form in our [Security Center](https://security.pinecone.io/). You'll need to request access first. Simply click the "Get Access" box on that page and enter your information. You should receive a link with further instructions shortly.

---

# Source: https://docs.pinecone.io/guides/production/data-deletion.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Data deletion on Pinecone

> Understand Pinecone's secure data deletion process.

Pinecone follows a secure process to ensure that customer data is permanently deleted from our system. This page gives an overview of the process.

As defined in the [Master Subscription Agreement](https://www.pinecone.io/legal/master-subscription-agreement/), customer data is data that you provide to Pinecone through the services of the Pinecone system, or such data provided on your behalf by connected systems. This includes objects such as [records](/guides/get-started/concepts#record), [indexes](/guides/get-started/concepts#index), [backups](/guides/get-started/concepts#backup-or-collection), [projects](/guides/get-started/concepts#project), [API keys](/guides/get-started/concepts#api-key), [users](/guides/get-started/concepts#user), [assistants](/guides/get-started/concepts#pinecone-assistant), and [organizations](/guides/get-started/concepts#organization).

## Deletion request

The deletion of customer data begins when you initiate a deletion request through the Pinecone API, console, or a connected service. A deletion request can delete a single resource, such as a record, or can delete a resource and all its dependent resources, such as an index and all its records.

Deletion of your customer data also occurs automatically when you end your relationship with Pinecone.

## Soft deletion

After you initiate a deletion request, Pinecone marks the data for deletion. The data is not immediately removed from the system. Instead, Pinecone retains the data for a maximum of 90 days. During this period, the data is not accessible to you or any other user.

## Permanent deletion

Before the end of the 90-day retention window, Pinecone permanently deletes the data from its system. Once the data is permanently deleted, it is no longer recoverable.

<Note>
  Pinecone creates an [audit log](/guides/production/security-overview#audit-logs) of user, service account, and API events. Events are captured within two hours of occurrence and are retained for 90 days, after which they are permanently deleted.
</Note>

## See also

* [Delete records](/guides/manage-data/delete-data)
* [Delete an index](/guides/manage-data/manage-indexes#delete-an-index)
* [Delete a project](/guides/projects/manage-projects#delete-a-project)
* [Delete an API key](/guides/projects/manage-api-keys#delete-an-api-key)
* [Delete a user](/guides/projects/manage-project-members#remove-members)
* [Delete an organization](/troubleshooting/delete-your-organization)
* [Master Subscription Agreement](https://www.pinecone.io/legal/master-subscription-agreement/)

---

# Source: https://docs.pinecone.io/guides/index-data/data-ingestion-overview.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Data ingestion overview

> Learn about the different ways to ingest data into Pinecone.

<Tip>
  To control costs when ingesting large datasets (10,000,000+ records), use [import](/guides/index-data/import-data) instead of upsert.
</Tip>

## Import from object storage

[Importing from object storage](/guides/index-data/import-data) is the most efficient and cost-effective method to load large numbers of records into an index. You store your data as Parquet files in object storage, integrate your object storage with Pinecone, and then start an asynchronous, long-running operation that imports and indexes your records.

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Standard and Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

## Upsert

For ongoing ingestion into an index, either one record at a time or in batches, use the [upsert](/guides/index-data/upsert-data) operation. [Batch upserting](/guides/index-data/upsert-data#upsert-in-batches) can improve throughput performance and is a good option for larger numbers of records if you cannot work around import's current [limitations](/guides/index-data/import-data#import-limits).

## Ingestion cost

* To understand how cost is calculated for imports, see [Import cost](/guides/manage-cost/understanding-cost#imports).
* To understand how cost is calculated for upserts, see [Upsert cost](/guides/manage-cost/understanding-cost#upsert).
* For up-to-date pricing information, see [Pricing](https://www.pinecone.io/pricing/).

## Data freshness

Pinecone is eventually consistent, so there can be a slight delay before new or changed records are visible to queries. You can view index stats to [check data freshness](/guides/index-data/check-data-freshness).

---

# Source: https://docs.pinecone.io/guides/index-data/data-modeling.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Data modeling

> Learn how to structure records for efficient data retrieval and management in Pinecone.

## Record format

<Tabs>
  <Tab title="Text">
    When you upsert raw text for Pinecone to convert to vectors automatically, each record consists of the following:

    * **ID**: A unique string identifier for the record.
    * **Text**: The raw text for Pinecone to convert to a dense vector for [semantic search](/guides/search/semantic-search) or a sparse vector for [lexical search](/guides/search/lexical-search), depending on the [embedding model](/guides/index-data/create-an-index#embedding-models) integrated with the index. This field name must match the `embed.field_map` defined in the index.
    * **Metadata** (optional): All additional fields are stored as record metadata. You can filter by metadata when searching or deleting records.

    <Note>
      Upserting raw text is supported only for [indexes with integrated embedding](/guides/index-data/indexing-overview#vector-embedding).
    </Note>

    Example:

    ```json  theme={null}
    {
      "_id": "document1#chunk1", 
      "chunk_text": "First chunk of the document content...", // Text to convert to a vector. 
      "document_id": "document1", // This and subsequent fields stored as metadata. 
      "document_title": "Introduction to Vector Databases",
      "chunk_number": 1,
      "document_url": "https://example.com/docs/document1", 
      "created_at": "2024-01-15",
      "document_type": "tutorial"
    }
    ```
  </Tab>

  <Tab title="Vectors">
    When you upsert pre-generated vectors, each record consists of the following:

    * **ID**: A unique string identifier for the record.
    * **Vector**: A dense vector for [semantic search](/guides/search/semantic-search), a sparse vector for [lexical search](/guides/search/lexical-search), or both for [hybrid search](/guides/search/hybrid-search) using a single hybrid index.
    * **Metadata** (optional):  A flat JSON document containing key-value pairs with additional information (nested objects are not supported). You can filter by metadata when searching or deleting records.

    <Note>
      When importing data from object storage, records must be in Parquet format. For more details, see [Import data](/guides/index-data/import-data#prepare-your-data).
    </Note>

    Example:

    <CodeGroup>
      ```json Dense theme={null}
      {
        "id": "document1#chunk1", 
        "values": [0.0236663818359375, -0.032989501953125, ..., -0.01041412353515625, 0.0086669921875], 
        "metadata": {
          "document_id": "document1",
          "document_title": "Introduction to Vector Databases",
          "chunk_number": 1,
          "chunk_text": "First chunk of the document content...",
          "document_url": "https://example.com/docs/document1",
          "created_at": "2024-01-15",
          "document_type": "tutorial"
        }
      }
      ```

      ```json Sparse theme={null}
      {
        "id": "document1#chunk1", 
        "sparse_values": {
          "values": [1.7958984, 0.41577148, ..., 4.4414062, 3.3554688],
          "indices": [822745112, 1009084850, ..., 3517203014, 3590924191]
        },
        "metadata": {
          "document_id": "document1",
          "document_title": "Introduction to Vector Databases",
          "chunk_number": 1,
          "chunk_text": "First chunk of the document content...",
          "document_url": "https://example.com/docs/document1",
          "created_at": "2024-01-15",
          "document_type": "tutorial"
        }
      }
      ```

      ```json Hybrid theme={null}
      {
        "id": "document1#chunk1", 
        "values": [0.0236663818359375, -0.032989501953125, ..., -0.01041412353515625, 0.0086669921875], 
        "sparse_values": {
          "values": [1.7958984, 0.41577148, ..., 4.4414062, 3.3554688],
          "indices": [822745112, 1009084850, ..., 3517203014, 3590924191]
        },
        "metadata": {
          "document_id": "document1",
          "document_title": "Introduction to Vector Databases",
          "chunk_number": 1,
          "chunk_text": "First chunk of the document content...",
          "document_url": "https://example.com/docs/document1",
          "created_at": "2024-01-15",
          "document_type": "tutorial"
        }
      }
      ```
    </CodeGroup>
  </Tab>
</Tabs>

## Use structured IDs

Use a structured, human-readable format for record IDs, including ID prefixes that reflect the type of data you're storing, for example:

* **Document chunks**: `document_id#chunk_number`
* **User data**: `user_id#data_type#item_id`
* **Multi-tenant data**: `tenant_id#document_id#chunk_id`

Choose a delimiter for your ID prefixes that won't appear elsewhere in your IDs. Common patterns include:

* `document1#chunk1` - Using hash delimiter
* `document1_chunk1` - Using underscore delimiter
* `document1:chunk1` - Using colon delimiter

Structuring IDs in this way provides several advantages:

* **Efficiency**: Applications can quickly identify which record it should operate on.
* **Clarity**: Developers can easily understand what they're looking at when examining records.
* **Flexibility**: ID prefixes enable list operations for fetching and updating records.

## Include metadata

Include [metadata key-value pairs](/guides/index-data/indexing-overview#metadata) that support your application's key operations, for example:

* **Enable query-time filtering**: Add fields for time ranges, categories, or other criteria for [filtering searches for increased accuracy and relevance](/guides/search/filter-by-metadata).
* **Link related chunks**: Use fields like `document_id` and `chunk_number` to keep track of related records and enable efficient [chunk deletion](#delete-chunks) and [document updates](#update-an-entire-document).
* **Link back to original data**: Include `chunk_text` or `document_url` for traceability and user display.

Metadata keys must be strings, and metadata values must be one of the following data types:

* String
* Number (integer or floating point, gets converted to a 64-bit floating point)
* Boolean (true, false)
* List of strings

<Note>
  Pinecone supports 40 KB of metadata per record.
</Note>

## Example

This example demonstrates how to manage document chunks in Pinecone using structured IDs and comprehensive metadata. It covers the complete lifecycle of chunked documents: upserting, searching, fetching, updating, and deleting chunks, and updating an entire document.

### Upsert chunks

When [upserting](/guides/index-data/upsert-data) documents that have been split into chunks, combine structured IDs with comprehensive metadata:

<Tabs>
  <Tab title="Upsert text">
    <Note>
      Upserting raw text is supported only for [indexes with integrated embedding](/guides/index-data/create-an-index#integrated-embedding).
    </Note>

    ```python Python theme={null}
    from pinecone.grpc import PineconeGRPC as Pinecone

    pc = Pinecone(api_key="YOUR_API_KEY")

    # To get the unique host for an index, 
    # see https://docs.pinecone.io/guides/manage-data/target-an-index
    index = pc.Index(host="INDEX_HOST")

    index.upsert_records(
      "example-namespace",
      [
        {
          "_id": "document1#chunk1", 
          "chunk_text": "First chunk of the document content...",
          "document_id": "document1",
          "document_title": "Introduction to Vector Databases",
          "chunk_number": 1,
          "document_url": "https://example.com/docs/document1",
          "created_at": "2024-01-15",
          "document_type": "tutorial"
        },
        {
          "_id": "document1#chunk2", 
          "chunk_text": "Second chunk of the document content...",
          "document_id": "document1",
          "document_title": "Introduction to Vector Databases", 
          "chunk_number": 2,
          "document_url": "https://example.com/docs/document1",
          "created_at": "2024-01-15",
          "document_type": "tutorial"
        },
        {
          "_id": "document1#chunk3", 
          "chunk_text": "Third chunk of the document content...",
          "document_id": "document1",
          "document_title": "Introduction to Vector Databases",
          "chunk_number": 3, 
          "document_url": "https://example.com/docs/document1",
          "created_at": "2024-01-15",
          "document_type": "tutorial"
        },
      ]
    )
    ```
  </Tab>

  <Tab title="Upsert vectors">
    ```python Python theme={null}
    from pinecone.grpc import PineconeGRPC as Pinecone

    pc = Pinecone(api_key="YOUR_API_KEY")

    # To get the unique host for an index, 
    # see https://docs.pinecone.io/guides/manage-data/target-an-index
    index = pc.Index(host="INDEX_HOST")

    index.upsert(
      namespace="example-namespace",
      vectors=[
        {
          "id": "document1#chunk1", 
          "values": [0.0236663818359375, -0.032989501953125, ..., -0.01041412353515625, 0.0086669921875], 
          "metadata": {
            "document_id": "document1",
            "document_title": "Introduction to Vector Databases",
            "chunk_number": 1,
            "chunk_text": "First chunk of the document content...",
            "document_url": "https://example.com/docs/document1",
            "created_at": "2024-01-15",
            "document_type": "tutorial"
          }
        },
        {
          "id": "document1#chunk2", 
          "values": [-0.0412445068359375, 0.028839111328125, ..., 0.01953125, -0.0174560546875],
          "metadata": {
            "document_id": "document1",
            "document_title": "Introduction to Vector Databases", 
            "chunk_number": 2,
            "chunk_text": "Second chunk of the document content...",
            "document_url": "https://example.com/docs/document1",
            "created_at": "2024-01-15",
            "document_type": "tutorial"
          }
        },
        {
          "id": "document1#chunk3", 
          "values": [0.0512237548828125, 0.041656494140625, ..., 0.02130126953125, -0.0394287109375],
          "metadata": {
            "document_id": "document1",
            "document_title": "Introduction to Vector Databases",
            "chunk_number": 3, 
            "chunk_text": "Third chunk of the document content...",
            "document_url": "https://example.com/docs/document1",
            "created_at": "2024-01-15",
            "document_type": "tutorial"
          }
        }
      ]
    )
    ```
  </Tab>
</Tabs>

### Search chunks

To search the chunks of a document, use a [metadata filter expression](/guides/search/filter-by-metadata#metadata-filter-expressions) that limits the search appropriately:

<Tabs>
  <Tab title="Search with text">
    <Note>
      Searching with text is supported only for [indexes with integrated embedding](/guides/index-data/create-an-index#integrated-embedding).
    </Note>

    ```python Python theme={null}
    from pinecone import Pinecone

    pc = Pinecone(api_key="YOUR_API_KEY")

    # To get the unique host for an index, 
    # see https://docs.pinecone.io/guides/manage-data/target-an-index
    index = pc.Index(host="INDEX_HOST")

    filtered_results = index.search(
        namespace="example-namespace", 
        query={
            "inputs": {"text": "What is a vector database?"}, 
            "top_k": 3,
            "filter": {"document_id": "document1"}
        },
        fields=["chunk_text"]
    )

    print(filtered_results)
    ```
  </Tab>

  <Tab title="Search with a vector">
    ```python Python theme={null}
    from pinecone.grpc import PineconeGRPC as Pinecone

    pc = Pinecone(api_key="YOUR_API_KEY")

    # To get the unique host for an index, 
    # see https://docs.pinecone.io/guides/manage-data/target-an-index
    index = pc.Index(host="INDEX_HOST")

    filtered_results = index.query(
        namespace="example-namespace",
        vector=[0.0236663818359375,-0.032989501953125, ..., -0.01041412353515625,0.0086669921875], 
        top_k=3,
        filter={
            "document_id": {"$eq": "document1"}
        },
        include_metadata=True,
        include_values=False
    )

    print(filtered_results)
    ```
  </Tab>
</Tabs>

### Fetch chunks

To retrieve all chunks for a specific document, first [list the record IDs](/guides/manage-data/list-record-ids) using the document prefix, and then [fetch](/guides/manage-data/fetch-data) the complete records:

```python Python theme={null}
from pinecone.grpc import PineconeGRPC as Pinecone

pc = Pinecone(api_key="YOUR_API_KEY")

# To get the unique host for an index, 
# see https://docs.pinecone.io/guides/manage-data/target-an-index
index = pc.Index(host="INDEX_HOST")

# List all chunks for document1 using ID prefix
chunk_ids = []
for record_id in index.list(prefix='document1#', namespace='example-namespace'):
    chunk_ids.append(record_id)

print(f"Found {len(chunk_ids)} chunks for document1")

# Fetch the complete records by ID
if chunk_ids:
    records = index.fetch(ids=chunk_ids, namespace='example-namespace')
    
    for record_id, record_data in records['vectors'].items():
        print(f"Chunk ID: {record_id}")
        print(f"Chunk text: {record_data['metadata']['chunk_text']}")
        # Process the vector values and metadata as needed
```

<Note>
  Pinecone is [eventually consistent](/guides/index-data/check-data-freshness), so it's possible that a write (upsert, update, or delete) followed immediately by a read (query, list, or fetch) may not return the latest version of the data. If your use case requires retrieving data immediately, consider implementing a small delay or [retry logic](/guides/production/error-handling#implement-retry-logic) after writes.
</Note>

### Update chunks

To [update](/guides/manage-data/update-data) specific chunks within a document, first list the chunk IDs, and then update individual records:

```python Python theme={null}
from pinecone.grpc import PineconeGRPC as Pinecone

pc = Pinecone(api_key="YOUR_API_KEY")

# To get the unique host for an index, 
# see https://docs.pinecone.io/guides/manage-data/target-an-index
index = pc.Index(host="INDEX_HOST")

# List all chunks for document1
chunk_ids = []
for record_id in index.list(prefix='document1#', namespace='example-namespace'):
    chunk_ids.append(record_id)

# Update specific chunks (e.g., update chunk 2)
if 'document1#chunk2' in chunk_ids:
    index.update(
        id='document1#chunk2',
        values=[<new dense vector>],
        set_metadata={
            "document_id": "document1",
            "document_title": "Introduction to Vector Databases - Revised",
            "chunk_number": 2,
            "chunk_text": "Updated second chunk content...",
            "document_url": "https://example.com/docs/document1",
            "created_at": "2024-01-15",
            "updated_at": "2024-02-15",
            "document_type": "tutorial"
        },
        namespace='example-namespace'
    )
    print("Updated chunk 2 successfully")
```

### Delete chunks

To [delete](/guides/manage-data/delete-data#delete-records-by-metadata) chunks of a document, use a [metadata filter expression](/guides/search/filter-by-metadata#metadata-filter-expressions) that limits the deletion appropriately:

```python Python theme={null}
from pinecone.grpc import PineconeGRPC as Pinecone

pc = Pinecone(api_key="YOUR_API_KEY")

# To get the unique host for an index, 
# see https://docs.pinecone.io/guides/manage-data/target-an-index
index = pc.Index(host="INDEX_HOST")

# Delete chunks 1 and 3
index.delete(
    namespace="example-namespace",
    filter={
        "document_id": {"$eq": "document1"},
        "chunk_number": {"$in": [1, 3]}
    }
)

# Delete all chunks for a document
index.delete(
    namespace="example-namespace",
    filter={
        "document_id": {"$eq": "document1"}
    }
)
```

### Update an entire document

When the amount of chunks or ordering of chunks for a document changes, the recommended approach is to first [delete all chunks using a metadata filter](/guides/manage-data/delete-data#delete-records-by-metadata), and then [upsert](/guides/index-data/upsert-data) the new chunks:

```python Python theme={null}
from pinecone.grpc import PineconeGRPC as Pinecone

pc = Pinecone(api_key="YOUR_API_KEY")

# To get the unique host for an index, 
# see https://docs.pinecone.io/guides/manage-data/target-an-index
index = pc.Index(host="INDEX_HOST")

# Step 1: Delete all existing chunks for the document
index.delete(
    namespace="example-namespace",
    filter={
        "document_id": {"$eq": "document1"}
    }
)

print("Deleted existing chunks for document1")

# Step 2: Upsert the updated document chunks
index.upsert(
  namespace="example-namespace", 
  vectors=[
    {
      "id": "document1#chunk1",
      "values": [<updated dense vector>],
      "metadata": {
        "document_id": "document1",
        "document_title": "Introduction to Vector Databases - Updated Edition",
        "chunk_number": 1,
        "chunk_text": "Updated first chunk with new content...",
        "document_url": "https://example.com/docs/document1",
        "created_at": "2024-02-15",
        "document_type": "tutorial",
        "version": "2.0"
      }
    },
    {
      "id": "document1#chunk2",
      "values": [<updated dense vector>],
      "metadata": {
        "document_id": "document1",
        "document_title": "Introduction to Vector Databases - Updated Edition",
        "chunk_number": 2,
        "chunk_text": "Updated second chunk with new content...",
        "document_url": "https://example.com/docs/document1",
        "created_at": "2024-02-15",
        "document_type": "tutorial",
        "version": "2.0"
      }
    }
    # Add more chunks as needed for the updated document
  ]
)

print("Successfully updated document1 with new chunks")
```

## Data freshness

Pinecone is [eventually consistent](/guides/index-data/check-data-freshness), so it's possible that a write (upsert, update, or delete) followed immediately by a read (query, list, or fetch) may not return the latest version of the data. If your use case requires retrieving data immediately, consider implementing a small delay or [retry logic](/guides/production/error-handling#implement-retry-logic) after writes.

## Design for multi-tenancy

Many applications have a concept of tenants—users, organizations, projects, or other groups that should only access their own data. How you model this access control significantly impacts query performance and cost.

### Use namespaces for tenant isolation

The most efficient way to implement multi-tenancy is to use [namespaces](/guides/index-data/indexing-overview#namespaces) to separate data by tenant. With this approach, each tenant has their own namespace, and queries only scan that tenant's data—resulting in better performance and lower costs.

For a complete implementation guide with examples across all SDKs, see [Implement multitenancy](/guides/index-data/implement-multitenancy).

<Accordion title="Why namespaces are more efficient">
  When you use namespaces for multi-tenancy:

  * **Lower query costs and faster performance**: Query cost is based on namespace size. If you have 100 tenants with 1 GB each, querying one tenant's namespace costs 1 RU and scans only 1 GB. With metadata filtering in a single namespace (100 GB total), the same query costs 100 RUs and scans all 100 GB, even though the filter narrows results.
  * **Natural isolation**: Reduces the risk of application bugs that could query the wrong tenant's data (for example, by passing an incorrect filter value).
</Accordion>

### Avoid filtering by high-cardinality IDs

A common anti-pattern is storing all data in a single namespace and using metadata filters to scope queries to specific users:

<CodeGroup>
  ```python Python theme={null}
  # Anti-pattern: Filtering by many user IDs
  query_vector = [0.1, 0.2, 0.3, ...]  # Your query vector
  results = index.query(
      vector=query_vector,
      top_k=10,
      filter={
          "allowed_user_ids": {"$in": ["user_1", "user_2", ..., "user_10000"]}
      }
  )
  ```

  ```javascript JavaScript theme={null}
  // Anti-pattern: Filtering by many user IDs
  const queryVector = [0.1, 0.2, 0.3, ...];  // Your query vector
  const results = await index.query({
    vector: queryVector,
    topK: 10,
    filter: {
      allowed_user_ids: { $in: ["user_1", "user_2", ..., "user_10000"] }
    }
  });
  ```

  ```java Java theme={null}
  // Anti-pattern: Filtering by many user IDs
  import com.google.protobuf.Struct;
  import com.google.protobuf.Value;
  import io.pinecone.clients.Index;
  import io.pinecone.clients.Pinecone;
  import io.pinecone.unsigned_indices_model.QueryResponseWithUnsignedIndices;
  import java.util.Arrays;
  import java.util.List;

  Pinecone pinecone = new Pinecone.Builder("YOUR_API_KEY").build();
  Index index = pinecone.getIndexConnection("your-index-name");
  List<Float> queryVector = Arrays.asList(0.1f, 0.2f, 0.3f, ...);  // Your query vector

  // Build filter with $in operator (up to 10,000 values)
  Struct.Builder filterBuilder = Struct.newBuilder();
  Value.Builder listValueBuilder = Value.newBuilder();
  listValueBuilder.getListValueBuilder()
      .addAllValues(Arrays.asList(
          Value.newBuilder().setStringValue("user_1").build(),
          Value.newBuilder().setStringValue("user_2").build()
          // ... up to 10,000 values
      ));
  filterBuilder.putFields("allowed_user_ids",
      Value.newBuilder()
          .setStructValue(Struct.newBuilder()
              .putFields("$in", listValueBuilder.build())
              .build())
          .build());
  Struct filter = filterBuilder.build();

  QueryResponseWithUnsignedIndices results = index.queryByVector(
      10,
      queryVector,
      null, // default namespace
      filter
  );
  ```

  ```go Go theme={null}
  // Anti-pattern: Filtering by many user IDs
  import (
      "context"
      "fmt"
      "log"
      "github.com/pinecone-io/go-pinecone/v5/pinecone"
      "google.golang.org/protobuf/types/known/structpb"
  )

  ctx := context.Background()

  clientParams := pinecone.NewClientParams{
      ApiKey: "YOUR_API_KEY",
  }
  pc, err := pinecone.NewClient(clientParams)
  if err != nil {
      log.Fatalf("Failed to create Client: %v", err)
  }

  idx, err := pc.DescribeIndex(ctx, "your-index-name")
  if err != nil {
      log.Fatalf("Failed to describe index: %v", err)
  }

  idxConnection, err := pc.Index(pinecone.NewIndexConnParams{
      Host: idx.Host,
  })
  if err != nil {
      log.Fatalf("Failed to create IndexConnection: %v", err)
  }

  queryVector := []float32{0.1, 0.2, 0.3, ...}  // Your query vector

  userIds := []interface{}{"user_1", "user_2", /* ... up to 10,000 values */}
  metadataMap := map[string]interface{}{
      "allowed_user_ids": map[string]interface{}{
          "$in": userIds,
      },
  }
  filter, err := structpb.NewStruct(metadataMap)
  if err != nil {
      log.Fatalf("Failed to create filter: %v", err)
  }

  queryReq := &pinecone.QueryByVectorValuesRequest{
      Vector:          queryVector,
      TopK:            10,
      MetadataFilter:  filter,
      IncludeMetadata: true,
  }
  results, err := idxConnection.QueryByVectorValues(ctx, queryReq)
  if err != nil {
      log.Fatalf("Failed to query: %v", err)
  }
  fmt.Printf("Found %d matches:\n", len(results.Matches))
  for _, match := range results.Matches {
      fmt.Printf("  ID: %s, Score: %.4f\n", match.Vector.Id, match.Score)
      if match.Vector.Metadata != nil {
          fmt.Printf("    Metadata: %v\n", match.Vector.Metadata)
      }
  }
  ```

  ```csharp C# theme={null}
  // Anti-pattern: Filtering by many user IDs
  using Pinecone;
  using System;
  using System.Linq;

  var queryVector = new[] { 0.1f, 0.2f, 0.3f, ... };  // Your query vector
  // index is your IndexClient instance

  var userIds = new[] { "user_1", "user_2", /* ... up to 10,000 values */ };
  var filter = new Metadata
  {
      { "allowed_user_ids", new MetadataValue(new Metadata { { "$in", userIds } }) }
  };

  var results = await index.QueryAsync(new QueryRequest
  {
      Vector = queryVector,
      TopK = 10,
      Filter = filter,
      IncludeMetadata = true
  });
  if (results == null)
  {
      throw new InvalidOperationException("Failed to query");
  }
  Console.WriteLine($"Found {results.Matches?.Count() ?? 0} matches:");
  if (results.Matches != null)
  {
      foreach (var match in results.Matches)
      {
          Console.WriteLine($"  ID: {match.Id}, Score: {match.Score:F4}");
          if (match.Metadata != null)
          {
              Console.WriteLine($"    Metadata: {match.Metadata}");
          }
      }
  }
  ```

  ```bash curl theme={null}
  # Anti-pattern: Filtering by many user IDs
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X POST "https://$INDEX_HOST/query" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H "Content-Type: application/json" \
       -H "X-Pinecone-Api-Version: 2025-10" \
       -d '{
             "vector": [0.1, 0.2, 0.3, ...],
             "topK": 10,
             "includeMetadata": true,
             "filter": {
               "allowed_user_ids": {
                 "$in": ["user_1", "user_2", ..., "user_10000"]
               }
             }
          }'
  ```
</CodeGroup>

This approach has several drawbacks:

* **Performance degradation**: Large `$in` filters increase network payload size and query latency.
* **Hard limits**: Each `$in` or `$nin` operator is limited to 10,000 values. Exceeding this limit will cause the request to fail. See [Metadata filter limits](/reference/api/database-limits#metadata-filter-limits).

### Use access control groups instead of individual IDs

If data must be shared across many tenants, design your access control using the smallest number of groups that describe a user's access:

<CodeGroup>
  ```python Python theme={null}
  # Better: Filter by organization or role instead of individual users
  query_vector = [0.1, 0.2, 0.3, ...]  # Your query vector
  results = index.query(
      vector=query_vector,
      top_k=10,
      filter={
          "$or": [
              {"organization_id": {"$eq": "org_A"}},
              {"project_id": {"$eq": "project_B"}}
          ]
      }
  )
  ```

  ```javascript JavaScript theme={null}
  // Better: Filter by organization or role instead of individual users
  const queryVector = [0.1, 0.2, 0.3, ...];  // Your query vector
  const results = await index.query({
    vector: queryVector,
    topK: 10,
    filter: {
      $or: [
        { organization_id: { $eq: "org_A" } },
        { project_id: { $eq: "project_B" } }
      ]
    }
  });
  ```

  ```java Java theme={null}
  // Better: Filter by organization or role instead of individual users
  import com.google.protobuf.Struct;
  import com.google.protobuf.Value;
  import io.pinecone.clients.Index;
  import io.pinecone.clients.Pinecone;
  import io.pinecone.unsigned_indices_model.QueryResponseWithUnsignedIndices;
  import java.util.Arrays;
  import java.util.List;

  Pinecone pinecone = new Pinecone.Builder("YOUR_API_KEY").build();
  Index index = pinecone.getIndexConnection("your-index-name");
  List<Float> queryVector = Arrays.asList(0.1f, 0.2f, 0.3f, ...);  // Your query vector

  // Build filter with $or operator
  Struct.Builder orgFilterBuilder = Struct.newBuilder();
  orgFilterBuilder.putFields("organization_id",
      Value.newBuilder()
          .setStructValue(Struct.newBuilder()
              .putFields("$eq", Value.newBuilder()
                  .setStringValue("org_A")
                  .build())
              .build())
          .build());

  Struct.Builder projectFilterBuilder = Struct.newBuilder();
  projectFilterBuilder.putFields("project_id",
      Value.newBuilder()
          .setStructValue(Struct.newBuilder()
              .putFields("$eq", Value.newBuilder()
                  .setStringValue("project_B")
                  .build())
              .build())
          .build());

  Struct.Builder orFilterBuilder = Struct.newBuilder();
  orFilterBuilder.putFields("$or",
      Value.newBuilder()
          .getListValueBuilder()
          .addValues(Value.newBuilder().setStructValue(orgFilterBuilder.build()).build())
          .addValues(Value.newBuilder().setStructValue(projectFilterBuilder.build()).build())
          .build());

  QueryResponseWithUnsignedIndices results = index.queryByVector(
      10,
      queryVector,
      null, // default namespace
      orFilterBuilder.build(),
      false, // includeValues
      true // includeMetadata
  );
  ```

  ```go Go theme={null}
  // Better: Filter by organization or role instead of individual users
  import (
      "context"
      "fmt"
      "log"
      "github.com/pinecone-io/go-pinecone/v5/pinecone"
      "google.golang.org/protobuf/types/known/structpb"
  )

  ctx := context.Background()

  clientParams := pinecone.NewClientParams{
      ApiKey: "YOUR_API_KEY",
  }
  pc, err := pinecone.NewClient(clientParams)
  if err != nil {
      log.Fatalf("Failed to create Client: %v", err)
  }

  idx, err := pc.DescribeIndex(ctx, "your-index-name")
  if err != nil {
      log.Fatalf("Failed to describe index: %v", err)
  }

  idxConnection, err := pc.Index(pinecone.NewIndexConnParams{
      Host: idx.Host,
  })
  if err != nil {
      log.Fatalf("Failed to create IndexConnection: %v", err)
  }

  queryVector := []float32{0.1, 0.2, 0.3, ...}  // Your query vector

  metadataMap := map[string]interface{}{
      "$or": []interface{}{
          map[string]interface{}{
              "organization_id": map[string]interface{}{
                  "$eq": "org_A",
              },
          },
          map[string]interface{}{
              "project_id": map[string]interface{}{
                  "$eq": "project_B",
              },
          },
      },
  }
  filter, err := structpb.NewStruct(metadataMap)
  if err != nil {
      log.Fatalf("Failed to create filter: %v", err)
  }

  queryReq := &pinecone.QueryByVectorValuesRequest{
      Vector:          queryVector,
      TopK:            10,
      MetadataFilter:  filter,
      IncludeMetadata: true,
  }
  results, err := idxConnection.QueryByVectorValues(ctx, queryReq)
  if err != nil {
      log.Fatalf("Failed to query: %v", err)
  }
  fmt.Printf("Found %d matches:\n", len(results.Matches))
  for _, match := range results.Matches {
      fmt.Printf("  ID: %s, Score: %.4f\n", match.Vector.Id, match.Score)
      if match.Vector.Metadata != nil {
          fmt.Printf("    Metadata: %v\n", match.Vector.Metadata)
      }
  }
  ```

  ```csharp C# theme={null}
  // Better: Filter by organization or role instead of individual users
  using Pinecone;
  using System;
  using System.Linq;

  var queryVector = new[] { 0.1f, 0.2f, 0.3f, ... };  // Your query vector
  // index is your IndexClient instance

  var filter = new Metadata
  {
      {
          "$or",
          new MetadataValue(new[]
          {
              new MetadataValue(new Metadata
              {
                  {
                      "organization_id",
                      new MetadataValue(new Metadata { { "$eq", "org_A" } })
                  }
              }),
              new MetadataValue(new Metadata
              {
                  {
                      "project_id",
                      new MetadataValue(new Metadata { { "$eq", "project_B" } })
                  }
              })
          })
      }
  };

  var results = await index.QueryAsync(new QueryRequest
  {
      Vector = queryVector,
      TopK = 10,
      Filter = filter,
      IncludeMetadata = true
  });
  if (results == null)
  {
      throw new InvalidOperationException("Failed to query");
  }
  Console.WriteLine($"Found {results.Matches?.Count() ?? 0} matches:");
  if (results.Matches != null)
  {
      foreach (var match in results.Matches)
      {
          Console.WriteLine($"  ID: {match.Id}, Score: {match.Score:F4}");
          if (match.Metadata != null)
          {
              Console.WriteLine($"    Metadata: {match.Metadata}");
          }
      }
  }
  ```

  ```bash curl theme={null}
  # Better: Filter by organization or role instead of individual users
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X POST "https://$INDEX_HOST/query" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H "Content-Type: application/json" \
       -H "X-Pinecone-Api-Version: 2025-10" \
       -d '{
             "vector": [0.1, 0.2, 0.3, ...],
             "topK": 10,
             "includeMetadata": true,
             "filter": {
               "$or": [
                 {"organization_id": {"$eq": "org_A"}},
                 {"project_id": {"$eq": "project_B"}}
               ]
             }
          }'
  ```
</CodeGroup>

Instead of passing thousands of user IDs, this filter uses only 2 group identifiers to achieve the same access control.

### Multitenancy patterns

The following table provides general guidelines for choosing a multitenancy approach. Evaluate your specific use case, access patterns, and requirements to determine the best fit for your application.

| Data pattern                              | Recommended approach                                                  | Query cost                             | Performance |
| :---------------------------------------- | :-------------------------------------------------------------------- | :------------------------------------- | :---------- |
| Each tenant's data is completely separate | One index, one namespace per tenant                                   | Lowest (scans only tenant namespace)   | Fastest     |
| Large tenants with many sub-groups        | One index per large tenant, namespaces for sub-groups                 | Low (scans only sub-group namespace)   | Fast        |
| Data shared across tenants                | One index, shared namespace, filter by group IDs (org, project, role) | Higher (scans entire shared namespace) | Slower      |

<Warning>
  Avoid filtering by large lists of individual user IDs (for example, `{"user_id": {"$in": ["user_1", "user_2", ..., "user_10000"]}}`). This approach has the following drawbacks:

  * Hard limits: Each `$in` or `$nin` operator is limited to 10,000 values. Exceeding this limit will cause requests to fail.
  * Performance: Large filters increase query latency.
  * Higher costs: You pay for scanning the entire shared namespace, even though the filter narrows results.

  Instead, consider these alternatives:

  * Use one namespace per tenant (see row 1 in the table above).
  * Filter by broader groups like organization, project, or role rather than individual user IDs (see row 3 in the table above).
  * Retrieve a larger top K without filtering (for example, top 1000), then filter the results client-side.
</Warning>

For a complete step-by-step implementation guide, see [Implement multitenancy](/guides/index-data/implement-multitenancy).

---

# Source: https://docs.pinecone.io/guides/get-started/database-architecture.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Architecture

> Learn how Pinecone's architecture enables fast, relevant vector search at any scale.

## Overview

Pinecone runs as a managed service on AWS, GCP, and Azure cloud platforms. When you send a request to Pinecone, it goes through an [API gateway](#api-gateway) that routes it to either a global [control plane](#control-plane) or a regional [data plane](#data-plane). All your vector data is stored in highly efficient, distributed [object storage](#object-storage).

<img className="block max-w-full dark:hidden" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview.svg?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=98d91bc4ea7c58f6b8d35fe679b68c0b" data-og-width="740" width="740" data-og-height="360" height="360" data-path="images/serverless-overview.svg" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview.svg?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=42998453798bb153dc3fa580414f9410 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview.svg?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=d807aaf9e8dfa5e7bdc6377507517155 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview.svg?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=aa2420a330195baad94c7c718f40eb73 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview.svg?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=024214355578dbdfcd939e0a0a77a946 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview.svg?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=4dc6a2bbfb7ee1df5cbc3f93a4190fea 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview.svg?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=1629f421c3ebbcca5ea344c751cdc8a7 2500w" />

<img className="hidden max-w-full dark:block" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview-dark.svg?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=49b931899a98acb5da51441f45593969" data-og-width="740" width="740" data-og-height="360" height="360" data-path="images/serverless-overview-dark.svg" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview-dark.svg?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=c1fbcd86c10d1a42a19369505d0edde5 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview-dark.svg?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=ef25b5617fddac002e73e7b47fe49101 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview-dark.svg?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=867f8807965bfb7910ac6e20bf88b4e4 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview-dark.svg?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=361c86e1fca74d10518658b0749d7307 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview-dark.svg?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=4a57619eeb614dd8e6126ab694cd966d 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-overview-dark.svg?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=ddb67f508a79e45b8128bac633ed62d9 2500w" />

### API gateway

Every request to Pinecone includes an [API key](/guides/projects/manage-api-keys) that's assigned to a specific [project](/guides/projects/understanding-projects). The API gateway first validates your API key to make sure you have permission to access the project. Once validated, it routes your request to either the global control plane (for managing projects and indexes) or a regional data plane (for reading and writing data), depending on what you're trying to do.

### Control plane

The global control plane manages your organizational resources like projects and indexes. It uses a dedicated database to keep track of all these objects. The control plane also handles billing, user management, and coordinates operations across different regions.

### Data plane

The data plane handles all requests to write and read records in [indexes](/guides/index-data/indexing-overview) within a specific [cloud region](/guides/index-data/create-an-index#cloud-regions). Each index is divided into one or more logical [namespaces](/guides/index-data/indexing-overview#namespaces), and all your read and write requests target a specific namespace.

Pinecone separates write and read operations into different paths, with each scaling independently based on demand. This separation ensures that your queries never slow down your writes, and your writes never slow down your queries.

### Object storage

For each namespace in a serverless index, Pinecone organizes records into immutable files called slabs. These slabs are [optimized for fast querying](#index-builder) and stored in distributed object storage that provides virtually unlimited scalability and high availability.

## Write path

<img className="block max-w-full dark:hidden" noZoom src="https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path.svg?fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=ecff622da831c3a66daa97bdf2697382" data-og-width="940" width="940" data-og-height="620" height="620" data-path="images/serverless-write-path.svg" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path.svg?w=280&fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=9808682d3f6e0e495f22c5b61deff011 280w, https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path.svg?w=560&fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=28edc7e9a311ef4e033f6deb7a0cbd75 560w, https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path.svg?w=840&fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=bb437ade62614b058dd09395e8a32362 840w, https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path.svg?w=1100&fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=3b0ee72f562e2f41c08ea3d2afdcc5ba 1100w, https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path.svg?w=1650&fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=e9b1813aae24f88359114325b4ab2a79 1650w, https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path.svg?w=2500&fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=4c725c7f6ca06a6bff68e6785fb1bb70 2500w" />

<img className="hidden max-w-full dark:block" noZoom src="https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path-dark.svg?fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=d456221c3a09d5ddc2decb7753cfebc3" data-og-width="940" width="940" data-og-height="620" height="620" data-path="images/serverless-write-path-dark.svg" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path-dark.svg?w=280&fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=988867bbb19ca047d50efe6354b5cefb 280w, https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path-dark.svg?w=560&fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=a85024b967a3f3327f28d00a61d15265 560w, https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path-dark.svg?w=840&fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=d1aad53d8a276f75ab844b337b3271ba 840w, https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path-dark.svg?w=1100&fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=62c9f8af12acb68707c5876f46be9547 1100w, https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path-dark.svg?w=1650&fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=39e08abbbe141f2da52783872a7ebb25 1650w, https://mintcdn.com/pinecone/JWPNAjHjOsqAgQOo/images/serverless-write-path-dark.svg?w=2500&fit=max&auto=format&n=JWPNAjHjOsqAgQOo&q=85&s=69e381ad399e323ec6d9bebc57d78578 2500w" />

### Request log

When you send a write request (to add, update, or delete records), the [data plane](#data-plane) first logs the request details with a unique sequence number (LSN). This ensures all operations happen in the correct order and provides a way to track the state of the index.

Pinecone immediately returns a `200 OK` response, guaranteeing that your write is durable and won't be lost. The system then processes your write in the background.

### Index builder

The index builder stores your write data in an in-memory structure called a memtable. This includes your vector data, any metadata you've attached, and the sequence number. If you're updating or deleting a record, the system also tracks how to handle the old version during queries.

Periodically, the index builder moves data from the memtable to permanent storage. In [object storage](#object-storage), your data is organized into immutable files called slabs. These slabs are optimized for query performance. Smaller slabs use fast indexing techniques that provide good performance with minimal resource requirements. As slabs grow, the system merges them into larger slabs that use more sophisticated methods that provide better performance at scale. This adaptive process both optimizes query performance for each slab and amortizes the cost of more expensive indexing through the lifetime of the namespace.

<Note>
  All read operations check the memtable first, so you can immediately search data that you've just written, even before it's moved to permanent storage. For more details, see [Query executors](#query-executors).
</Note>

## Read path

<img className="block max-w-full dark:hidden" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path.svg?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=72b1e07346aed7b8d1ba57c3921671dc" data-og-width="780" width="780" data-og-height="620" height="620" data-path="images/serverless-read-path.svg" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path.svg?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=c79e043ee7f88f12f84c203004bc76f2 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path.svg?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=d8621e6936b3e164256f77ed60a26e96 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path.svg?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=b1fdeb25ef032032553a0ef1b39e47ee 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path.svg?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=8b53c093092386d6519fb38b4051caa5 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path.svg?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=74f3a707be64e5ad09cf059ebe85195a 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path.svg?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=cff6162f9c68c34998e3771f4303c6a6 2500w" />

<img className="hidden max-w-full dark:block" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path-dark.svg?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=550f3a697b7158fe3d424c70df7c8242" data-og-width="780" width="780" data-og-height="620" height="620" data-path="images/serverless-read-path-dark.svg" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path-dark.svg?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=eaaa5125f10a76ba2d40b27ccfd3daf7 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path-dark.svg?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=9a3e9166c8c861e804f02e505984975b 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path-dark.svg?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=9c6ec5b4fe6e7b94ae312c45cd48f051 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path-dark.svg?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=c1f67150c9832d7f8b9701485c912a24 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path-dark.svg?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=f019f8d47b3c908021c20d0cb0325a3f 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/serverless-read-path-dark.svg?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=1d475b0bbef6850adb84b6993acadc5e 2500w" />

### Query routers

When you send a search query, the [data plane](#data-plane) first validates your request and checks that it meets system limits like [rate and object limits](/reference/api/database-limits.mdx). The query router then identifies which slabs contain relevant data and routes your query to the appropriate executors. It also searches the memtable for any recent data that hasn't been moved to permanent storage yet.

### Query executors

Each query executor searches through its assigned slabs and returns the most relevant candidates to the query router. If your query includes metadata filters, the executors exclude records that don't match your criteria before finding the best matches.

Most of the time, the slabs are cached in memory or on local SSD, which provides very fast query performance. If a slab isn't cached (which happens when it's accessed for the first time or hasn't been used recently), the executor fetches it from object storage and caches it for future queries.

The query router then combines results from all executors, removes duplicates, merges them with results from the memtable, and returns the final set of best matches to you.

---

# Source: https://docs.pinecone.io/reference/api/database-limits.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Pinecone Database limits

This page describes different types of limits for Pinecone Database.

## Rate limits

Rate limits help protect your applications from misuse and maintain the health of our shared serverless infrastructure. These limits are designed to support typical production workloads while ensuring reliable performance for all users.

**Most rate limits can be adjusted upon request.** If you need higher limits to scale your application, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket) with details about your use case. Pinecone is committed to supporting your growth and can often accommodate higher throughput requirements.

Rate limits vary based on [pricing plan](https://www.pinecone.io/pricing/) and apply to [serverless indexes](/guides/index-data/indexing-overview) only.

<Note>
  Indexes built on [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) are not subject to read unit limits for query, fetch, and list operations. For sizing and capacity planning guidance, see the [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) guide.
</Note>

### Data plane operations: request-per-second limits

Pinecone enforces rate limits on the number of API requests per second at the namespace level for data plane operations (query, upsert, delete, and update). These limits provide protection against excessive request rates.

#### Affected operations

The following operations are subject to request-per-second rate limiting:

| Operation | Scope         | Limit |
| --------- | ------------- | ----- |
| Query     | Per namespace | 100   |
| Upsert    | Per namespace | 100   |
| Delete    | Per namespace | 100   |
| Update    | Per namespace | 100   |

#### Error response

When you exceed the request-per-second limit, you'll receive an HTTP `429 - TOO_MANY_REQUESTS` response. The error message indicates which operation exceeded the limit and includes the namespace name and limit value. See the individual limit sections below for specific error message formats.

#### How request-per-second limits work with limits on read and write units

Request-per-second limits are enforced in addition to existing read unit and write unit limits. Requests must not exceed any applicable limits:

* Index-level limits - read and write unit limits, per index
* Namespace-level limits - read and write unit limits, per namespace
* Request-per-second limits - requests per second, per namespace

If any limit is exceeded, the request fails with a 429 error.

#### Recommendations

If you're hitting request-per-second limits:

1. Implement retry logic. Use exponential backoff to handle rate limit errors gracefully. See [Error Handling Guide](/guides/production/error-handling#implement-retry-logic).
2. Pace your requests. Add client-side rate limiting to stay under limits.
3. Consider [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes), which don't have request-per-second limits and provide dedicated capacity for high-throughput workloads.
4. If you need higher limits, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket) with details about your use case.

### All rate limits

| Metric                                                                                                        | Starter plan   | Standard plan  | Enterprise plan |
| :------------------------------------------------------------------------------------------------------------ | :------------- | :------------- | :-------------- |
| [Read units per month per project](#read-units-per-month-per-project)                                         | 1,000,000      | Unlimited      | Unlimited       |
| [Write units per month per project](#write-units-per-month-per-project)                                       | 2,000,000      | Unlimited      | Unlimited       |
| [Upsert size per second per namespace](#upsert-size-per-second-per-namespace)                                 | 50 MB          | 50 MB          | 50 MB           |
| [Query read units per second per index](#query-read-units-per-second-per-index)                               | 2,000          | 2,000          | 2,000           |
| [Query requests per second per namespace](#query-requests-per-second-per-namespace)                           | 100            | 100            | 100             |
| [Update records per second per namespace](#update-records-per-second-per-namespace)                           | 100            | 100            | 100             |
| [Update requests per second per namespace](#update-requests-per-second-per-namespace)                         | 100            | 100            | 100             |
| [Update by metadata requests per second per namespace](#update-by-metadata-requests-per-second-per-namespace) | 5              | 5              | 5               |
| [Update by metadata requests per second per index](#update-by-metadata-requests-per-second-per-index)         | 500            | 500            | 500             |
| [Upsert requests per second per namespace](#upsert-requests-per-second-per-namespace)                         | 100            | 100            | 100             |
| [Fetch requests per second per index](#fetch-requests-per-second-per-index)                                   | 100            | 100            | 100             |
| [List requests per second per index](#list-requests-per-second-per-index)                                     | 200            | 200            | 200             |
| [Describe index stats requests per second per index](#describe-index-stats-requests-per-second-per-index)     | 100            | 100            | 100             |
| [Delete requests per second per namespace](#delete-requests-per-second-per-namespace)                         | 100            | 100            | 100             |
| [Delete records per second per namespace](#delete-records-per-second-per-namespace)                           | 5,000          | 5,000          | 5,000           |
| [Delete records per second per index](#delete-records-per-second-per-index)                                   | 5,000          | 5,000          | 5,000           |
| [Delete by metadata requests per second per namespace](#delete-by-metadata-requests-per-second-per-namespace) | 5              | 5              | 5               |
| [Delete by metadata requests per second per index](#delete-by-metadata-requests-per-second-per-index)         | 500            | 500            | 500             |
| [Embedding tokens per minute per model](#embedding-tokens-per-minute-per-model)                               | Model-specific | Model-specific | Model-specific  |
| [Embedding tokens per month per model](#embedding-tokens-per-month-per-model)                                 | 5,000,000      | Unlimited      | Unlimited       |
| [Rerank requests per minute per model](#rerank-requests-per-minute-per-model)                                 | Model-specific | Model-specific | Model-specific  |
| [Rerank requests per month per model](#rerank-requests-per-month-per-model)                                   | 500            | Model-specific | Model-specific  |

### Read units per month per project

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 1,000,000    | Unlimited     | Unlimited       |

[Read units](/guides/manage-cost/understanding-cost#read-units) measure the compute, I/O, and network resources used by [fetch](/guides/manage-data/fetch-data), [query](/guides/search/search-overview), and [list](/guides/manage-data/list-record-ids) requests to serverless indexes. When you reach the monthly read unit limit for a project, fetch, query, and list requests to serverless indexes in the project will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached your read unit limit for the current month limit. 
To continue reading data, upgrade your plan. 
```

To continue reading from serverless indexes in the project, [upgrade your plan](/guides/organizations/manage-billing/upgrade-billing-plan).

To check how close you are to the monthly read unit limit for a project, do the following:

1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).
2. Select the project.
3. Select any index in the project.
4. Look under **Starter Usage**.

<Note>
  Indexes built on [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) are not subject to read unit limits for query, fetch, and list operations. For sizing and capacity planning guidance, see the [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) guide.
</Note>

### Write units per month per project

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 2,000,000    | Unlimited     | Unlimited       |

[Write units](/guides/manage-cost/understanding-cost#write-units) measure the storage and compute resources used by [upsert](/guides/index-data/upsert-data), [update](/guides/manage-data/update-data), and [delete](/guides/manage-data/delete-data) requests to serverless indexes. When you reach the monthly write unit limit for a project, upsert, update, and delete requests to serverless indexes in the project will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached your write unit limit for the current month. 
To continue writing data, upgrade your plan.
```

To continue writing data to serverless indexes in the project, [upgrade your plan](/guides/organizations/manage-billing/upgrade-billing-plan).

To check how close you are to the monthly read unit limit for a project, do the following:

1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).
2. Select the project.
3. Select any index in the project.
4. Look under **Starter Usage**.

### Upsert size per second per namespace

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 50 MB        | 50 MB         | 50 MB           |

When you reach the per second [upsert](/guides/index-data/upsert-data) size for a namespace in an index, additional upserts will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max upsert size limit per second for index <index name>. 
Pace your upserts or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Query read units per second per index

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 2,000        | 2,000         | 2,000           |

Pinecone measures [query](/guides/search/search-overview) usage in [read units](/guides/manage-cost/understanding-cost#read-units). When you reach the per second limit for queries across all namespaces in an index, additional queries will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max query read units per second for index <index name>. 
Pace your queries or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

To check how many read units a query consumes, [check the query response](/guides/manage-cost/monitor-usage-and-costs#read-units).

<Note>
  Indexes built on [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) are not subject to read unit limits for query, fetch, and list operations. For sizing and capacity planning guidance, see the [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) guide.
</Note>

### Query requests per second per namespace

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 100          | 100           | 100             |

When you reach the per second [query](/guides/search/search-overview) limit for a namespace in an index, additional queries will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the query QPS limit for namespace {namespace_name} ({limit} QPS). Pace your queries,
consider Dedicated Read Nodes for your index, or contact Pinecone Support
(https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

<Note>
  Indexes built on [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) are not subject to read unit limits for query, fetch, and list operations. For sizing and capacity planning guidance, see the [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) guide.
</Note>

### Update records per second per namespace

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 100          | 100           | 100             |

When you reach the per second [update](/guides/manage-data/update-data) limit for a namespace in an index, additional updates will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max update records per second for namespace <namespace name>. 
Pace your update requests or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Update requests per second per namespace

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 100          | 100           | 100             |

When you reach the per second [update](/guides/manage-data/update-data) request limit for a namespace in an index, additional update requests will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the update QPS limit for namespace {namespace_name} ({limit} QPS). Pace your update requests or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Update by metadata requests per second per namespace

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 5            | 5             | 5               |

When you reach the per second [update by metadata](/guides/manage-data/update-data#update-metadata-across-multiple-records) request limit for a namespace in an index, additional update by metadata requests will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max update by metadata requests per second for namespace <namespace name>. Pace your update by metadata requests or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Update by metadata requests per second per index

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 500          | 500           | 500             |

When you reach the per second [update by metadata](/guides/manage-data/update-data#update-metadata-across-multiple-records) request limit across all namespaces in an index, additional update by metadata requests will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max update by metadata requests per second for index <index name>. Pace your update by metadata requests or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Upsert requests per second per namespace

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 100          | 100           | 100             |

When you reach the per second [upsert](/guides/index-data/upsert-data) request limit for a namespace in an index, additional upsert requests will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the upsert QPS limit for namespace {namespace_name} ({limit} QPS). Pace your upsert requests or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Fetch requests per second per index

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 100          | 100           | 100             |

When you reach the per second [fetch](/guides/manage-data/fetch-data) limit across all namespaces in an index, additional fetch requests will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max fetch requests per second for index <index name>.
Pace your fetch requests or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

<Note>
  Indexes built on [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) are not subject to read unit limits for query, fetch, and list operations. For sizing and capacity planning guidance, see the [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) guide.
</Note>

### List requests per second per index

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 200          | 200           | 200             |

When you reach the per second [list](/guides/manage-data/list-record-ids) limit across all namespaces in an index, additional list requests will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max list requests per second for index <index name>.
Pace your list requests or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

<Note>
  Indexes built on [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) are not subject to read unit limits for query, fetch, and list operations. For sizing and capacity planning guidance, see the [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) guide.
</Note>

### Describe index stats requests per second per index

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 100          | 100           | 100             |

When you reach the per second [describe index stats](/reference/api/2024-10/data-plane/describeindexstats) limit across all namespaces in an index, additional list requests will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max describe_index_stats requests per second for index <index>. 
Pace your describe_index_stats requests or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Delete requests per second per namespace

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 100          | 100           | 100             |

When you reach the per second [delete](/guides/manage-data/delete-data) request limit for a namespace in an index, additional delete requests will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the delete QPS limit for namespace {namespace_name} ({limit} QPS). Pace your delete requests or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Delete records per second per namespace

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 5000         | 5000          | 5000            |

When you reach the per second [delete](/guides/manage-data/delete-data) limit for a namespace in an index, additional deletes will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max delete records per second for namespace <namespace name>. 
Pace your delete requests or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Delete records per second per index

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 5000         | 5000          | 5000            |

When you reach the per second [delete](/guides/manage-data/delete-data) limit across all namespaces in an index, additional deletes will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max delete records per second for index <index name>. 
Pace your delete requests or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Delete by metadata requests per second per namespace

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 5            | 5             | 5               |

When you reach the per second [delete by metadata](/guides/manage-data/delete-data#delete-records-by-metadata) request limit for a namespace in an index, additional delete by metadata requests will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max delete by metadata requests per second for namespace <namespace name>. Pace your delete by metadata requests or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Delete by metadata requests per second per index

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 500          | 500           | 500             |

When you reach the per second [delete by metadata](/guides/manage-data/delete-data#delete-records-by-metadata) request limit across all namespaces in an index, additional delete by metadata requests will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max delete by metadata requests per second for index <index name>. Pace your delete by metadata requests or contact Pinecone Support (https://app.pinecone.io/organizations/-/settings/support/ticket) to request a higher limit.
```

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). If you need a higher limit for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Embedding tokens per minute per model

| Embedding model              | Input type | Starter plan | Standard plan | Enterprise plan |
| :--------------------------- | :--------- | :----------- | :------------ | :-------------- |
| `llama-text-embed-v2`        | Passage    | 250,000      | 1,000,000     | 1,000,000       |
|                              | Query      | 50,000       | 250,000       | 250,000         |
| `multilingual-e5-large`      | Passage    | 250,000      | 1,000,000     | 1,000,000       |
|                              | Query      | 50,000       | 250,000       | 250,000         |
| `pinecone-sparse-english-v0` | Passage    | 250,000      | 3,000,000     | 3,000,000       |
|                              | Query      | 250,000      | 3,000,000     | 3,000,000       |

When you reach the per minute token limit for an [embedding model](/guides/index-data/create-an-index#embedding-models) hosted by Pinecone, additional embeddings will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max embedding tokens per minute (<limit>) model '<model name>'' and input type '<passage|query>' for the current project. 
To increase this limit, upgrade your plan.
```

To increase this limit, [upgrade your plan](/guides/organizations/manage-billing/upgrade-billing-plan). Otherwise, you can handle this limit by [implementing retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic).

### Embedding tokens per month per model

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 5,000,000    | Unlimited     | Unlimited       |

When you reach the monthly token limit for an [embedding model](/guides/index-data/create-an-index#embedding-models) hosted by Pinecone, additional embeddings will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the embedding token limit (<limit>) for model <model name> for the current month. 
To continue using this model, upgrade your plan.
```

To increase this limit, [upgrade your plan](/guides/organizations/manage-billing/upgrade-billing-plan) or [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Rerank requests per minute per model

| Reranking model      | Starter plan  | Standard plan | Enterprise plan |
| :------------------- | :------------ | :------------ | :-------------- |
| `cohere-rerank-3.5`  | Not available | 300           | 300             |
| `bge-reranker-v2-m3` | 60            | 60            | 60              |
| `pinecone-rerank-v0` | 60            | 60            | 60              |

When you reach the per minute request limit for a [reranking model](/guides/search/rerank-results#reranking-models) hosted by Pinecone, additional reranking requests will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max rerank requests per minute (<limit>) for model '<model name>' for the current project. 
To increase this limit, upgrade your plan.
```

To increase this limit, [upgrade your plan](/guides/organizations/manage-billing/upgrade-billing-plan).

### Rerank requests per month per model

| Reranking model      | Starter plan  | Standard plan | Enterprise plan |
| :------------------- | :------------ | :------------ | :-------------- |
| `cohere-rerank-3.5`  | Not available | Unlimited     | Unlimited       |
| `bge-reranker-v2-m3` | 500           | Unlimited     | Unlimited       |
| `pinecone-rerank-v0` | 500           | Unlimited     | Unlimited       |

When you reach the monthly request limit for a [reranking model](/guides/search/rerank-results#reranking-models) hosted by Pinecone, additional reranking requests will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the rerank request limit (<limit>) for model <model name> for the current month. 
To continue using this model, upgrade your plan.
```

To increase this limit, [upgrade your plan](/guides/organizations/manage-billing/upgrade-billing-plan) or [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Inference requests per second or minute, per project

| Metric                        | Starter plan | Standard plan | Enterprise plan |
| :---------------------------- | :----------- | :------------ | :-------------- |
| Inference requests per second | 100          | 100           | 100             |
| Inference requests per minute | 2000         | 2000          | 2000            |

When you reach the per second or per minute request limit, inference requests will fail and return a `429 - TOO_MANY_REQUESTS` status with the following error:

```
Request failed. You've reached the max inference requests per second (<limit>) for the current project.
```

<Note>
  This error indicates per second or per minute, as applicable.
</Note>

To handle this limit, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic).

## Object limits

Object limits are restrictions on the number or size of objects in Pinecone. Object limits vary based on [pricing plan](https://www.pinecone.io/pricing/).

| Metric                                                                         | Starter plan | Standard plan | Enterprise plan |
| :----------------------------------------------------------------------------- | :----------- | :------------ | :-------------- |
| [Projects per organization](#projects-per-organization)                        | 1            | 20            | 100             |
| [Serverless indexes per project](#serverless-indexes-per-project) <sup>1</sup> | 5            | 20            | 200             |
| [Serverless index storage per project](#serverless-index-storage-per-project)  | 2 GB         | N/A           | N/A             |
| [Namespaces per serverless index](#namespaces-per-serverless-index)            | 100          | 100,000       | 100,000         |
| [Serverless backups per project](#serveless-backups-per-project)               | N/A          | 500           | 1000            |
| [Collections per project](#collections-per-project)                            | 100          | N/A           | N/A             |

<sup>1 On the Starter plan, all serverless must be in the `us-east-1` region of AWS.</sup><br />

### Projects per organization

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 1            | 20            | 100             |

When you reach this quota for an organization, trying to [create projects](/guides/projects/create-a-project) will fail and return a `403 - QUOTA_EXCEEDED` status with the following error:

```
Request failed. You've reached the max projects allowed in organization <org name>. 
To add more projects, upgrade your plan.
```

To increase this quota, [upgrade your plan](/guides/organizations/manage-billing/upgrade-billing-plan) or [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

### Serverless indexes per project

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 5            | 20            | 200             |

When you reach this quota for a project, trying to [create serverless indexes](/guides/index-data/create-an-index#create-a-serverless-index) in the project will fail and return a `403 - QUOTA_EXCEEDED` status with the following error:

```
Request failed. You've reached the max serverless indexes allowed in project <project>. 
Use namespaces to partition your data into logical groups, or upgrade your plan to add more serverless indexes.
```

To stay under this quota, consider using [namespaces](/guides/index-data/create-an-index#namespaces) instead of creating multiple indexes. Namespaces let you partition your data into logical groups within a single index. This approach not only helps you stay within index limits, but can also improve query performance and lower costs by limiting searches to relevant data subsets.

To increase this quota, [upgrade your plan](/guides/organizations/manage-billing/upgrade-billing-plan).

### Serverless index storage per project

<Note>This limit applies to organizations on the Starter plan only.</Note>

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 2 GB         | N/A           | N/A             |

When you've reached this quota for a project, updates and upserts into serverless indexes will fail and return a `403 - QUOTA_EXCEEDED` status with the following error:

```
Request failed. You've reached the max storage allowed for project <project name>. 
To update or upsert new data, delete records or upgrade your plan.
```

To continue writing data into your serverless indexes, [delete records](/guides/manage-data/delete-data) to bring your project under the limit or [upgrade your plan](/guides/organizations/manage-billing/upgrade-billing-plan).

### Namespaces per serverless index

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 100          | 100,000       | 100,000         |

When you reach this quota for a serverless index, trying to [upsert records into a new namespace](/guides/index-data/upsert-data) in the index will fail and return a `403 - QUOTA_EXCEEDED` status with the following error:

```
Request failed. You've reached the max namespaces allowed in serverless index <index name>. 
To add more namespaces, upgrade your plan.
```

To increase this quota, [upgrade your plan](/guides/organizations/manage-billing/upgrade-billing-plan).

<Note>
  While the Standard and Enterprise plans support up to [100,000 namespaces per index](/reference/api/database-limits#namespaces-per-serverless-index), Pinecone can accommodate million-scale namespaces and beyond for specific use cases. If your application requires more than 100,000 namespaces, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).
</Note>

### Serverless backups per project

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| N/A          | 500           | 1000            |

When you reach this quota for a project, trying to [create serverless backups](/guides/manage-data/back-up-an-index) in the project will fail and return a `403 - QUOTA_EXCEEDED` status with the following error:

```
Backup failed to create. Quota for number of backups per index exceeded.
```

### Collections per project

| Starter plan | Standard plan | Enterprise plan |
| ------------ | ------------- | --------------- |
| 100          | N/A           | N/A             |

When you reach this quota for a project, trying to [create collections](/guides/manage-data/back-up-an-index) in the project will fail and return a `403 - QUOTA_EXCEEDED` status with the following error:

```
Request failed. You've reached the max collections allowed in project <project name>. 
To add more collections, upgrade your plan.
```

To increase this quota, [upgrade your plan](/guides/organizations/manage-billing/upgrade-billing-plan).

## Operation limits

Operation limits are restrictions on the size, number, or other characteristics of operations in Pinecone. Operation limits are fixed and do not vary based on pricing plan.

### Upsert limits

| Metric                                                             | Limit                                                         |
| :----------------------------------------------------------------- | :------------------------------------------------------------ |
| Max [batch size](/guides/index-data/upsert-data#upsert-in-batches) | 2 MB or 1000 records with vectors <br /> 96 records with text |
| Max metadata size per record                                       | 40 KB                                                         |
| Max length for a record ID                                         | 512 characters                                                |
| Max dimensionality for dense vectors                               | 20,000                                                        |
| Max non-zero values for sparse vectors                             | 2048                                                          |
| Max dimensionality for sparse vectors                              | 4.2 billion                                                   |

### Import limits

<Note>
  If your import exceeds these limits, you'll get an `Exceeds system limit` error. Pinecone can help unblock these imports quickly. [Contact Pinecone support](https://app.pinecone.io/organizations/-/settings/support/ticket) for assistance.
</Note>

| Metric                    | Limit   |
| :------------------------ | :------ |
| Max namespaces per import | 10,000  |
| Max size per namespace    | 500 GB  |
| Max files per import      | 100,000 |
| Max size per file         | 10 GB   |

### Query limits

| Metric            | Limit  |
| :---------------- | :----- |
| Max `top_k` value | 10,000 |
| Max result size   | 4MB    |

The query result size is affected by the dimension of the dense vectors and whether or not dense vector values and metadata are included in the result.

<Tip>
  If a query fails due to exceeding the 4MB result size limit, choose a lower `top_k` value, or use `include_metadata=False` or `include_values=False` to exclude metadata or values from the result. For better performance, especially with higher `top_k` values, avoid including vector values unless you need them.
</Tip>

### Fetch limits

| Metric                           | Limit |
| :------------------------------- | :---- |
| Max record IDs per fetch request | 1,000 |

### Delete limits

| Metric                            | Limit |
| :-------------------------------- | :---- |
| Max record IDs per delete request | 1,000 |

### Metadata filter limits

The following limits apply to [metadata filter expressions](/guides/search/filter-by-metadata#metadata-filter-expressions) used in query, delete, update, and fetch operations.

| Limit                                       | Value  | Description                                                                                                                                                                                                      |
| :------------------------------------------ | :----- | :--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Maximum values per `$in` or `$nin` operator | 10,000 | Each `$in` or `$nin` operator accepts up to 10,000 values in its array. This limit applies per operator—if you have multiple `$in` operators in a single filter, each is independently limited to 10,000 values. |

When you exceed this limit, the request will fail and return a `400 - BAD_REQUEST` error.

#### Rationale

Large `$in` operators can impact query performance and cost. Filters with thousands of values increase request payload size and end-to-end latency. Additionally, using large filters typically indicates a shared namespace architecture, which increases query costs—queries scan the entire namespace regardless of filters.

#### Alternative approaches

If you need to filter by more than 10,000 values, consider these alternatives:

* **Use namespaces for tenant isolation**: Instead of filtering by tenant IDs within a single namespace, create separate namespaces for each tenant or tenant group. This can also reduce query costs. See [Design for multi-tenancy](/guides/index-data/data-modeling#design-for-multi-tenancy).
* **Use broader access control groups**: Instead of filtering by individual user IDs, filter by organization, project, or role. This reduces the number of values in your `$in` filter. See [Design for multi-tenancy](/guides/index-data/data-modeling#use-access-control-groups-instead-of-individual-ids).
* **Post-filter client-side**: Retrieve a larger top K without filtering (for example, top 1000), then filter results client-side.
* **Run multiple queries**: Split your filter into multiple queries with smaller `$in` operators and combine the results client-side.

<Tip>
  To avoid hitting this limit in production, validate the size of your `$in` and `$nin` arrays in your application code before making the request to Pinecone.
</Tip>

## Identifier limits

An identifier is a string of characters used to identify "named" [objects in Pinecone](/guides/get-started/concepts). The following Pinecone objects use strings as identifiers:

| Object                                                    | Field       | Max # characters | Allowed characters                                                                                                                        |
| --------------------------------------------------------- | ----------- | ---------------- | ----------------------------------------------------------------------------------------------------------------------------------------- |
| [Organization](/guides/get-started/concepts#organization) | `name`      | 512              | <ul><li>UTF-8 except `\0`</li><li>Cannot be empty</li></ul>                                                                               |
| [Project](/guides/get-started/concepts#project)           | `name`      | 512              | <ul><li>UTF-8 except `\0`</li><li>Cannot be empty</li></ul>                                                                               |
| [Index](/guides/get-started/concepts#index)               | `name`      | 45               | <ul><li>`a-z`, `0-9`, and `-`</li><li>Must be lowercase</li><li>Cannot start or end with `-`</li><li>Cannot be empty</li></ul>            |
| [Namespace](/guides/get-started/concepts#namespace)       | `namespace` | 512              | <ul><li>ASCII except `\0`</li><li>For the default namespace, use `""` (or `"__default__"`, in API versions `2025-04` and later)</li></ul> |
| [Record](/guides/get-started/concepts#record)             | `id`        | 512              | <ul><li>ASCII except `\0`</li><li>Cannot be empty</li></ul>                                                                               |

---

# Source: https://docs.pinecone.io/integrations/databricks.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Databricks

> Using Databricks and Pinecone to create and index vector embeddings at scale

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Databricks is a Unified Analytics Platform on top of Apache Spark. The primary advantage of using Spark is its ability to distribute workloads across a cluster of machines. By adding more machines or increasing the number of cores on each machine, it is easy to horizontally scale a cluster to handle computationally intensive tasks like vector embedding, where parallelization can save many hours of precious computation time and resources. Leveraging GPUs with Spark can produce even better results — enjoying the benefits of the fast computation of a GPU combined with parallelization will ensure optimal performance.

Efficiently create, ingest, and update vector embeddings at scale with Databricks and Pinecone.

<PrimarySecondaryCTA
  primaryLabel={"Get started"}
  primaryHref={
  "/reference/tools/pinecone-spark-connector"
}
  secondaryLabel={"View setup guide"}
  secondaryHref={
  "#setup-guide"
}
/>

## Setup guide

In this guide, you'll create embeddings based on the [sentence-transformers/all-MiniLM-L6-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2) model from [Hugging Face](https://huggingface.co/), but the approach demonstrated here should work with any other model and dataset.

### Before you begin

Ensure you have the following:

* A [Databricks cluster](https://docs.databricks.com/en/compute/configure.html)
* A [Pinecone account](https://app.pinecone.io/)
* A [Pinecone API key](/guides/projects/understanding-projects#api-keys)

### 1. Install the Spark-Pinecone connector

<Tabs>
  <Tab title="Databricks platform">
    1. [Install the Spark-Pinecone connector as a library](https://docs.databricks.com/en/libraries/cluster-libraries.html#install-a-library-on-a-cluster).
    2. Configure the library as follows:
       1. Select **File path/S3** as the **Library Source**.

       2. Enter the S3 URI for the Pinecone assembly JAR file:

          ```
          s3://pinecone-jars/1.1.0/spark-pinecone-uberjar.jar  
          ```

          <Note>
            Databricks platform users must use the Pinecone assembly jar listed above to ensure that the proper dependecies are installed.
          </Note>

       3. Click **Install**.
  </Tab>

  <Tab title="Databricks on AWS">
    1. [Install the Spark-Pinecone connector as a library](https://docs.databricks.com/en/libraries/cluster-libraries.html#install-a-library-on-a-cluster).
    2. Configure the library as follows:
       1. Select **File path/S3** as the **Library Source**.

       2. Enter the S3 URI for the Pinecone assembly JAR file:

          ```
          s3://pinecone-jars/1.1.0/spark-pinecone-uberjar.jar  
          ```

       3. Click **Install**.
  </Tab>

  <Tab title="Databricks on GCP / Azure">
    1. [Install the Spark-Pinecone connector as a library](https://docs.databricks.com/en/libraries/cluster-libraries.html#install-a-library-on-a-cluster).
    2. Configure the library as follows:
       1. [Download the Pinecone assembly JAR file](https://repo1.maven.org/maven2/io/pinecone/spark-pinecone_2.12/1.1.0/).
       2. Select **Workspace** as the **Library Source**.
       3. Upload the JAR file.
       4. Click **Install**.
  </Tab>
</Tabs>

### 2. Load the dataset into partitions

As your example dataset, use a collection of news articles from Hugging Face's datasets library:

1. [Create a new notebook](https://docs.databricks.com/en/notebooks/notebooks-manage.html#create-a-notebook) attached to your cluster.

2. Install dependencies:

   ```
   pip install datasets transformers pinecone torch  
   ```

3. Load the dataset:

   ```Python Python theme={null}
   from datasets import list_datasets, load_dataset  
   dataset_name = "allenai/multinews_sparse_max"  
   dataset = load_dataset(dataset_name, split="train")  
   ```

4. Convert the dataset from the Hugging Face format and repartition it:

   ```Python Python theme={null}
   dataset.to_parquet("/dbfs/tmp/dataset_parquet.pq")  
   num_workers = 10  
   dataset_df = spark.read.parquet("/tmp/dataset_parquet.pq").repartition(num_workers)  
   ```

   Once the repartition is complete, you get back a DataFrame, which is a distributed collection of the data organized into named columns. It is conceptually equivalent to a table in a relational database or a dataframe in R/Python, but with richer optimizations under the hood. As mentioned above, each partition in the dataframe has an equal amount of the original data.

5. The dataset doesn't have identifiers associated with each document, so add them:

   ```Python Python theme={null}
   from pyspark.sql.types import StringType  
   from pyspark.sql.functions import monotonically_increasing_id  
   dataset_df = dataset_df.withColumn("id", monotonically_increasing_id().cast(StringType()))  
   ```

   As its name suggests, `withColumn` adds a column to the dataframe, containing a simple increasing identifier that you cast to a string.

### 3. Create the vector embeddings

1. Create a UDF (User-Defined Function) to create the embeddings, using the AutoTokenizer and AutoModel classes from the Hugging Face transformers library:

   ```Python Python theme={null}
   from transformers import AutoTokenizer, AutoModel  
   def create_embeddings(partitionData):  
   tokenizer = AutoTokenizer.from_pretrained("sentence-transformers/all-MiniLM-L6-v2")  
   model = AutoModel.from_pretrained("sentence-transformers/all-MiniLM-L6-v2")  
   for row in partitionData:  
       document = str(row.document)  
       inputs = tokenizer(document, padding=True, truncation=True, return_tensors="pt", max_length=512)  
       result = model(**inputs)  
       embeddings = result.last_hidden_state[:, 0, :].cpu().detach().numpy()  
       lst = embeddings.flatten().tolist()  
       yield [row.id, lst, "", "{}", None]  
   ```

2. Apply the UDF to the data:

   ```Python Python theme={null}
   embeddings = dataset_df.rdd.mapPartitions(create_embeddings)  
   ```

   A dataframe in Spark is a higher-level abstraction built on top of a more fundamental building block called a resilient distributed dataset (RDD). Here, you use the `mapPartitions` function, which provides finer control over the execution of the UDF by explicitly applying it to each partition of the RDD.

3. Convert the resulting RDD back into a dataframe with the schema required by Pinecone:

   ```Python Python theme={null}
   from pyspark.sql.types import StructType, StructField, StringType, ArrayType, FloatType, IntegerType  
   schema = StructType([  
       StructField("id",StringType(),True),  
       StructField("values",ArrayType(FloatType()),True),  
       StructField("namespace",StringType(),True),  
       StructField("metadata", StringType(), True),  
       StructField("sparse_values", StructType([  
           StructField("indices", ArrayType(LongType(), False), False),  
           StructField("values", ArrayType(FloatType(), False), False)  
       ]), True)  
   ])  
   embeddings_df = spark.createDataFrame(data=embeddings,schema=schema)  
   ```

### 4. Save the embeddings in Pinecone

1. Initialize the connection to Pinecone:

   ```Python Python theme={null}
   from pinecone.grpc import PineconeGRPC as Pinecone
   from pinecone import ServerlessSpec

   pc = Pinecone(api_key="YOUR_API_KEY")
   ```

2. Create an index for your embeddings:

   ```Python Python theme={null}
   pc.create_index(
   name="news",
   dimension=1536,
   metric="cosine",
   spec=ServerlessSpec(
       cloud="aws",
       region="us-east-1"
   )
   )
   ```

3. Use the Spark-Pinecone connector to save the embeddings to your index:

   ```Python Python theme={null}
   (  
       embeddings_df.write  
       .option("pinecone.apiKey", api_key) 
       .option("pinecone.indexName", index_name)  
       .format("io.pinecone.spark.pinecone.Pinecone")  
       .mode("append")  
       .save()  
   )  
   ```

   The process of writing the embeddings to Pinecone should take approximately 15 seconds. When it completes, you'll see the following:

   ```
   spark: org.apache.spark.sql.SparkSession = org.apache.spark.sql.SparkSession@41638051  
   pineconeOptions: scala.collection.immutable.Map[String,String] = Map(pinecone.apiKey -><YOUR API KEY>, pinecone.indexName -> "news")  
   ```

   This means the process was completed successfully and the embeddings have been stored in Pinecone.

4. Perform a similarity search using the embeddings you loaded into Pinecone by providing a set of vector values or a vector ID. The [query endpoint](/reference/api/2025-10/data-plane/query) will return the IDs of the most similar records in the index, along with their similarity scores:
   ```Python Python theme={null}
       index.query(
           namespace="example-namespace",
           vector=[0.3, 0.3, 0.3, 0.3, ... 0.3],
           top_k=3,
           include_values=True
       )
   ```
   <Note>
     If you want to make a query with a text string (e.g., `"Summarize this article"`), use the [`search` endpoint via integrated inference](/reference/api/2025-10/data-plane/search_records).
   </Note>

---

# Source: https://docs.pinecone.io/integrations/datadog.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Datadog

> Monitoring Pinecone with Datadog

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

<Note>
  This feature is available on [Standard and Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

Datadog is a monitoring and analytics tool that can be used to determine performance metrics as well as event monitoring for infrastructure and cloud services. Use Datadog to:

* Optimize performance and control usage: Observe and track specific actions (e.g., request count) within Pinecone to identify application requests with high latency or usage. Monitor trends and gain actionable insights to improve resource utilization and reduce spend.
* Automatically alert on metrics: Get alerted when index fullness reaches a certain threshold. You can also create your own customized monitors to alert on specific metrics and thresholds.
* Locate and triage unexpected spikes in usage or latency: Quickly visualize anomalies in usage or latency in Pinecone's Datadog dashboard. View metrics over time to better understand trends and determine the severity of a spike.

<PrimarySecondaryCTA secondaryHref={"#setup-guide"} secondaryLabel={"View setup guide"} />

## Setup guide

Follow these steps to monitor a Pinecone project with Datadog:

1. Go to the [Pinecone integration](https://app.datadoghq.com/integrations/pinecone) tile in Datadog.
2. Go to the **Configure** tab.
3. Click **+ Add New**.
4. Enter a project name to identify your project in Datadog.
5. Do not select an environment. This is a legacy setting.
6. Enter an [API key](/guides/projects/understanding-projects#api-keys) for the Pinecone project you want to monitor.
7. Enter the [project ID](/guides/projects/understanding-projects#project-ids) of the Pinecone project you want to monitor.
8. Save the configuration.

On the **Monitoring Resources** tab, you'll find dashboards for the pod-based and serverless indexes in your project and recommendations for [configuring monitors](https://docs.datadoghq.com/monitors/configuration/?tab=thresholdalert) using [Pinecone's metrics](/guides/production/monitoring#available-metrics).

---

# Source: https://docs.pinecone.io/integrations/datavolo.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Datavolo

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

[Datavolo](https://datavolo.io/) helps data teams build multimodal data pipelines to support their AI initiatives. Every organization has their own private data that they need to incorporate into their AI apps, and a predominant pattern to do so has emerged: retrieval augmented generation (RAG).

Datavolo sources, transforms, and enriches data in a continuous, composable and customizable manner, landing the data in Pinecone for retrieval. This ensures organizations can securely access their unstructured data.

<PrimarySecondaryCTA primaryLabel="Get started" primaryHref="https://docs.datavolo.io/docs/product/Integrations/pinecone-setup" />

---

# Source: https://docs.pinecone.io/troubleshooting/debug-model-vs-pinecone-recall-issues.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Debug model vs. Pinecone recall issues

## **Step 1: Establish the evaluation framework**

Before starting, establish an evaluation framework for your model and Pinecone recall issues. You will need to query a dataset of at least 10 samples and a source dataset of 100k samples, and choose an evaluation metric that is appropriate for your use case. Pinecone recommends [Evaluation Measures in Information Retrieval](https://www.pinecone.io/learn/offline-evaluation/) as a guide for choosing an evaluation metric. Label the "right answers" in the source dataset for each query.

## **Step 2: Generate embeddings for queries + source dataset with the model**

Use your model to generate embeddings for your queries and the source dataset. Run the model on the source dataset to create the vector dataset and query vectors.

## **Step 3: Calculate brute force vector distance to evaluate model quality**

Run a brute force search using query vectors over the vector dataset via FAISS or numpy and record the record IDs for each query. Evaluate the returned list using your evaluation metric and the set of "right answers" labeled in step 1. If this metric is unacceptable, it indicates a model issue.

## **Step 4: Upload vector dataset to Pinecone + query**

Upload the vector dataset to Pinecone and query it using your queries. Record the vector IDs returned for each query.

## **Step 5: Calculate Pinecone recall**

For each query, compare the % of vector IDs that Pinecone recalled compared to the brute force search. This will be the % recall for each query. You can then average across all queries to get average recall. Typically, average recall should be close to 0.99 for s1/p1 indexes.

## **Step 6: If recall is too low, reach out to Pinecone Support (reproducible dataset and queries)**

If the recall metric is too low for your use case, reach out to Pinecone product and engineering with the query and vector dataset that reproduces the issue for further investigation. Pinecone's team will investigate.

---

# Source: https://docs.pinecone.io/guides/optimize/decrease-latency.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Decrease latency

> Learn techniques to decrease latency for search and upsert operations.

## Use namespaces

When you divide records into [namespaces](/guides/index-data/indexing-overview#namespaces) in a logical way, you speed up queries by ensuring only relevant records are scanned. The same applies to [fetching records](/guides/manage-data/fetch-data), [listing record IDs](/guides/manage-data/list-record-ids), and other data operations.

## Filter by metadata

In addition to increasing search accuracy and relevance, [searching with metadata filters](/guides/search/filter-by-metadata) can also help decrease latency by retrieving only records that match the filter.

## Target indexes by host

When you target an index by name for data operations such as `upsert` and `query`, the SDK gets the unique DNS host for the index using the `describe_index` operation. This is convenient for testing but should be avoided in production because `describe_index` uses a different API than data operations and therefore adds an additional network call and point of failure. Instead, you should get an index host once and cache it for reuse or specify the host directly.

You can get index hosts in the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/indexes) or using the [`describe_index`](/guides/manage-data/manage-indexes#describe-an-index) operation.

The following example shows how to target an index by host directly:

<Note>
  When using Private Endpoints for private connectivity between your application and Pinecone, you must target the index using the [Private Endpoint URL](/guides/production/connect-to-aws-privatelink#run-data-plane-commands) for the host.
</Note>

<CodeGroup>
  ```Python Python {5} theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  index = pc.Index(host="INDEX_HOST")
  ```

  ```javascript JavaScript {6} theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  // For the Node.js SDK, you must specify both the index host and name.
  const index = pc.index("INDEX_NAME", "INDEX_HOST");
  ```

  ```java Java {11} theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;

  public class TargetIndexByHostExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          // For the Java SDK, you must specify both the index host and name.
          Index index = new Index(connection, "INDEX_NAME");
      }
  }
  ```

  ```go Go {21} theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host %v: %v", idx.Host, err)
      }
  }
  ```

  ```csharp C# {5} theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var index = pinecone.Index(host: "INDEX_HOST");
  ```
</CodeGroup>

## Reuse connections

When you target an index for upserting or querying, the client establishes a TCP connection, which is a three-step process. To avoid going through this process on every request, and reduce average request latency, [cache and reuse the index connection object](/reference/api/authentication#initialize-a-client) whenever possible.

## Use a cloud environment

If you experience slow uploads or high query latencies, it might be because you are accessing Pinecone from your home network. To decrease latency, access Pinecone/deploy your application from a cloud environment instead, ideally from the same [cloud and region](/guides/index-data/create-an-index#cloud-regions) as your index.

## Avoid including vector values when not needed

For on-demand indexes, since vector values are retrieved from object storage, including vector values in query responses (`include_values=true`) adds latency, especially with higher `top_k` values. If you don't need the vector values in your response, set `include_values=false` to improve query performance. This applies to [`query`](/reference/api/latest/data-plane/query) and [`fetch`](/reference/api/latest/data-plane/fetch) operations.

<Note>
  This optimization applies to on-demand indexes. DRN indexes cache values locally and are not affected.
</Note>

## Work with database limits

Pinecone has [rate limits](/reference/api/database-limits#rate-limits) to protect your applications and maintain infrastructure health. Rate limits vary based on pricing plan and apply to serverless indexes only.

<Note>
  Indexes built on [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) are not subject to read unit limits for query, fetch, and list operations. For sizing and capacity planning guidance, see the [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) guide.
</Note>

To handle rate limits effectively:

* [Implement retry logic with exponential backoff](/guides/production/error-handling#handle-rate-limits-429).
* If you need higher limits for your use case, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket). Most limits can be adjusted to accommodate your scaling needs.

---

# Source: https://docs.pinecone.io/guides/index-data/dedicated-read-nodes.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Dedicated Read Nodes

> Dedicated read nodes use provisioned hardware for read operations, providing predictable, low-latency performance at high query volumes.

<Note>
  This feature is in [public preview](/release-notes/feature-availability).
</Note>

<Callout icon="calendar" color="#16A34A" iconType="regular">
  Join us for [**Solving for Size and Speed**](https://lu.ma/event/evt-wKqJf1SVpEW95Zv), a 60-minute webinar on Wednesday, February 25, to learn how Dedicated Read Nodes deliver consistent throughput and predictable low latency.
</Callout>

## Overview

Pinecone indexes built on dedicated read nodes use provisioned read hardware to provide predictable, consistent performance at sustained, high query volumes. They're designed for large-scale vector workloads such as semantic search, recommendation engines, and mission-critical services.

Dedicated read nodes differ from on-demand indexes in how they handle read operations. While on-demand indexes use shared, multi-tenant capacity for reads, dedicated read nodes provision exclusive hardware for reads—memory, local SSDs, and compute. Both index types use Pinecone's serverless infrastructure for writes and storage.

When you create a dedicated read nodes index, Pinecone provisions resources based on your choice of [node type](#node-types), number of [shards](#shards), and number of [replicas](#replicas). These resources include local SSDs and memory that cache all your index data, and provide dedicated query executors to handle read operations (query, fetch, list). This architecture eliminates cold starts and ensures consistent low-latency performance, even under heavy load.

Dedicated read nodes support dense, sparse, or hybrid indexes, giving you flexibility in your search and retrieval strategy. Because storage (shards) and compute (replicas) scale independently, you can optimize for your specific workload characteristics.

<Frame caption="Read path for dedicated read nodes">
  <img className="block max-w-full dark:hidden" noZoom src="https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_light.png?fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=d139d1364ec4349c0028f6036769a61a" data-og-width="2644" width="2644" data-og-height="2424" height="2424" data-path="images/drn-read-path_light.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_light.png?w=280&fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=7ef1f770fe848a8b1918c628337b77a2 280w, https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_light.png?w=560&fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=6a0d4ce47fbadc8c41422f7e005c8b3e 560w, https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_light.png?w=840&fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=a11a20221fee5d49941bcc54dd83c585 840w, https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_light.png?w=1100&fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=23d20a9afa3ca750c1f4f2d4d069d170 1100w, https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_light.png?w=1650&fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=89a841ed98525c6c9a5c4ecad5ead82c 1650w, https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_light.png?w=2500&fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=efdefccb8b01946f3a7dc3a1fa3172e7 2500w" />

  <img className="hidden max-w-full dark:block" noZoom src="https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_dark.png?fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=b9c36be64e8b3803db2536eaa0a5e29e" data-og-width="2644" width="2644" data-og-height="2424" height="2424" data-path="images/drn-read-path_dark.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_dark.png?w=280&fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=869d14475bb485a3f02735bce67b078c 280w, https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_dark.png?w=560&fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=465fdd9ccc654440b3d80ce5cafc8049 560w, https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_dark.png?w=840&fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=c16f6cc75f0bd738a34693d2dda105fe 840w, https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_dark.png?w=1100&fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=ff8e223f08b47dcbfabd2a213915bb2e 1100w, https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_dark.png?w=1650&fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=ee1ef2f97e47a743f076e0734eb28eaf 1650w, https://mintcdn.com/pinecone/id0yJFEuv83Kq3F9/images/drn-read-path_dark.png?w=2500&fit=max&auto=format&n=id0yJFEuv83Kq3F9&q=85&s=69f7f3e080b9db4c378cf347dbc76256 2500w" />
</Frame>

## On-demand vs dedicated

On-demand indexes and dedicated read nodes are both built on Pinecone's serverless infrastructure. They use the same write path, storage layer, and data operations API.

However, every dedicated read nodes index has isolated hardware for read operations (query, fetch, list), allowing these operations to run on dedicated query executors. This affects performance, cost, and how you scale:

| Feature                 | On-demand                                                                                                                                      | Dedicated read nodes                                                                                                                                               |
| :---------------------- | :--------------------------------------------------------------------------------------------------------------------------------------------- | :----------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Read infrastructure** | Multi-tenant compute resources shared across customers                                                                                         | Isolated, provisioned query executors dedicated to your index                                                                                                      |
| **Read costs**          | Pay per [read unit](/guides/manage-cost/understanding-cost#serverless-indexes) (1 RU per 1 GB of namespace size per query, minimum 0.25 RU)    | Fixed hourly rate for read capacity based on node type, shards, and replicas                                                                                       |
| **Other costs**         | [Storage](/guides/manage-cost/understanding-cost#storage) and [write](/guides/manage-cost/understanding-cost#write-units) costs based on usage | [Storage](/guides/manage-cost/understanding-cost#storage) and [write](/guides/manage-cost/understanding-cost#write-units) costs based on usage (same as on-demand) |
| **Caching**             | Best-effort; frequently accessed data is cached, but cold queries fetch from object storage                                                    | Guaranteed; all index data always warm in memory and on local SSDs                                                                                                 |
| **Read rate limits**    | [2,000 RUs/second per index (adjustable)](/reference/api/database-limits#rate-limits)                                                          | No read rate limits (only bounded by CPU capacity)                                                                                                                 |
| **Scaling**             | Automatic; Pinecone handles capacity                                                                                                           | Manual; add [shards](#shards) for storage, add [replicas](#replicas) for throughput                                                                                |
| **Best for**            | Variable workloads, multi-tenant applications with many namespaces, low to moderate query rates                                                | Sustained high query rates, large single-namespace workloads, predictable performance and cost                                                                     |

## When to use dedicated read nodes

Dedicated read nodes are ideal for workloads with millions to billions of records and predictable query rates. They provide performance and cost benefits compared to on-demand for high-throughput workloads, and may be required when your workload exceeds on-demand rate limits.

There's no universal formula for choosing between on-demand and dedicated read nodes—performance and cost vary by workload (vector dimensionality, metadata filtering, and query patterns). Consider the following factors when making your decision:

<AccordionGroup>
  <Accordion title="Predictable, consistent performance and cost">
    With dedicated read nodes, you allocate dedicated read hardware for your index, and your data is cached in memory and on local SSDs. This provides:

    * Consistent low latency under heavy load.
    * No cold starts (fetching data from object storage).
    * Performance isolation from other workloads.
    * Linear scaling by adding replicas.
    * Predictable costs based on fixed hourly rates for provisioned hardware.

    If predictable performance and cost are critical for your application, dedicated read nodes may be a better fit than on-demand.
  </Accordion>

  <Accordion title="High throughput without rate limits or throttling">
    On-demand indexes are subject to [read unit rate limits](/reference/api/database-limits#rate-limits) (default: 2,000 RUs/second per index).

    A high query volume on a large index can exceed these limits. For example, a 15 GB namespace at 150 QPS requires approximately 2,250 RUs/second (`15 RUs per query × 150 QPS`), which exceeds the default rate limit.

    Dedicated read nodes have no read rate limits and provide dedicated capacity for predictable QPS without throttling (bounded only by CPU capacity), making them better suited for high-throughput workloads.
  </Accordion>

  <Accordion title="Recommendation engines and real-time use cases">
    Recommendation engines for use cases such as e-commerce and media require very high throughput and low latency to maintain positive user experiences. Dedicated read nodes are purpose-built for these use cases, providing:

    * Consistent performance for thousands of queries per second
    * Low latency for real-time recommendations
    * Scalability to billion-vector datasets
    * No performance degradation during traffic spikes

    Similar requirements apply to other real-time use cases like semantic search at scale, personalization engines, and mission-critical services with strict performance SLOs.
  </Accordion>

  <Accordion title="Single namespace workload (multi-namespace coming soon)">
    Dedicated read nodes indexes support only a single namespace. If your application requires multiple namespaces, on-demand is a better fit.

    <Note>
      Multi-namespace support is coming soon. For early access, [contact us](https://www.pinecone.io/contact/).
    </Note>
  </Accordion>

  <Accordion title="When NOT to use dedicated read nodes">
    On-demand indexes are better suited for workloads with unpredictable or highly variable traffic patterns. For example:

    * RAG systems with variable query volumes
    * Agentic applications with sporadic usage
    * Prototypes and development environments with intermittent activity
    * Scheduled jobs with infrequent, batch-style queries

    Additionally, on-demand is better for indexes with many namespaces, even if you have high query volumes. Dedicated read nodes currently only support single-namespace indexes, so multi-tenant applications requiring namespace-based isolation should use on-demand until multi-namespace support is available.

    For these scenarios, on-demand's elasticity and usage-based pricing provide better cost efficiency than provisioning dedicated capacity.

    <Note>
      Dedicated read nodes **can** handle predictable traffic spikes efficiently if you scale replicas proactively via the API. For example, you can provision extra replicas before a scheduled email campaign and scale back down afterward. Auto-scaling will be available in a future release.
    </Note>
  </Accordion>

  <Accordion title="Cost considerations">
    On-demand and dedicated read nodes have different cost structures. The key difference is read costs: on-demand uses usage-based pricing, while dedicated read nodes use a fixed hourly rate based on provisioned hardware. Write and storage costs are usage-based for both modes.

    Dedicated read nodes become cost-effective when you have predictable, sustained query volumes that make full use of your provisioned capacity. With unpredictable or low query volumes, you pay hourly rates even when your machines sit idle, making on-demand's usage-based pricing more economical.

    For detailed cost information, comparison tables, and estimation tools, see the [Cost](#cost) section of this guide.
  </Accordion>

  <Accordion title="Test results for your workload">
    Performance depends on your specific workload — index size, vector dimensionality, metadata filtering, query patterns, throughput requirements, and latency requirements. Testing is the only way to know for sure whether dedicated read nodes are right for your scenario.

    For a step-by-step guide to testing, see [Test your workload](#test-your-workload).
  </Accordion>
</AccordionGroup>

If you need guidance choosing a capacity mode (on-demand or dedicated read nodes) or sizing your index configuration, [contact us](https://www.pinecone.io/contact/).

## Key concepts

Before creating a dedicated read nodes index, understand the configuration options that determine capacity and performance.

### Node types

A node is the basic unit of compute and cache storage capacity for a dedicated read nodes index. Each shard runs on one node, so the node type you choose determines the performance characteristics and cost of your index. The total number of nodes in your index is calculated as `shards × replicas`. For example, an index with two shards and two replicas uses four nodes.

There are two node types: `b1` and `t1`. Both are suitable for large-scale and demanding workloads, but they differ in processing power and memory capacity, and they cache different data.

|                      | **b1 (Balanced)**                                                               | **t1 (Performance)**                                                                                            |
| -------------------- | ------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------- |
| **Memory caching**   | Vector index stored in memory                                                   | Vector index + vector projections cached in memory                                                              |
| **Use case**         | Predictable performance for sustained query rates with balanced cost efficiency | Highest performance for the most demanding workloads with extreme query volumes and strict latency requirements |
| **Storage**          | 250 GB per shard                                                                | 250 GB per shard                                                                                                |
| **Compute & memory** | Base-level compute and memory resources                                         | \~4x more compute and memory than `b1`                                                                          |
| **Cost**             | Lower-cost option                                                               | \~3x the cost of `b1`                                                                                           |

Consider using `t1` nodes if your performance requirements are not met by `b1` nodes, or if `t1` nodes are more cost-effective than `b1` nodes for your workload.

<Note>
  When choosing a node type, remember that:

  * Both types of nodes provide 250 GB of storage per shard. The difference is in compute and memory, which affects query performance.
  * Because `t1` nodes cache more data in memory than `b1` nodes, an index may require more shards on `t1` than on `b1` (for the same data).
  * You can [change node types](#change-node-types) after creating your index.
</Note>

### Shards

Shards determine the storage capacity of an index. Each shard provides 250 GB of storage, and data is split across all the shards in an index. To respond to a query, the index gathers data from all shards as needed. To determine how many shards you need, [calculate your index size](#calculate-the-size-of-your-index) and then [calculate the number of shards](#number-of-shards).

<Warning>
  It's your responsibility to allocate enough shards to accommodate the size of your index. If your index exceeds the capacity of its shards, write operations (upsert, update, delete) are blocked, but read operations continue to work normally.
</Warning>

### Replicas

Replicas multiply the compute resources and data of an index, allowing for higher query throughput and availability. Each replica is a complete copy of your index data and has its own dedicated compute resources.

* Throughput scales approximately linearly with replicas. For example, if one replica handles 50 QPS at your target latency, two replicas should handle approximately 100 QPS.
* You can scale replicas up or down with no downtime using the API. See [Add or remove replicas](#add-or-remove-replicas).
* Minimum: 0 replicas ([pauses the index](#pause-an-index)).
* For high availability, use at least two replicas. The recommended approach is to allocate `n+1` replicas where `n` is your minimum for throughput. Pinecone distributes replicas across availability zones (up to three per region), so if one zone fails, remaining replicas continue serving queries.

To determine how many replicas you need, [test your workload](#test-your-workload) and then [calculate the number of replicas](#number-of-replicas).

<Note>
  Actual performance varies based on workload characteristics (query complexity, vector dimensions, metadata characteristics), [metadata filter](/guides/search/filter-by-metadata) selectivity, and [node type](#node-types) (`b1` vs `t1`). Always test with your specific workload.
</Note>

### Index fullness

Index fullness measures how much of your index's allocated capacity is being used. To ensure predictable performance, dedicated read nodes cache your data in memory and on local SSD.

* You can use Pinecone's API to [check index fullness](#monitor-index-fullness). There are three metrics to monitor: `memoryFullness`, `storageFullness`, and `indexFullness`.
  <Note>
    `indexFullness` is the maximum of `memoryFullness` and `storageFullness`.
  </Note>
* Usually, storage fills up first. However, memory can be the limiting factor when you have `b1` nodes with many low-dimension vectors, or when you have `t1` nodes with high-dimension vectors and lots of metadata.
* Monitor fullness regularly and [add shards](#add-or-remove-shards) before your index reaches capacity. When `indexFullness` reaches 1.0 (100%), write operations (upsert, update, delete) are blocked, but read operations continue to work normally.

<Note>
  Add shards when [index fullness](#index-fullness) reaches 70-80%, especially if you expect continued growth. Adding shards reduces storage fullness (index data is spread across shards, so each stores less) and memory fullness (with less data per shard, there's less to cache in memory), helping you avoid write failures.
</Note>

## Test your workload

To choose between on-demand and dedicated read nodes, or to optimize your dedicated read nodes configuration, test with your actual workload. Performance varies based on factors such as the size of your index,vector dimensionality, metadata characteristics, and query patterns.

<Steps>
  <Step title="Calculate the size of your index">
    [Calculate the size of your index](#calculate-the-size-of-your-index) to determine how many shards it requires.
  </Step>

  <Step title="Create and populate a test index">
    [Create a dedicated read nodes index](#create-a-dedicated-read-nodes-index) with representative data for your workload. You'll use this index for testing.

    <Tip>
      If you don't restore your test index from a backup, you can [upsert](/guides/index-data/upsert-data) or [import](/guides/index-data/import-data) your data.
    </Tip>
  </Step>

  <Step title="Migrate your test index to dedicated read nodes (if necessary)">
    If your test index is on-demand, [migrate it to dedicated read nodes](#migrate-to-dedicated-read-nodes). To start, use a single `b1` replica.

    <Warning>
      Don't migrate your production index yet. At this point, you're just testing your workload.
    </Warning>
  </Step>

  <Step title="Run a load test">
    Run realistic query patterns against your test index, gradually increasing QPS. For example, start at 10 QPS for about 30 minutes, then increase by 10 QPS increments while monitoring latency. Identify the QPS where latency exceeds your target threshold.
  </Step>

  <Step title="Calculate replicas">
    Throughput scales approximately linearly with replicas. For example, if one replica handles 50 QPS at your target latency, two replicas should handle approximately 100 QPS. However, performance can vary based on metadata filter selectivity.

    To calculate the number of replicas required for your target QPS, use this formula, rounding up:

    ```
    Minimum replicas = (Required QPS) / (QPS per replica)
    ```

    For more information, see [Number of replicas](#number-of-replicas).
  </Step>

  <Step title="Adjust and re-test">
    To meet your performance and cost goals, adjust your configuration as needed and re-test:

    * [Add or remove shards](#add-or-remove-shards) for storage capacity
    * [Add or remove replicas](#add-or-remove-replicas) for throughput
    * [Change node types](#change-node-types) for different performance characteristics

    Continue iterating until you meet your requirements with room for growth.
  </Step>
</Steps>

## Calculate the size of your index

To determine how many [shards](#shards) your index requires, calculate your index size and then use the formula in the [section](#number-of-shards) below.

### Index size

A record can include a dense vector, a sparse vector, or both. Use the formula for your index type to calculate total size:

<div className="border border-gray-200 dark:border-gray-700 rounded-lg p-4 my-4">
  <Tabs>
    <Tab title="Dense index">
      A [dense index](/guides/index-data/indexing-overview#dense-indexes) contains records with one dense vector each.

      <Note>
        Dense index records can also contain sparse vectors (when the index metric is set to `dotproduct`), which can be useful for [hybrid search](/guides/search/hybrid-search#use-a-single-hybrid-index). To learn how to calculate the size of a hybrid index, see [Hybrid index](#hybrid-index).
      </Note>

      **Calculate dense index size (assuming no sparse vectors)**

      ```
      Index size = Number of records × (
                     ID size + 
                     Metadata size +
                     Dense vector dimensions × 4 bytes
                   )
      ```

      Where:

      * `ID size` and `Metadata size` are measured in bytes, averaged across all records.
      * Each `Dense vector dimension` uses 4 bytes.

      **Example dense index calculations**

      These examples assume 8-byte IDs:

      | Records    | Dense vector dimensions | Avg metadata size | Index size |
      | :--------- | :---------------------- | :---------------- | :--------- |
      | 500,000    | 768                     | 500 bytes         | 1.79 GB    |
      | 1,000,000  | 1536                    | 1,000 bytes       | 7.15 GB    |
      | 5,000,000  | 1024                    | 15,000 bytes      | 95.5 GB    |
      | 10,000,000 | 1536                    | 1,000 bytes       | 71.5 GB    |

      <Note>
        Example: 500,000 records × (8-byte ID + (768 dense vector dimensions × 4 bytes) + 500 bytes of metadata) = 1.79 GB
      </Note>
    </Tab>

    <Tab title="Sparse index">
      A [sparse index](/guides/index-data/indexing-overview#sparse-indexes) contains records with one sparse vector each.

      **Calculate sparse index size**

      ```
      Index size = Number of records × (
                     ID size + 
                     Metadata size +
                     Number of non-zero sparse values × 8 bytes
                   )
      ```

      Where:

      * `ID size` and `Metadata size` are measured in bytes, averaged across all records.
      * `Number of non-zero sparse values`: Average number across all records. To find the count for a single record, check the length of the sparse vector's `indices` or `values` array. Each non-zero value uses 8 bytes.

      **Example sparse index calculations**

      These examples assume 8-byte IDs:

      | Records    | Avg number of non-zero sparse values | Avg metadata size | Index size |
      | :--------- | :----------------------------------- | :---------------- | :--------- |
      | 500,000    | 10                                   | 500 bytes         | 0.29 GB    |
      | 1,000,000  | 50                                   | 1,000 bytes       | 1.41 GB    |
      | 5,000,000  | 100                                  | 15,000 bytes      | 79.0 GB    |
      | 10,000,000 | 50                                   | 1,000 bytes       | 14.1 GB    |

      <Note>
        Example: 500,000 records × (8-byte ID + (10 non-zero sparse values × 8 bytes) + 500 bytes of metadata) = 0.29 GB
      </Note>
    </Tab>

    <Tab title="Hybrid index">
      A [hybrid index](/guides/search/hybrid-search#use-a-single-hybrid-index) contains records that each have one dense vector and an optional sparse vector.

      **Calculate hybrid index size**

      ```
      Index size = Number of records × (
                     ID size + 
                     Metadata size +
                     Dense vector dimensions × 4 bytes + 
                     Number of non-zero sparse values × 8 bytes
                   )
      ```

      Where:

      * `ID size` and `Metadata size` are measured in bytes, averaged across all records.
      * Each `Dense vector dimension` uses 4 bytes.
      * `Number of non-zero sparse values`: Average number across all records, including those without sparse vectors. To find the count for a single record, check the length of the sparse vector's `indices` or `values` array. Each non-zero value uses 8 bytes.

      **Example hybrid index calculations**

      These examples assume 8-byte IDs:

      | Records    | Dense vector dimensions | Avg number of non-zero sparse values | Avg metadata size | Index size |
      | :--------- | :---------------------- | :----------------------------------- | :---------------- | :--------- |
      | 500,000    | 768                     | 10                                   | 500 bytes         | 1.83 GB    |
      | 1,000,000  | 1536                    | 50                                   | 1,000 bytes       | 7.54 GB    |
      | 5,000,000  | 1024                    | 100                                  | 15,000 bytes      | 99.5 GB    |
      | 10,000,000 | 1536                    | 50                                   | 1,000 bytes       | 75.4 GB    |

      <Note>
        Example: 500,000 records × (8-byte ID + (768 dense vector dimensions × 4 bytes) + (10 non-zero sparse values × 8 bytes) + 500 bytes of metadata) = 1.83 GB
      </Note>
    </Tab>
  </Tabs>
</div>

### Number of shards

To calculate the number of shards your index requires, divide the size of your index by 250 GB and round up:

```
Minimum shards = (Index size) / (250 GB per shard)
```

To maintain optimal performance, provision additional shards to keep your index at 70-80% capacity. For example, a 500 GB index should have three shards (750 GB capacity = 67% full), not two shards (500 GB capacity = 100% full).

**Example shard calculations**

| Index size   | Minimum shards       | Recommended shards   |
| :----------- | :------------------- | :------------------- |
| **\~71 GB**  | 1 (250 GB; 28% full) | 1 (250 GB; 28% full) |
| **\~300 GB** | 2 (500 GB; 60% full) | 2 (500 GB; 60% full) |
| **\~400 GB** | 2 (500 GB; 80% full) | 3 (750 GB; 53% full) |

**Other considerations**

* Every index must have at least one shard. However, you can [pause an index](#pause-an-index) by reducing its replicas to 0.
* After you've created your index, [monitor its fullness](#monitor-index-fullness). When your index approaches capacity, you can [add shards](#add-or-remove-shards).

<Note>
  Add shards when [index fullness](#index-fullness) reaches 70-80%, especially if you expect continued growth. Adding shards reduces storage fullness (index data is spread across shards, so each stores less) and memory fullness (with less data per shard, there's less to cache in memory), helping you avoid write failures.
</Note>

### Number of replicas

To calculate the number of replicas your index requires, first [test your workload](#test-your-workload) to find the QPS a single replica can handle at your target latency. Then, use this formula, and round up:

```
Minimum replicas = (Required QPS) / (QPS per replica)
```

For example, if one replica handles 50 QPS at your target latency and you need 150 QPS, you need three replicas.

**Other considerations**

* Throughput scales approximately linearly with replicas, but performance can vary based on metadata filter selectivity.
* For high availability, allocate `n+1` replicas where `n` is your minimum for throughput. Pinecone distributes replicas across availability zones.

## Create a dedicated read nodes index

You can create a dedicated read nodes index from scratch or from a backup of an existing index.

### From scratch

To create a new dedicated read nodes index from scratch, call [Create an index](/reference/api/2025-10/control-plane/create_index). In the request body, in the `spec.serverless.read_capacity` object, set the following fields:

| Field                           | Value                                            | Notes                                                  |
| :------------------------------ | :----------------------------------------------- | :----------------------------------------------------- |
| **`mode`**                      | `Dedicated`                                      |                                                        |
| **`dedicated.node_type`**       | `b1` or `t1`                                     | See [node types](#node-types)                          |
| **`dedicated.scaling`**         | `Manual`                                         | Currently the only option                              |
| **`dedicated.manual.shards`**   | Number of [shards](#number-of-shards) needed     | Minimum 1 shard; each shard provides 250 GB of storage |
| **`dedicated.manual.replicas`** | Number of [replicas](#number-of-replicas) needed | Minimum 0 (this [pauses](#pause-an-index) the index)   |

<Note>
  To learn how to determine the number of shards and replicas your index requires, see [Calculate the size of your index](#calculate-the-size-of-your-index).
</Note>

**Example**

<CodeGroup>
  ```bash curl expandable theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -X POST "https://api.pinecone.io/indexes" \
       -H "Accept: application/json" \
       -H "Content-Type: application/json" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H "X-Pinecone-Api-Version: 2025-10" \
       -d '{
             "name": "example-dedicated-index",
             "dimension": 1024,
             "metric": "cosine",
             "deletion_protection": "enabled",
             "tags": {
               "environment": "production"
             },
             "vector_type": "dense",
             "spec": {
               "serverless": {
                 "cloud": "aws",
                 "region": "us-east-1",
                 "read_capacity": {
                   "mode": "Dedicated",
                   "dedicated": {
                     "node_type": "b1",
                     "scaling": "Manual",
                     "manual": {
                       "shards": 2,
                       "replicas": 1
                     }
                   }
                 }
               }
             }
           }'
  ```
</CodeGroup>

Example response:

<CodeGroup>
  ```json curl highlight={21,27} expandable theme={null}
  {
    "name": "example-dedicated-index",
    "vector_type": "dense",
    "metric": "cosine",
    "dimension": 1024,
    "status": {
      "ready": false,
      "state": "Initializing"
    },
    "host": "example-dedicated-index-1c6ab6aa.svc.aped-4627-b74a.pinecone.io",
    "spec": {
      "serverless": {
        "region": "us-east-1",
        "cloud": "aws",
        "read_capacity": {
          "mode": "Dedicated",
          "dedicated": {
            "node_type": "b1",
            "scaling": "Manual",
            "manual": {
              "shards": 2, // <---- desired state
              "replicas": 1
            }
          },
          "status": {
            "state": "Migrating",
            "current_shards": null, // <---- current state
            "current_replicas": null
          }
        }
      }
    },
    "deletion_protection": "enabled",
    "tags": {
      "environment": "production"
    }
  }
  ```
</CodeGroup>

The response includes two status fields:

| Field                                            | Description                                                                |
| :----------------------------------------------- | :------------------------------------------------------------------------- |
| **`status.state`**                               | Overall index status (for example, `Initializing`, `Ready`, `Terminating`) |
| **`spec.serverless.read_capacity.status.state`** | Read capacity status (`Migrating`, `Scaling`, `Ready`, `Error`)            |

<Note>
  When creating a dedicated read nodes index, `status.state` transitions to `Ready` as soon as the index is ready for reads and writes.

  However, `spec.serverless.read_capacity.status.state` remains `Migrating` until the index scales to its full read capacity, at which point it transitions to `Ready`.
</Note>

After creating the index, [upsert](/guides/index-data/upsert-data) or [import](/guides/index-data/import-data) your data.

<Tip>
  To upsert and search with text instead of vectors, you can configure your index to use a [hosted embedding model](/guides/index-data/create-an-index#embedding-models). Call [Configure an index](/reference/api/2025-10/control-plane/configure_index) and specify the `embed` object in the request body.
</Tip>

### From a backup

To create a dedicated read nodes index from a backup:

1. [Restore the backup](/guides/manage-data/restore-an-index). This creates a new on-demand index with the same data as the original.
2. If the restored index has multiple namespaces, [delete](/reference/api/latest/data-plane/deletenamespace) all of them except the one you want to keep. Dedicated read nodes currently only support [one namespace](#namespace-limits).
3. [Migrate the index to dedicated read nodes](#migrate-to-dedicated-read-nodes).

## Migrate to dedicated read nodes

To migrate an existing on-demand index to dedicated read nodes, follow these steps:

<Steps>
  <Step title="Create a backup of your index">
    [Create a backup](/guides/manage-data/back-up-an-index) of your index. If you later find that on-demand is preferable, you can restore the backup to a new on-demand index or [contact support](https://app.pinecone.io/organizations/-/settings/support/ticket) to migrate back.
  </Step>

  <Step title="Delete extra namespaces">
    If your index has multiple namespaces, [delete](/reference/api/latest/data-plane/deletenamespace) all of them except the one you want to keep. Dedicated read nodes currently only support a [single namespace](#namespace-limits).

    <Warning>
      If this is a production index, be sure to make a [backup](/guides/manage-data/back-up-an-index) before deleting namespaces. Or, if you need multiple namespaces, [contact support](https://app.pinecone.io/organizations/-/settings/support/ticket) to discuss early access to multi-namespace support for dedicated read nodes.
    </Warning>
  </Step>

  <Step title="Calculate your index size">
    [Calculate your index size](#index-size) to determine how many [shards](#number-of-shards) you need.
  </Step>

  <Step title="Migrate the index">
    To migrate the index, call [Configure an index](/reference/api/2025-10/control-plane/configure_index). In the request body, in the `spec.serverless.read_capacity` object, set the following fields:

    | Field                           | Value                                            | Notes                                                  |
    | :------------------------------ | :----------------------------------------------- | :----------------------------------------------------- |
    | **`mode`**                      | `Dedicated`                                      |                                                        |
    | **`dedicated.node_type`**       | `b1` or `t1`                                     | See [node types](#node-types)                          |
    | **`dedicated.scaling`**         | `Manual`                                         | Currently the only option                              |
    | **`dedicated.manual.shards`**   | Number of [shards](#number-of-shards) needed     | Minimum 1 shard; each shard provides 250 GB of storage |
    | **`dedicated.manual.replicas`** | Number of [replicas](#number-of-replicas) needed | Minimum 0 (this [pauses](#pause-an-index) the index)   |

    **Example**

    This example migrates an index to dedicated read nodes using `b1` nodes, one shard, and one replica:

    <CodeGroup>
      ```bash curl expandable theme={null}
      PINECONE_API_KEY="YOUR_API_KEY"
      INDEX_NAME="YOUR_INDEX_NAME"

      curl -X PATCH "https://api.pinecone.io/indexes/$INDEX_NAME" \
           -H "Accept: application/json" \
           -H "Content-Type: application/json" \
           -H "Api-Key: $PINECONE_API_KEY" \
           -H "X-Pinecone-Api-Version: 2025-10" \
           -d '{
                 "spec": {
                   "serverless": {
                     "read_capacity": {
                       "mode": "Dedicated",
                       "dedicated": {
                         "node_type": "b1",
                         "scaling": "Manual",
                         "manual": {
                           "shards": 1,
                           "replicas": 1
                         }
                       }
                     }
                   }
                 }
               }'
      ```
    </CodeGroup>

    Example response:

    <CodeGroup>
      ```json curl highlight={21,27} expandable theme={null}
      {
        "name": "example-index-to-migrate",
        "vector_type": "dense",
        "metric": "cosine",
        "dimension": 1024,
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "host": "example-index-to-migrate-1c6ab6aa.svc.aped-4627-b74a.pinecone.io",
        "spec": {
          "serverless": {
            "region": "us-east-1",
            "cloud": "aws",
            "read_capacity": {
              "mode": "Dedicated",
              "dedicated": {
                "node_type": "b1",
                "scaling": "Manual",
                "manual": {
                  "shards": 1, // <---- desired state
                  "replicas": 1
                }
              },
              "status": {
                "state": "Migrating",
                "current_shards": null, //<---- current state
                "current_replicas": null
              }
            }
          }
        },
        "deletion_protection": "disabled",
        "tags": null,
        "embed": {
          "model": "llama-text-embed-v2",
          "field_map": {
            "text": "text"
          },
          "dimension": 1024,
          "metric": "cosine",
          "write_parameters": {
            "dimension": 1024,
            "input_type": "passage",
            "truncate": "END"
          },
          "read_parameters": {
            "dimension": 1024,
            "input_type": "query",
            "truncate": "END"
          },
          "vector_type": "dense"
        }
      }
      ```
    </CodeGroup>

    The response includes two status fields:

    | Field                                            | Description                                                                |
    | :----------------------------------------------- | :------------------------------------------------------------------------- |
    | **`status.state`**                               | Overall index status (for example, `Initializing`, `Ready`, `Terminating`) |
    | **`spec.serverless.read_capacity.status.state`** | Read capacity status (`Migrating`, `Scaling`, `Ready`, `Error`)            |

    <Warning>
      If `status.state` is set to `Error`, the allocated number of shards was insufficient for the size of the index. Try again, adding more shards as needed.
    </Warning>
  </Step>

  <Step title="Monitor the migration">
    [Monitor](#check-the-status-of-a-change) the status of the migration. When the migration is complete, `spec.serverless.read_capacity.status.state` is `Ready`.
  </Step>

  <Step title="Monitor index performance">
    After migrating, monitor your index performance to verify that it meets expectations.
  </Step>
</Steps>

## Manage your index

The following sections describe how to manage a dedicated read nodes index using version `2025-10` of the Pinecone API.

<AccordionGroup>
  <Accordion title="Add a hosted embedding model">
    To upsert and search with text instead of vectors, you can configure your index to use a [hosted embedding model](/guides/index-data/create-an-index#embedding-models). To do this, call [Configure an index](/reference/api/2025-10/control-plane/configure_index) and provide an `embed` object in the request body. In this object:

    * For the `text` field, specify the name of the field in your data that contains the text to be embedded.
    * Specify a model whose dimension requirements match the dimensions of your index.

    **Example**

    Example request:

    <CodeGroup>
      ```bash curl expandable theme={null}
      PINECONE_API_KEY="YOUR_API_KEY"
      INDEX_NAME="YOUR_INDEX_NAME"

      curl -X PATCH "https://api.pinecone.io/indexes/$INDEX_NAME" \
           -H "Content-Type: application/json" \
           -H "Api-Key: $PINECONE_API_KEY" \
           -H "X-Pinecone-Api-Version: 2025-10" \
           -d '{
                 "embed": {
                   "field_map": {
                     "text": "chunk_text"
                   },
                   "model": "llama-text-embed-v2",
                   "read_parameters": {
                     "input_type": "query",
                     "truncate": "NONE"
                   },
                   "write_parameters": {
                     "input_type": "passage"
                   }
                 }
               }'
      ```
    </CodeGroup>

    Example response:

    <CodeGroup>
      ```json curl expandable theme={null}
      {
        "name": "example-dedicated-index",
        "vector_type": "dense",
        "metric": "cosine",
        "dimension": 1024,
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "host": "example-dedicated-index-1c6ab6aa.svc.aped-4627-b74a.pinecone.io",
        "spec": {
          "serverless": {
            "region": "us-east-1",
            "cloud": "aws",
            "read_capacity": {
              "mode": "Dedicated",
              "dedicated": {
                "node_type": "b1",
                "scaling": "Manual",
                "manual": {
                  "shards": 2,
                  "replicas": 1
                }
              },
              "status": {
                "state": "Ready",
                "current_shards": 2,
                "current_replicas": 1
              }
            }
          }
        },
        "deletion_protection": "enabled",
        "tags": {
          "environment": "testing"
        },
        "embed": {
          "model": "llama-text-embed-v2",
          "field_map": {
            "text": "chunk_text"
          },
          "dimension": 1024,
          "metric": "cosine",
          "write_parameters": {
            "dimension": 1024,
            "input_type": "passage",
            "truncate": "END"
          },
          "read_parameters": {
            "dimension": 1024,
            "input_type": "query",
            "truncate": "NONE"
          },
          "vector_type": "dense"
        }
      }
      ```
    </CodeGroup>

    <Note>
      You can also create a dedicated read nodes index when calling [Create an index with integrated embedding](/reference/api/2025-10/control-plane/create_for_model). In the request body, use the `read_capacity` object to configure node type, shards, and replicas for dedicated read nodes.
    </Note>
  </Accordion>

  <Accordion title="Monitor index fullness">
    To check [index fullness](#index-fullness), call [Get index stats](/reference/api/2025-10/data-plane/describeindexstats).

    **Example**

    Example request:

    <CodeGroup>
      ```bash curl theme={null}
      # To get the unique host for an index,
      # see https://docs.pinecone.io/guides/manage-data/target-an-index
      PINECONE_API_KEY="YOUR_API_KEY"
      INDEX_HOST="YOUR_INDEX_HOST"

      curl -X GET "https://$INDEX_HOST/describe_index_stats" \
           -H "Api-Key: $PINECONE_API_KEY" \
           -H "X-Pinecone-Api-Version: 2025-10"
      ```
    </CodeGroup>

    Example response:

    <CodeGroup>
      ```json curl expandable theme={null}
      {
        "namespaces": {
          "__default__": {
            "vectorCount": 705000
          }
        },
        "indexFullness": 0.01,
        "totalVectorCount": 705000,
        "dimension": 1536,
        "metric": "cosine",
        "vectorType": "dense",
        "memoryFullness": 0.01,
        "storageFullness": 0.01
      }
      ```
    </CodeGroup>

    In the response, `indexFullness` describes how full the index is, on a scale of 0 to 1. It's set to the greater of `memoryFullness` and `storageFullness`.
  </Accordion>

  <Accordion title="Add or remove shards">
    To add or remove [shards](#shards), call [Configure an index](/reference/api/2025-10/control-plane/configure_index). This operation does not require downtime, but can take up to 30 minutes to complete. In the request body, set the following fields:

    | Field                                                       | Value                               | Notes                                 |
    | :---------------------------------------------------------- | :---------------------------------- | :------------------------------------ |
    | **`spec.serverless.read_capacity.mode`**                    | `Dedicated`                         |                                       |
    | **`spec.serverless.read_capacity.dedicated.scaling`**       | `Manual`                            |                                       |
    | **`spec.serverless.read_capacity.dedicated.manual.shards`** | Desired number of [shards](#shards) | Each shard provides 250 GB of storage |

    **Example**

    Example request:

    <CodeGroup>
      ```bash curl expandable theme={null}
      PINECONE_API_KEY="YOUR_API_KEY"
      INDEX_NAME="YOUR_INDEX_NAME"

      curl -X PATCH "https://api.pinecone.io/indexes/$INDEX_NAME" \
           -H "Accept: application/json" \
           -H "Content-Type: application/json" \
           -H "Api-Key: $PINECONE_API_KEY" \
           -H "X-Pinecone-Api-Version: 2025-10" \
           -d '{
                 "spec": {
                   "serverless": {
                     "read_capacity": {
                       "mode": "Dedicated",
                       "dedicated": {
                         "scaling": "Manual",
                         "manual": {
                           "shards": 3
                         }
                       }
                     }
                   }
                 }
               }'
      ```
    </CodeGroup>

    Example response:

    <CodeGroup>
      ```json curl highlight={21,27} expandable theme={null}
      {
        "name": "example-dedicated-index",
        "vector_type": "dense",
        "metric": "cosine",
        "dimension": 1024,
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "host": "example-dedicated-index-1c6ab6aa.svc.aped-4627-b74a.pinecone.io",
        "spec": {
          "serverless": {
            "region": "us-east-1",
            "cloud": "aws",
            "read_capacity": {
              "mode": "Dedicated",
              "dedicated": {
                "node_type": "b1",
                "scaling": "Manual",
                "manual": {
                  "shards": 3, // <---- desired state
                  "replicas": 1
                }
              },
              "status": {
                "state": "Scaling",
                "current_shards": 2, // <---- current state
                "current_replicas": 1
              }
            }
          }
        },
        "deletion_protection": "disabled",
        "tags": null,
        "embed": {
          "model": "llama-text-embed-v2",
          "field_map": {
            "text": "text"
          },
          "dimension": 1024,
          "metric": "cosine",
          "write_parameters": {
            "dimension": 1024,
            "input_type": "passage",
            "truncate": "END"
          },
          "read_parameters": {
            "dimension": 1024,
            "input_type": "query",
            "truncate": "END"
          },
          "vector_type": "dense"
        }
      }
      ```
    </CodeGroup>

    <Note>
      Configuration change limits:

      * You can make one configuration change every ten minutes, but you can batch multiple changes (node type, shards, and replicas) in a single request.
      * A new configuration change can only be initiated after the previous configuration change has completed.
      * Each configuration change can take up to 30 minutes to complete.
      * Read and write operations continue normally during configuration changes.
    </Note>
  </Accordion>

  <Accordion title="Add or remove replicas">
    To add or remove [replicas](#replicas), call [Configure an index](/reference/api/2025-10/control-plane/configure_index). This operation does not require downtime, but can take up to 30 minutes to complete. In the request body, set the following fields:

    | Field                                                         | Value                                   | Notes                                     |
    | :------------------------------------------------------------ | :-------------------------------------- | :---------------------------------------- |
    | **`spec.serverless.read_capacity.mode`**                      | `Dedicated`                             |                                           |
    | **`spec.serverless.read_capacity.dedicated.scaling`**         | `Manual`                                |                                           |
    | **`spec.serverless.read_capacity.dedicated.manual.replicas`** | Desired number of [replicas](#replicas) | Add replicas to increase query throughput |

    **Example**

    Example request:

    <CodeGroup>
      ```bash curl expandable theme={null}
      PINECONE_API_KEY="YOUR_API_KEY"
      INDEX_NAME="YOUR_INDEX_NAME"

      curl -X PATCH "https://api.pinecone.io/indexes/$INDEX_NAME" \
           -H "Accept: application/json" \
           -H "Content-Type: application/json" \
           -H "Api-Key: $PINECONE_API_KEY" \
           -H "X-Pinecone-Api-Version: 2025-10" \
           -d '{
                 "spec": {
                   "serverless": {
                     "read_capacity": {
                       "mode": "Dedicated",
                       "dedicated": {
                         "scaling": "Manual",
                         "manual": {
                           "replicas": 2
                         }
                       }
                     }
                   }
                 }
               }'
      ```
    </CodeGroup>

    Example response:

    <CodeGroup>
      ```json curl highlight={22,28} expandable theme={null}
      {
        "name": "example-dedicated-index",
        "vector_type": "dense",
        "metric": "cosine",
        "dimension": 1024,
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "host": "example-dedicated-index-1c6ab6aa.svc.aped-4627-b74a.pinecone.io",
        "spec": {
          "serverless": {
            "region": "us-east-1",
            "cloud": "aws",
            "read_capacity": {
              "mode": "Dedicated",
              "dedicated": {
                "node_type": "b1",
                "scaling": "Manual",
                "manual": {
                  "shards": 1,
                  "replicas": 2 // <---- desired state
                }
              },
              "status": {
                "state": "Scaling",
                "current_shards": 1,
                "current_replicas": 1 // <---- current state
              }
            }
          }
        },
        "deletion_protection": "disabled",
        "tags": null,
        "embed": {
          "model": "llama-text-embed-v2",
          "field_map": {
            "text": "text"
          },
          "dimension": 1024,
          "metric": "cosine",
          "write_parameters": {
            "dimension": 1024,
            "input_type": "passage",
            "truncate": "END"
          },
          "read_parameters": {
            "dimension": 1024,
            "input_type": "query",
            "truncate": "END"
          },
          "vector_type": "dense"
        }
      }
      ```
    </CodeGroup>

    <Note>
      Configuration change limits:

      * You can make one configuration change every ten minutes, but you can batch multiple changes (node type, shards, and replicas) in a single request.
      * A new configuration change can only be initiated after the previous configuration change has completed.
      * Each configuration change can take up to 30 minutes to complete.
      * Read and write operations continue normally during configuration changes.
    </Note>
  </Accordion>

  <Accordion title="Change node types">
    You can change node types in either direction (`b1` → `t1` or `t1` → `b1`). This operation does not require downtime, but can take up to 30 minutes to complete.

    <Note>
      The most predictable way to increase throughput is by increasing [replicas](#replicas).
    </Note>

    <Warning>
      `t1` nodes [cache more data in memory](#node-types) than `b1` nodes. Because of this, switching from `b1` to `t1` may require more shards.

      If your new configuration doesn't have enough shards, the configuration change will fail with an error telling you how many shards are required. Update the request and retry.

      In the meantime, your index will continue to function normally in its original configuration.
    </Warning>

    To change node types, call [Configure an index](/reference/api/2025-10/control-plane/configure_index). In the request body, set the following fields:

    | Field                                                   | Value        | Notes                         |
    | :------------------------------------------------------ | :----------- | :---------------------------- |
    | **`spec.serverless.read_capacity.mode`**                | `Dedicated`  |                               |
    | **`spec.serverless.read_capacity.dedicated.node_type`** | `b1` or `t1` | See [node types](#node-types) |

    **Example**

    Example request to change from `b1` to `t1`:

    <CodeGroup>
      ```bash curl expandable theme={null}
      PINECONE_API_KEY="YOUR_API_KEY"
      INDEX_NAME="YOUR_INDEX_NAME"

      curl -X PATCH "https://api.pinecone.io/indexes/$INDEX_NAME" \
           -H "Accept: application/json" \
           -H "Content-Type: application/json" \
           -H "Api-Key: $PINECONE_API_KEY" \
           -H "X-Pinecone-Api-Version: 2025-10" \
           -d '{
                 "spec": {
                   "serverless": {
                     "read_capacity": {
                       "mode": "Dedicated",
                       "dedicated": {
                         "node_type": "t1"
                       }
                     }
                   }
                 }
               }'
      ```
    </CodeGroup>

    Example response:

    <CodeGroup>
      ```json curl expandable theme={null}
      {
        "name": "example-dedicated-index",
        "vector_type": "dense",
        "metric": "cosine",
        "dimension": 1024,
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "host": "example-dedicated-index-1c6ab6aa.svc.aped-4627-b74a.pinecone.io",
        "spec": {
          "serverless": {
            "region": "us-east-1",
            "cloud": "aws",
            "read_capacity": {
              "mode": "Dedicated",
              "dedicated": {
                "node_type": "t1",
                "scaling": "Manual",
                "manual": {
                  "shards": 1,
                  "replicas": 1
                }
              },
              "status": {
                "state": "Scaling",
                "current_shards": 1,
                "current_replicas": 1
              }
            }
          }
        },
        "deletion_protection": "disabled",
        "tags": null,
        "embed": {
          "model": "llama-text-embed-v2",
          "field_map": {
            "text": "text"
          },
          "dimension": 1024,
          "metric": "cosine",
          "write_parameters": {
            "dimension": 1024,
            "input_type": "passage",
            "truncate": "END"
          },
          "read_parameters": {
            "dimension": 1024,
            "input_type": "query",
            "truncate": "END"
          },
          "vector_type": "dense"
        }
      }
      ```
    </CodeGroup>

    <Note>
      Configuration change limits:

      * You can make one configuration change every ten minutes, but you can batch multiple changes (node type, shards, and replicas) in a single request.
      * A new configuration change can only be initiated after the previous configuration change has completed.
      * Each configuration change can take up to 30 minutes to complete.
      * Read and write operations continue normally during configuration changes.
    </Note>
  </Accordion>

  <Accordion title="Pause an index">
    To pause an index, [set the number of replicas](#add-or-remove-replicas) to 0. This operation can take up to 30 minutes to complete.

    <Note>
      While an index is paused, you cannot write to it or read from it. For a paused index, you're billed for storage, but not for node costs, reads, or writes.
    </Note>
  </Accordion>

  <Accordion title="Check the status of a configuration change">
    After making a configuration change to a dedicated read nodes index (changing shards, replicas, or node type), check the status of the change by calling [Describe an index](/reference/api/2025-10/control-plane/describe_index).

    **Example**

    Example request:

    <CodeGroup>
      ```bash curl theme={null}
      PINECONE_API_KEY="YOUR_API_KEY"
      INDEX_NAME="YOUR_INDEX_NAME"

      curl -X GET "https://api.pinecone.io/indexes/$INDEX_NAME" \
           -H "Api-Key: $PINECONE_API_KEY" \
           -H "X-Pinecone-Api-Version: 2025-10" 
      ```
    </CodeGroup>

    Example response (index scaling from one to two replicas):

    <CodeGroup>
      ```json curl highlight={22,28} expandable theme={null}
      {
        "name": "example-dedicated-index",
        "vector_type": "dense",
        "metric": "cosine",
        "dimension": 1536,
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "host": "example-dedicated-index-1c6ab6aa.svc.aped-4627-b74a.pinecone.io",
        "spec": {
          "serverless": {
            "region": "us-east-1",
            "cloud": "aws",
            "read_capacity": {
              "mode": "Dedicated",
              "dedicated": {
                "node_type": "b1",
                "scaling": "Manual",
                "manual": {
                  "shards": 1,
                  "replicas": 2 // <---- desired state
                }
              },
              "status": {
                "state": "Scaling", 
                "current_shards": 1,
                "current_replicas": 1 // <---- current state
              }
            }
          }
        },
        "deletion_protection": "enabled",
        "tags": {
          "tag0": "value0"
        }
      }
      ```
    </CodeGroup>

    The response includes two status fields:

    | Field                                            | Description                                                                |
    | :----------------------------------------------- | :------------------------------------------------------------------------- |
    | **`status.state`**                               | Overall index status (for example, `Initializing`, `Ready`, `Terminating`) |
    | **`spec.serverless.read_capacity.status.state`** | Read capacity status (`Migrating`, `Scaling`, `Ready`, `Error`)            |

    When changing node types, shards, or replicas, monitor the read capacity status (`spec.serverless.read_capacity.status.state`). Possible values:

    | State           | Description                                                                                                                                                                                   |
    | :-------------- | :-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
    | **`Ready`**     | The change is complete and the index is ready to serve queries at full capacity.                                                                                                              |
    | **`Scaling`**   | A change to the number of shards or replicas is in progress.                                                                                                                                  |
    | **`Migrating`** | A change to the node type or read capacity <Tooltip tip="Dedicated or OnDemand">mode</Tooltip> is in progress.                                                                                |
    | **`Error`**     | The operation failed. For migrations to dedicated, this typically means you didn't allocate enough shards for your index size. Check `error_message` for details, and retry with more shards. |

    <Note>
      During changes to shards, replicas, and node type, the index-level status (`status.state`) remains `Ready`. This is because the index can handle reads and writes while its dedicated read capacity scales.
    </Note>

    <Note>
      Configuration change limits:

      * You can make one configuration change every ten minutes, but you can batch multiple changes (node type, shards, and replicas) in a single request.
      * A new configuration change can only be initiated after the previous configuration change has completed.
      * Each configuration change can take up to 30 minutes to complete.
      * Read and write operations continue normally during configuration changes.
    </Note>
  </Accordion>

  <Accordion title="Migrate from dedicated to on-demand">
    To change a dedicated read nodes index to on-demand, [contact support](https://app.pinecone.io/organizations/-/settings/support/ticket). This can't be done with the API.
  </Accordion>
</AccordionGroup>

## Limits

The following limits apply to dedicated read nodes:

<AccordionGroup>
  <Accordion title="Read limits">
    Dedicated read nodes indexes are not subject to [read-operation rate limits](/reference/api/database-limits#rate-limits), like on-demand indexes are. However, if your query rate exceeds the compute capacity of your index, you may observe decreased query throughput. In such cases, consider [adding replicas](#add-or-remove-replicas) to increase the compute resources of the index.
  </Accordion>

  <Accordion title="Write limits">
    On dedicated read nodes indexes, write operations (upsert, update, delete) have the same [rate limits](/reference/api/database-limits#rate-limits) as on-demand indexes.

    Writes that would cause your index to exceed its storage capacity are blocked. In such cases, consider [adding shards](#add-or-remove-shards) to increase available storage. To determine how close to the write limit you are, [check index fullness](#monitor-index-fullness).
  </Accordion>

  <Accordion title="Namespace limits">
    Currently, dedicated read nodes indexes only support a single namespace. However, multi-namespace support is coming soon. For early access, [contact support](https://app.pinecone.io/organizations/-/settings/support/ticket).
  </Accordion>

  <Accordion title="Shard, replica, and node limits">
    **Shards**

    The minimum number of [shards](#shards) per index is 1.

    **Replicas**

    The minimum number of [replicas](#replicas) per index is 0, which [pauses the index](#pause-an-index).

    **Nodes**

    The maximum number of [nodes](#node-types) per project is 20. This is a **project** limit, not an index limit.

    To calculate your total node count, multiply `shards × replicas` for each of your project's indexes, and then sum the results. This total must not exceed 20. For example, if you have two indexes that each have two shards and three replicas, your total node count is `(2 × 3) + (2 × 3) = 12` nodes.

    To increase your project's node limit, [contact support](https://app.pinecone.io/organizations/-/settings/support/ticket).

    <Note>
      Configuration change limits:

      * You can make one configuration change every ten minutes, but you can batch multiple changes (node type, shards, and replicas) in a single request.
      * A new configuration change can only be initiated after the previous configuration change has completed.
      * Each configuration change can take up to 30 minutes to complete.
      * Read and write operations continue normally during configuration changes.
    </Note>
  </Accordion>

  <Accordion title="memoryFullness">
    `memoryFullness` is an approximation and doesn't yet account for metadata. For more information, see [Index fullness](#index-fullness).
  </Accordion>

  <Accordion title="Migrating from dedicated to on-demand">
    To migrate an index from dedicated to on-demand, [contact support](https://app.pinecone.io/organizations/-/settings/support/ticket). This cannot be done with the API.
  </Accordion>
</AccordionGroup>

## Cost

<Note>
  For the latest pricing information, see the [Pinecone pricing page](https://www.pinecone.io/pricing/).
</Note>

The cost of an index has three components: read costs, write costs, and storage costs.

On-demand and dedicated read nodes share infrastructure for writes and storage, so these costs are the same. However, dedicated read nodes provision dedicated hardware for read operations (query, fetch, list), which changes how read costs are calculated.

| Cost component    | On-demand                                                                                                | Dedicated read nodes                                                                               |
| :---------------- | :------------------------------------------------------------------------------------------------------- | :------------------------------------------------------------------------------------------------- |
| **Read costs**    | [Usage-based](/guides/manage-cost/understanding-cost#read-units): 1 RU per 1 GB namespace size per query | Fixed hourly rate: Based on [node type](#node-types), [shards](#shards), and [replicas](#replicas) |
| **Write costs**   | [Usage-based](/guides/manage-cost/understanding-cost#write-units)                                        | [Usage-based](/guides/manage-cost/understanding-cost#write-units) (same as on-demand)              |
| **Storage costs** | [Usage-based](/guides/manage-cost/understanding-cost#storage)                                            | [Usage-based](/guides/manage-cost/understanding-cost#storage) (same as on-demand)                  |

<Note>
  If you use a hosted model for search or reranking, there are additional [inference costs](https://www.pinecone.io/pricing).
</Note>

<AccordionGroup>
  <Accordion title="Calculate dedicated read nodes costs">
    To calculate the total cost of a dedicated read nodes index, use this formula:

    ```
    (Node rate × shards × replicas) + storage costs + write costs
    ```

    | Term              | Description                                                                                                                                           |
    | :---------------- | :---------------------------------------------------------------------------------------------------------------------------------------------------- |
    | **Node rate**     | Monthly rate for the [node type](#node-types) (`b1` or `t1`), which varies by cloud region. See [Pinecone pricing](https://www.pinecone.io/pricing/). |
    | **Shards**        | Number of [shards](#shards) allocated                                                                                                                 |
    | **Replicas**      | Number of [replicas](#replicas) allocated                                                                                                             |
    | **Storage costs** | [Usage-based](/guides/manage-cost/understanding-cost#storage), same as on-demand                                                                      |
    | **Write costs**   | [Usage-based](/guides/manage-cost/understanding-cost#write-units), same as on-demand                                                                  |

    <Tip>
      For help estimating costs, use the [Pinecone pricing calculator](https://www.pinecone.io/pricing/estimate/) or [contact us](https://www.pinecone.io/contact/).
    </Tip>

    **Example:** If the rate for `b1` nodes on `aws-us-east-1` is \$336.42/month (\$0.46/hour), an index with two shards and two replicas would cost:

    ```
    336.42 × 2 × 2 = $1,345.68/month, plus storage and write costs
    ```
  </Accordion>
</AccordionGroup>

---

# Source: https://docs.pinecone.io/guides/manage-data/delete-data.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Delete records

> Delete records by ID or metadata filter from indexes

This page shows you how to [delete](/reference/api/latest/data-plane/delete) records from an index [namespace](/guides/index-data/indexing-overview#namespaces).

## Delete records by ID

Since Pinecone records can always be efficiently accessed using their ID, deleting by ID is the most efficient way to remove specific records from a namespace.

<Note>
  To remove records from the default namespace, specify `"__default__"` as the namespace in your request.
</Note>

<CodeGroup>
  ```Python Python theme={null}
  # pip install "pinecone[grpc]"
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  index.delete(ids=["id-1", "id-2"], namespace='example-namespace')
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  const ns = index.namespace('example-namespace')
  // Delete one record by ID.
  await ns.deleteOne('id-1');
  // Delete more than one record by ID.
  await ns.deleteMany(['id-2', 'id-3']);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;

  import java.util.Arrays;
  import java.util.List;

  public class DeleteExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");
          List<String> ids = Arrays.asList("id-1", "id-2");
          index.deleteByIds(ids, "example-namespace");
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      id1 := "id-1"
      id2 := "id-2"

      err = idxConnection.DeleteVectorsById(ctx, []string{id1, id2})
      if err != nil {
          log.Fatalf("Failed to delete vector with ID %v: %v", id, err)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var deleteResponse = await index.DeleteAsync(new DeleteRequest {
      Ids = new List<string> { "id-1", "id-2" },
      Namespace = "example-namespace",
  });
  ```

  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/vectors/delete" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "ids": [
        "id-1", 
        "id-2"
      ],
      "namespace": "example-namespace"
    }
  '
  ```
</CodeGroup>

## Delete records by metadata

To delete records from a namespace based on their metadata values, pass a [metadata filter expression](/guides/index-data/indexing-overview#metadata-filter-expressions) to the `delete` operation. This deletes all records in the namespace that match the filter expression.

For example, the following code deletes all records with a `genre` field set to `documentary` from namespace `example-namespace`:

<CodeGroup>
  ```Python Python theme={null}
  # pip install "pinecone[grpc]"
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  index.delete(
      filter={
          "genre": {"$eq": "documentary"}
      },
      namespace="example-namespace" 
  )
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  const ns = index.namespace('example-namespace')

  await ns.deleteMany({
    genre: { $eq: "documentary" },
  });
  ```

  ```java Java theme={null}
  import com.google.protobuf.Struct;
  import com.google.protobuf.Value;
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;

  import java.util.Arrays;
  import java.util.List;

  public class DeleteExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");
          Struct filter = Struct.newBuilder()
                  .putFields("genre", Value.newBuilder()
                          .setStructValue(Struct.newBuilder()
                                  .putFields("$eq", Value.newBuilder()
                                          .setStringValue("documentary")
                                          .build()))
                          .build())
                  .build();
          index.deleteByFilter(filter, "example-namespace");
          
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      metadataFilter := map[string]interface{}{
  		"genre": map[string]interface{}{
  			"$eq": "documentary",
  		},
      }

      filter, err := structpb.NewStruct(metadataFilter)
      if err != nil {
          log.Fatalf("Failed to create metadata filter: %v", err)
      }

      err = idxConnection.DeleteVectorsByFilter(ctx, filter)
      if err != nil {
          log.Fatalf("Failed to delete vector(s) with filter %+v: %v", filter, err)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var deleteResponse = await index.DeleteAsync(new DeleteRequest {
      Namespace = "example-namespace",
      Filter = new Metadata
      {
          ["genre"] =
              new Metadata
              {
                  ["$eq"] = "documentary"
              }
      }
  });
  ```

  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -i "https://$INDEX_HOST/vectors/delete" \
    -H 'Api-Key: $PINECONE_API_KEY' \
    -H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "filter": {"genre": {"$eq": "documentary"}},
      "namespace": "example-namespace"
    }'
  ```
</CodeGroup>

## Delete all records in a namespace

To delete all of the records in a namespace but not the namespace itself, provide a `namespace` parameter and specify the appropriate `deleteAll` parameter for your SDK. To target the default namespace, set `namespace` to `"__default__"`.

<CodeGroup>
  ```Python Python theme={null}
  # pip install "pinecone[grpc]"
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  index.delete(delete_all=True, namespace='example-namespace')
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  await index.namespace('example-namespace').deleteAll();
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;

  import java.util.Arrays;
  import java.util.List;

  public class DeleteExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");
          index.deleteAll("example-namespace");
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      err = idxConnection.DeleteAllVectorsInNamespace(ctx)
      if err != nil {
          log.Fatalf("Failed to delete all vectors in namespace %v: %v", namespace, err)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var deleteResponse = await index.DeleteAsync(new DeleteRequest {
      DeleteAll = true,
      Namespace = "example-namespace",
  });
  ```

  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/vectors/delete" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "deleteAll": true,
      "namespace": "example-namespace"
    }
  '
  ```
</CodeGroup>

## Delete an entire namespace

To delete an entire namespace and all of its records, see [Delete a namespace](/guides/manage-data/manage-namespaces#delete-a-namespace).

## Delete an entire index

To remove all records from an index, [delete the index](/guides/manage-data/manage-indexes#delete-an-index) and [recreate it](/guides/index-data/create-an-index).

## Delete limits

**Delete by ID limits:**

| Metric              | Limit                                          |
| :------------------ | :--------------------------------------------- |
| Max IDs per request | 1000 IDs                                       |
| Max request rate    | 5000 records per second per index or namespace |

**Delete by metadata limits:**

| Metric           | Limit                                                                      |
| :--------------- | :------------------------------------------------------------------------- |
| Max request rate | 5 requests per second per namespace<br />500 requests per second per index |

## Data freshness

Pinecone is eventually consistent, so there can be a slight delay before new or changed records are visible to queries. You can view index stats to [check data freshness](/guides/index-data/check-data-freshness).

---

# Source: https://docs.pinecone.io/troubleshooting/delete-your-account.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Delete your account

To delete your Pinecone account, you need to remove your user from all organizations and delete any organizations in which you are the sole member.

<Warning>These actions cannot be undone.</Warning>

* [How do I remove myself from an organization?](/guides/organizations/manage-organization-members#remove-a-member)
* [How do I delete an organization?](/troubleshooting/delete-your-organization)

Once you've removed yourself from or deleted all organizations associated with your user, your Pinecone account no longer exists.

---

# Source: https://docs.pinecone.io/troubleshooting/delete-your-organization.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Delete your organization

If you want to delete your Pinecone organization entirely, you'll need to delete all projects, which first requires deleting all indexes and collections, and downgrade to the Starter plan.

* [Delete an index](/guides/manage-data/manage-indexes#delete-an-index)
* [Delete a project](/guides/projects/manage-projects#delete-a-project)
* [Downgrade to the Starter plan](/guides/organizations/manage-billing/downgrade-billing-plan)

Once you've downgraded to the Starter plan and deleted your indexes and projects, you can delete your organization.

<Note>This action cannot be undone.</Note>

1. Go to [Settings > Manage](https://app.pinecone.io/organizations/-/settings/manage) in the Pinecone console.
2. Click **Delete this organization**.
3. Type the organization name and confirm the deletion.

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/delete.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Delete vectors

> Delete vectors by id from a single namespace.

For guidance and examples, see [Delete data](https://docs.pinecone.io/guides/manage-data/delete-data).

<RequestExample>
  ```shell curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/vectors/delete" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "ids": [
        "id-1", 
        "id-2"
      ],
      "namespace": "example-namespace"
    }
  '
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {}
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml post /vectors/delete
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /vectors/delete:
    post:
      tags:
        - Vector Operations
      summary: Delete vectors
      description: >-
        Delete vectors by id from a single namespace.


        For guidance and examples, see [Delete
        data](https://docs.pinecone.io/guides/manage-data/delete-data).
      operationId: deleteVectors
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/DeleteRequest'
        required: true
      responses:
        '200':
          description: A successful response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/DeleteResponse'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    DeleteRequest:
      description: The request for the `delete` operation.
      type: object
      properties:
        ids:
          example:
            - id-0
            - id-1
          description: Vectors to delete.
          type: array
          items:
            type: string
          minLength: 1
          maxLength: 1000
        deleteAll:
          example: false
          description: >-
            This indicates that all vectors in the index namespace should be
            deleted.
          default: false
          type: boolean
        namespace:
          example: example-namespace
          description: The namespace to delete vectors from, if applicable.
          type: string
        filter:
          description: >-
            If specified, the metadata filter here will be used to select the
            vectors to delete. This is mutually exclusive with specifying ids to
            delete in the ids param or using delete_all=True. See [Delete
            data](https://docs.pinecone.io/guides/manage-data/delete-data#delete-records-by-metadata).
          type: object
    DeleteResponse:
      description: The response for the `delete` operation.
      type: object
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/admin/delete_api_key.md

# Source: https://docs.pinecone.io/reference/api/2025-10/admin-assistant/delete_api_key.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Delete an API key

> Delete an API key from a project.

<RequestExample>
  ```bash curl theme={null}
  PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"
  PINECONE_API_KEY_ID="YOUR_KEY_ID"

  curl -X DELETE "https://api.pinecone.io/admin/api-keys/$PINECONE_API_KEY_ID" \
  	-H "X-Pinecone-Api-Version: 2025-10" \
  	-H "Authorization: Bearer $PINECONE_ACCESS_TOKEN"
  ```
</RequestExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/admin_2025-10.oas.yaml delete /admin/api-keys/{api_key_id}
openapi: 3.0.3
info:
  title: Pinecone Admin API
  description: |
    Provides an API for managing a Pinecone organization and its resources.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - BearerAuth: []
tags:
  - name: API Keys
    description: Actions that manage API Keys.
  - name: Organizations
    description: Actions that manage organizations.
  - name: Projects
    description: Actions that manage projects.
paths:
  /admin/api-keys/{api_key_id}:
    delete:
      tags:
        - API Keys
      summary: Delete an API key
      description: Delete an API key from a project.
      operationId: delete_api_key
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: api_key_id
          description: API key ID
          required: true
          schema:
            type: string
            format: uuid
          style: simple
      responses:
        '202':
          description: API key deletion request accepted.
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '403':
          description: Forbidden
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '404':
          description: Not found
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
        4XX:
          description: Unexpected error on request.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
components:
  schemas:
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, or `UNPROCESSABLE_ENTITY`.        
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    BearerAuth:
      type: http
      scheme: bearer
      description: >
        An [access
        token](https://docs.pinecone.io/guides/organizations/manage-service-accounts#retrieve-an-access-token)
        must be provided in the `Authorization` header using the `Bearer`
        scheme.

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/assistant/delete_assistant.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Delete an assistant

> Delete an existing assistant.

For guidance and examples, see [Manage assistants](https://docs.pinecone.io/guides/assistant/manage-assistants#delete-an-assistant)

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl -X DELETE "https://api.pinecone.io/assistant/assistants/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \ 
    -H "X-Pinecone-Api-Version: 2025-10" 
  ```
</RequestExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/assistant_control_2025-10.oas.yaml delete /assistants/{assistant_name}
openapi: 3.0.3
info:
  title: Pinecone Assistant Control Plane API
  description: >-
    Pinecone Assistant Engine is a context engine to store and retrieve relevant
    knowledge  from millions of documents at scale. This API supports creating
    and managing assistants. 
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io/assistant
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Assistants
    description: Actions that manage Assistants
paths:
  /assistants/{assistant_name}:
    delete:
      tags:
        - Manage Assistants
      summary: Delete an assistant
      description: >-
        Delete an existing assistant.


        For guidance and examples, see [Manage
        assistants](https://docs.pinecone.io/guides/assistant/manage-assistants#delete-an-assistant)
      operationId: delete_assistant
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: assistant_name
          description: The name of the assistant to delete.
          required: true
          schema:
            type: string
          example: my-assistant
          style: simple
      responses:
        '200':
          description: The request to delete the assistant has been accepted.
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Assistant not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                assistant-not-found:
                  summary: Assistant not found.
                  value:
                    error:
                      code: NOT_FOUND
                      message: Assistant "example-assistant" not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: Uploaded file can only currently be either a pdf or txt file
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: Pinecone API Key

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/delete_backup.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Delete a backup

> Delete a backup.

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  BACKUP_ID="9947520e-d5a1-4418-a78d-9f464c9969da"

  curl -X DELETE "https://api.pinecone.io/backups/$BACKUP_ID" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample />


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml delete /backups/{backup_id}
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /backups/{backup_id}:
    delete:
      tags:
        - Manage Indexes
      summary: Delete a backup
      description: Delete a backup.
      operationId: delete_backup
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: backup_id
          description: The ID of the backup to delete.
          required: true
          schema:
            type: string
          example: 670e8400-e29b-41d4-a716-446655440000
          style: simple
      responses:
        '202':
          description: The request to delete the backup has been accepted.
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Backup not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                backup-not-found:
                  summary: Backup not found
                  value:
                    error:
                      code: NOT_FOUND
                      message: Backup bkp_123abc not found.
                    status: 404
        '412':
          description: There is a pending restore job created from this backup.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                pending-restore:
                  summary: There is a pending restore job from this backup.
                  value:
                    error:
                      code: FAILED_PRECONDITION
                      message: >-
                        Unable to delete backup. There are pending restore jobs
                        for this backup:
                        ['670e8400-e29b-41d4-a716-446655440000']
                    status: 412
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/delete_collection.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Delete a collection

> Delete an existing collection.
Serverless indexes do not support collections.


<RequestExample>
  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X DELETE "https://api.pinecone.io/collections/example-collection" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample />


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml delete /collections/{collection_name}
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /collections/{collection_name}:
    delete:
      tags:
        - Manage Indexes
      summary: Delete a collection
      description: |
        Delete an existing collection.
        Serverless indexes do not support collections.
      operationId: delete_collection
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: collection_name
          description: The name of the collection.
          required: true
          schema:
            type: string
          example: test-collection
          style: simple
      responses:
        '202':
          description: The collection has been successfully deleted.
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Collection not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                collection-not-found:
                  summary: Collection not found.
                  value:
                    error:
                      code: NOT_FOUND
                      message: Collection example-collection not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/assistant/delete_file.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Delete an uploaded file

> Delete an uploaded file from an assistant.

For guidance and examples, see [Manage files](https://docs.pinecone.io/guides/assistant/manage-files#delete-a-file).

<Note>You can delete a file while it is still processing. You do not need to wait for processing to complete.</Note>

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"
  FILE_ID="070513b3-022f-4966-b583-a9b12e0290ff"

  curl -X DELETE "https://prod-1-data.ke.pinecone.io/assistant/files/$ASSISTANT_NAME/$FILE_ID" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/assistant_data_2025-10.oas.yaml delete /files/{assistant_name}/{assistant_file_id}
openapi: 3.0.3
info:
  title: Pinecone Assistant Data Plane API
  description: >-
    Pinecone Assistant Engine is a context engine to store and retrieve relevant
    knowledge from millions of documents at scale. This API supports
    interactions with assistants.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{assistant_host}
    variables:
      assistant_host:
        default: unknown
        description: host of the created assistant
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Assistants
    description: Actions that manage Assistants
paths:
  /files/{assistant_name}/{assistant_file_id}:
    delete:
      tags:
        - Manage Assistants
      summary: Delete an uploaded file
      description: >-
        Delete an uploaded file from an assistant.


        For guidance and examples, see [Manage
        files](https://docs.pinecone.io/guides/assistant/manage-files#delete-a-file).
      operationId: delete_file
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: assistant_name
          description: The name of the assistant to upload files to.
          required: true
          schema:
            type: string
          example: test-model
          style: simple
        - in: path
          name: assistant_file_id
          description: The uuid of the file to be described.
          required: true
          schema:
            type: string
          example: 72490b32-46d9-4db1-b48b-666e9176d9be
          style: simple
      responses:
        '200':
          description: The request to delete the file has been accepted.
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: File not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                file-not-found:
                  summary: File not found.
                  value:
                    error:
                      code: NOT_FOUND
                      message: >-
                        File with id 72490b32-46d9-4db1-b48b-666e9176d9be  not
                        found in provided assistant
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: Uploaded file can only currently be either a pdf or txt file
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: Pinecone API Key

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/delete_index.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Delete an index

> Delete an existing index.

<RequestExample>
  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X DELETE "https://api.pinecone.io/indexes/docs-example" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample />


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml delete /indexes/{index_name}
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /indexes/{index_name}:
    delete:
      tags:
        - Manage Indexes
      summary: Delete an index
      description: Delete an existing index.
      operationId: delete_index
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: index_name
          description: The name of the index to delete.
          required: true
          schema:
            type: string
          example: test-index
          style: simple
      responses:
        '202':
          description: The request to delete the index has been accepted.
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '403':
          description: Forbidden. Deletion protection enabled.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                deletion-protection:
                  summary: Forbidden
                  value:
                    error:
                      code: FORBIDDEN
                      message: >-
                        Deletion protection is enabled for this index. Disable
                        deletion protection before retrying.
                    status: 403
        '404':
          description: Index not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-not-found:
                  summary: Index not found
                  value:
                    error:
                      code: NOT_FOUND
                      message: Index example-index not found.
                    status: 404
        '412':
          description: There is a pending collection created from this index.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                pending-collection:
                  summary: There is a pending collection from this index.
                  value:
                    error:
                      code: FAILED_PRECONDITION
                      message: >-
                        Unable to delete an index. There are pending collections
                        for this index: ['test-collection']
                    status: 412
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/admin/delete_project.md

# Source: https://docs.pinecone.io/reference/api/2025-10/admin-assistant/delete_project.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Delete a project

> Delete a project and all its associated configuration.
Before deleting a project, you must delete all indexes, assistants, backups, and collections associated with the project. Other project resources, such as API keys, are automatically deleted when the project is deleted.


<RequestExample>
  ```bash curl theme={null}
  PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"
  PROJECT_ID="YOUR_PROJECT_ID"

  curl -X DELETE "https://api.pinecone.io/admin/projects/$PROJECT_ID" \
  	-H "X-Pinecone-Api-Version: 2025-10" \
  	-H "Authorization: Bearer $PINECONE_ACCESS_TOKEN"
  ```
</RequestExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/admin_2025-10.oas.yaml delete /admin/projects/{project_id}
openapi: 3.0.3
info:
  title: Pinecone Admin API
  description: |
    Provides an API for managing a Pinecone organization and its resources.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - BearerAuth: []
tags:
  - name: API Keys
    description: Actions that manage API Keys.
  - name: Organizations
    description: Actions that manage organizations.
  - name: Projects
    description: Actions that manage projects.
paths:
  /admin/projects/{project_id}:
    delete:
      tags:
        - Projects
      summary: Delete a project
      description: >
        Delete a project and all its associated configuration.

        Before deleting a project, you must delete all indexes, assistants,
        backups, and collections associated with the project. Other project
        resources, such as API keys, are automatically deleted when the project
        is deleted.
      operationId: delete_project
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: project_id
          description: Project ID
          required: true
          schema:
            type: string
            format: uuid
          style: simple
      responses:
        '202':
          description: Project deletion request accepted.
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '403':
          description: Forbidden
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '404':
          description: Not found
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
        4XX:
          description: Unexpected error on request.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
components:
  schemas:
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, or `UNPROCESSABLE_ENTITY`.        
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    BearerAuth:
      type: http
      scheme: bearer
      description: >
        An [access
        token](https://docs.pinecone.io/guides/organizations/manage-service-accounts#retrieve-an-access-token)
        must be provided in the `Authorization` header using the `Bearer`
        scheme.

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/deletenamespace.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Delete a namespace

> Delete a namespace from a serverless index. Deleting a namespace is irreversible; all data in the namespace is permanently deleted.

For guidance and examples, see [Manage namespaces](https://docs.pinecone.io/guides/manage-data/manage-namespaces).

**Note:** This operation is not supported for pod-based indexes.

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="YOUR_INDEX_HOST"
  NAMESPACE="YOUR_NAMESPACE" # To target the default namespace, use "__default__".

  curl -X DELETE "https://$INDEX_HOST/namespaces/$NAMESPACE" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample />


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml delete /namespaces/{namespace}
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /namespaces/{namespace}:
    delete:
      tags:
        - Namespace Operations
      summary: Delete a namespace
      description: >-
        Delete a namespace from a serverless index. Deleting a namespace is
        irreversible; all data in the namespace is permanently deleted.


        For guidance and examples, see [Manage
        namespaces](https://docs.pinecone.io/guides/manage-data/manage-namespaces).


        **Note:** This operation is not supported for pod-based indexes.
      operationId: deleteNamespace
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: namespace
          description: The namespace to delete.
          required: true
          schema:
            type: string
          style: simple
      responses:
        '200':
          description: A successful response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/DeleteResponse'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    DeleteResponse:
      description: The response for the `delete` operation.
      type: object
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/assistant/describe_assistant.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Check assistant status

> Get the status of an assistant.

For guidance and examples, see [Manage assistants](https://docs.pinecone.io/guides/assistant/manage-assistants#get-the-status-of-an-assistant)

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl -X GET "https://api.pinecone.io/assistant/assistants/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \ 
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "name": "example-assistant",
    "instructions": "Use American English for spelling and grammar.",
    "metadata": {"team": "customer-support", "version": "1.0"},
    "status": "Ready",
    "host": "https://prod-1-data.ke.pinecone.io",
    "created_at": "2025-10-01T12:30:00Z",
    "updated_at": "2025-10-01T12:45:00Z"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/assistant_control_2025-10.oas.yaml get /assistants/{assistant_name}
openapi: 3.0.3
info:
  title: Pinecone Assistant Control Plane API
  description: >-
    Pinecone Assistant Engine is a context engine to store and retrieve relevant
    knowledge  from millions of documents at scale. This API supports creating
    and managing assistants. 
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io/assistant
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Assistants
    description: Actions that manage Assistants
paths:
  /assistants/{assistant_name}:
    get:
      tags:
        - Manage Assistants
      summary: Check assistant status
      description: >-
        Get the status of an assistant.


        For guidance and examples, see [Manage
        assistants](https://docs.pinecone.io/guides/assistant/manage-assistants#get-the-status-of-an-assistant)
      operationId: get_assistant
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: assistant_name
          description: The name of the assistant to get a status on.
          required: true
          schema:
            type: string
          example: test-model
          style: simple
      responses:
        '200':
          description: Poll request successful.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Assistant'
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Assistant not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                assistant-not-found:
                  summary: Assistant not found.
                  value:
                    error:
                      code: NOT_FOUND
                      message: Assistant "example-assistant" not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    Assistant:
      description: Describes the configuration and status of a Pinecone Assistant.
      type: object
      properties:
        name:
          example: example-assistant
          description: >
            The name of the assistant. Resource name must be 1-63 characters
            long, start and end with an alphanumeric character, and consist only
            of lower case alphanumeric characters or '-'.
          type: string
          minLength: 1
          maxLength: 63
        instructions:
          nullable: true
          description: >-
            Description or directive for the assistant to apply to all
            responses.
          type: string
        metadata:
          description: >-
            Optional metadata associated with the assistant. Metadata is a JSON
            object that can store custom organizational data, tags, and
            attributes.
          nullable: true
          type: object
        status:
          x-enum:
            - Initializing
            - Failed
            - Ready
            - Terminating
            - InitializationFailed
          type: string
        host:
          description: The host where the assistant is deployed.
          type: string
        created_at:
          type: string
          format: date-time
        updated_at:
          type: string
          format: date-time
      required:
        - name
        - status
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: Uploaded file can only currently be either a pdf or txt file
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: Pinecone API Key

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/describe_backup.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Describe a backup

> Get a description of a backup.

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  BACKUP_ID="8c85e612-ed1c-4f97-9f8c-8194e07bcf71"

  curl -X GET "https://api.pinecone.io/backups/$BACKUP_ID" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -H "accept: application/json"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "backup_id":"8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
    "source_index_id":"f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
    "source_index_name":"docs-example",
    "tags":{},
    "name":"example-backup",
    "description":"Monthly backup of production index",
    "status":"Ready",
    "cloud":"aws",
    "region":"us-east-1",
    "dimension":1024,
    "record_count":98,
    "namespace_count":3,
    "size_bytes":1069169,
    "created_at":"2025-03-11T18:29:50.549505Z"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml get /backups/{backup_id}
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /backups/{backup_id}:
    get:
      tags:
        - Manage Indexes
      summary: Describe a backup
      description: Get a description of a backup.
      operationId: describe_backup
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: backup_id
          description: The ID of the backup to describe.
          required: true
          schema:
            type: string
          example: 670e8400-e29b-41d4-a716-446655440000
          style: simple
      responses:
        '200':
          description: Configuration information and deployment status of the backup.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/BackupModel'
              examples:
                backup:
                  summary: A backup
                  value:
                    backup_id: 670e8400-e29b-41d4-a716-446655440000
                    cloud: aws
                    created_at: '2024-03-15T10:30:00.000Z'
                    description: Monthly backup of production index
                    dimension: 1536
                    metric: cosine
                    name: backup_2024_03_15
                    namespace_count: 3
                    record_count: 120000
                    region: us-east-1
                    schema:
                      fields:
                        genre:
                          filterable: true
                        title:
                          filterable: true
                    size_bytes: 10000000
                    source_index_id: 670e8400-e29b-41d4-a716-446655440001
                    source_index_name: my-index
                    status: Ready
                    tags:
                      environment: production
                      type: monthly
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Backup not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                backup-not-found:
                  summary: Backup not found
                  value:
                    error:
                      code: NOT_FOUND
                      message: Backup bkp_123abc not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    BackupModel:
      description: >-
        The BackupModel describes the configuration and status of a Pinecone
        backup.
      type: object
      properties:
        backup_id:
          example: 670e8400-e29b-41d4-a716-446655440001
          description: Unique identifier for the backup.
          type: string
        source_index_name:
          example: my-index
          description: Name of the index from which the backup was taken.
          type: string
        source_index_id:
          example: 670e8400-e29b-41d4-a716-446655440000
          description: ID of the index.
          type: string
        name:
          example: backup-2025-02-04
          description: Optional user-defined name for the backup.
          type: string
        description:
          example: Backup before bulk update.
          description: Optional description providing context for the backup.
          type: string
        status:
          example: Ready
          description: Current status of the backup (e.g., Initializing, Ready, Failed).
          type: string
        cloud:
          example: aws
          description: Cloud provider where the backup is stored.
          type: string
        region:
          example: us-east-1
          description: Cloud region where the backup is stored.
          type: string
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If the 'vector_type' is
            'sparse', the metric must be 'dotproduct'. If the `vector_type` is
            `dense`, the metric defaults to 'cosine'.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        schema:
          $ref: '#/components/schemas/MetadataSchema'
        record_count:
          example: 120000
          description: Total number of records in the backup.
          type: integer
        namespace_count:
          example: 3
          description: Number of namespaces in the backup.
          type: integer
        size_bytes:
          example: 10000000
          description: Size of the backup in bytes.
          type: integer
        tags:
          $ref: '#/components/schemas/IndexTags'
        created_at:
          description: Timestamp when the backup was created.
          type: string
      required:
        - backup_id
        - source_index_name
        - source_index_id
        - status
        - cloud
        - region
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    MetadataSchema:
      example:
        fields:
          description:
            filterable: true
          genre:
            filterable: true
          year:
            filterable: true
      description: >-
        Schema for the behavior of Pinecone's internal metadata index. By
        default, all metadata is indexed; when `schema` is present, only fields
        which are present in the `fields` object with a `filterable: true` are
        indexed. Note that `filterable: false` is not currently supported.
      type: object
      properties:
        fields:
          description: >-
            A map of metadata field names to their configuration. The field name
            must be a valid metadata field name. The field name must be unique.
          type: object
          additionalProperties:
            type: object
            properties:
              filterable:
                description: >-
                  Whether the field is filterable. If true, the field is indexed
                  and can be used in filters. Only true values are allowed.
                type: boolean
      required:
        - fields
    IndexTags:
      example:
        tag0: val0
        tag1: val1
      description: >-
        Custom user tags added to an index. Keys must be 80 characters or less.
        Values must be 120 characters or less. Keys must be alphanumeric, '_',
        or '-'.  Values must be alphanumeric, ';', '@', '_', '-', '.', '+', or '
        '. To unset a key, set the value to be an empty string.
      type: object
      additionalProperties:
        type: string
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/describe_collection.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Describe a collection

> Get a description of a collection.
Serverless indexes do not support collections.


<RequestExample>
  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X GET "https://api.pinecone.io/collections/tiny-collection" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
      "name": "example-collection",
      "status": "Ready",
      "environment": "us-east-1-aws",
      "size": 3075398,
      "vector_count": 99,
      "dimension": 1536
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml get /collections/{collection_name}
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /collections/{collection_name}:
    get:
      tags:
        - Manage Indexes
      summary: Describe a collection
      description: |
        Get a description of a collection.
        Serverless indexes do not support collections.
      operationId: describe_collection
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: collection_name
          description: The name of the collection to be described.
          required: true
          schema:
            type: string
          example: tiny-collection
          style: simple
      responses:
        '200':
          description: Configuration information and status of the collection.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/CollectionModel'
              examples:
                tiny-collection:
                  summary: A small collection.
                  value:
                    dimension: 3
                    environment: us-east1-gcp
                    name: tiny-collection
                    size: 3126700
                    status: Ready
                    vector_count: 99
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Collection not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                collection-not-found:
                  summary: Collection not found.
                  value:
                    error:
                      code: NOT_FOUND
                      message: Collection example-collection not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    CollectionModel:
      description: >-
        The CollectionModel describes the configuration and status of a Pinecone
        collection.
      type: object
      properties:
        name:
          example: example-collection
          description: The name of the collection.
          type: string
        size:
          example: 10000000
          description: The size of the collection in bytes.
          type: integer
          format: int64
        status:
          example: Initializing
          description: |-
            The status of the collection.
            Possible values: `Initializing`, `Ready`, or `Terminating`.
          x-enum:
            - Initializing
            - Ready
            - Terminating
          type: string
        dimension:
          example: 1536
          description: >-
            The dimension of the vectors stored in each record held in the
            collection.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        vector_count:
          example: 120000
          description: The number of records stored in the collection.
          type: integer
          format: int32
        environment:
          example: us-east1-gcp
          description: The environment where the collection is hosted.
          type: string
      required:
        - name
        - status
        - environment
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/assistant/describe_file.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Describe a file upload

> Get the status and metadata of a file uploaded to an assistant.

For guidance and examples, see [Manage files](https://docs.pinecone.io/guides/assistant/manage-files#get-the-status-of-a-file).

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"
  FILE_ID="3c90c3cc-0d44-4b50-8888-8dd25736052a"

  # Describe a file. 
  # To get a signed URL in the response, set `include_url` to `true`.
  curl -X GET "https://prod-1-data.ke.pinecone.io/assistant/files/$ASSISTANT_NAME/$FILE_ID?include_url=true" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/assistant_data_2025-10.oas.yaml get /files/{assistant_name}/{assistant_file_id}
openapi: 3.0.3
info:
  title: Pinecone Assistant Data Plane API
  description: >-
    Pinecone Assistant Engine is a context engine to store and retrieve relevant
    knowledge from millions of documents at scale. This API supports
    interactions with assistants.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{assistant_host}
    variables:
      assistant_host:
        default: unknown
        description: host of the created assistant
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Assistants
    description: Actions that manage Assistants
paths:
  /files/{assistant_name}/{assistant_file_id}:
    get:
      tags:
        - Manage Assistants
      summary: Describe a file upload
      description: >-
        Get the status and metadata of a file uploaded to an assistant.


        For guidance and examples, see [Manage
        files](https://docs.pinecone.io/guides/assistant/manage-files#get-the-status-of-a-file).
      operationId: describe_file
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: assistant_name
          description: The name of the assistant to upload files to.
          required: true
          schema:
            type: string
          example: test-model
          style: simple
        - in: path
          name: assistant_file_id
          description: The uuid of the file to be described.
          required: true
          schema:
            type: string
          example: 72490b32-46d9-4db1-b48b-666e9176d9be
          style: simple
        - in: query
          name: include_url
          description: Include the signed URL of the file in the response.
          schema:
            x-enum:
              - 'true'
              - 'false'
            type: string
          style: form
      responses:
        '200':
          description: Poll request successful.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/AssistantFileModel'
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: File not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                file-not-found:
                  summary: File not found.
                  value:
                    error:
                      code: NOT_FOUND
                      message: >-
                        File with id 72490b32-46d9-4db1-b48b-666e9176d9be  not
                        found in provided assistant
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    AssistantFileModel:
      description: The response format for a successful file upload request.
      type: object
      properties:
        name:
          type: string
        id:
          type: string
          format: uuid
        metadata:
          nullable: true
          type: object
        created_on:
          type: string
          format: date-time
        updated_on:
          type: string
          format: date-time
        status:
          description: >-
            The current state of the uploaded file. Possible values:

            - `Processing`: File is being processed (parsed, chunked, embedded)

            - `Available`: Processing completed successfully; file is ready for
            use

            - `Deleting`: Deletion has been initiated but not yet completed

            - `ProcessingFailed`: Processing failed with an error


            Note: Once a file is deleted, the API returns 404 Not Found instead
            of a file object.
          x-enum:
            - Processing
            - Available
            - Deleting
            - ProcessingFailed
          type: string
        percent_done:
          nullable: true
          description: The percentage of the file that has been processed
          type: number
          format: double
        signed_url:
          nullable: true
          example: https://storage.googleapis.com/bucket/file.pdf?...
          description: >-
            A [signed
            URL](https://cloud.google.com/storage/docs/access-control/signed-urls)
            that provides temporary, read-only access to the underlying file.
            Anyone with the link can access the file, so treat it as sensitive
            data. Expires after a short time.
          type: string
        error_message:
          nullable: true
          description: >-
            A message describing any error during file processing. Provided only
            if an error occurs.
          type: string
        multimodal:
          description: Indicates whether the file was processed as multimodal.
          type: boolean
      required:
        - id
        - name
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: Uploaded file can only currently be either a pdf or txt file
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: Pinecone API Key

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/describe_import.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Describe an import

> Return details of a specific import operation.

For guidance and examples, see [Import data](https://docs.pinecone.io/guides/index-data/import-data).

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Standard and Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

<RequestExample>
  ```bash curl theme={null}
  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X GET "https://$INDEX_HOST/bulk/imports/101" \
    -H 'Api-Key: $PINECONE_API_KEY' \
    -H 'X-Pinecone-Api-Version: 2025-10'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "id": "101",
    "uri": "s3://BUCKET_NAME/PATH/TO/DIR",
    "status": "Pending",
    "created_at": "2024-08-19T20:49:00.754Z",
    "finished_at": "2024-08-19T20:49:00.754Z",
    "percent_complete": 42.2,
    "records_imported": 1000000
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml get /bulk/imports/{id}
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /bulk/imports/{id}:
    get:
      tags:
        - Bulk Operations
      summary: Describe an import
      description: >-
        Return details of a specific import operation.


        For guidance and examples, see [Import
        data](https://docs.pinecone.io/guides/index-data/import-data).
      operationId: describeBulkImport
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: id
          description: Unique identifier for the import operation.
          required: true
          schema:
            type: string
            minLength: 1
            maxLength: 1000
          example: '101'
          style: simple
      responses:
        '200':
          description: Details of the import operation.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ImportModel'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    ImportModel:
      description: The model for an import operation.
      type: object
      properties:
        id:
          example: '101'
          description: Unique identifier for the import operation.
          type: string
          minLength: 1
          maxLength: 1000
        uri:
          description: The URI from where the data is imported.
          type: string
        status:
          example: Pending
          description: >-
            The status of the operation.

            Possible values: `Pending`, `InProgress`, `Failed`, `Completed`, or
            `Cancelled`.
          x-enum:
            - Pending
            - InProgress
            - Failed
            - Completed
            - Cancelled
          type: string
        createdAt:
          description: The start time of the import operation.
          type: string
          format: date-time
        finishedAt:
          description: The end time of the import operation.
          type: string
          format: date-time
        percentComplete:
          example: 42.2
          description: The progress made by the operation, as a percentage.
          type: number
          format: float
          minimum: 0
          maximum: 100
        recordsImported:
          example: 1000000
          description: The number of records successfully imported.
          type: integer
          format: int64
        error:
          description: The error message if the import process failed.
          type: string
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/describe_index.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Describe an index

> Get a description of an index.

<RequestExample>
  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_NAME="YOUR_INDEX_NAME"

  curl "https://api.pinecone.io/indexes/$INDEX_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" 
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  // EXAMPLE RESPONSE 1: Serverless index (on-demand)
  {
    "name": "example-serverless-ondemand-index",
    "vector_type": "dense",
    "metric": "cosine",
    "dimension": 1024,
    "status": {
      "ready": true,
      "state": "Ready"
    },
    "host": "example-serverless-ondemand-index-bhnyigt.svc.aped-4627-b74a.pinecone.io",
    "spec": {
      "serverless": {
        "region": "us-east-1",
        "cloud": "aws",
        "read_capacity": {
          "mode": "OnDemand",
          "status": {
            "state": "Ready",
            "current_shards": null,
            "current_replicas": null
          }
        }
      }
    },
    "deletion_protection": "enabled",
    "tags": {
      "tag1": "value1",
      "tag2": "value2"
    },
    "embed": {
      "model": "llama-text-embed-v2",
      "field_map": {
        "text": "text"
      },
      "dimension": 1024,
      "metric": "cosine",
      "write_parameters": {
        "dimension": 1024,
        "input_type": "passage",
        "truncate": "END"
      },
      "read_parameters": {
        "dimension": 1024,
        "input_type": "query",
        "truncate": "END"
      },
      "vector_type": "dense"
    }
  }

  // EXAMPLE RESPONSE 2: Serverless index (dedicated)
  {
    "name": "example-serverless-dedicated-index",
    "vector_type": "dense",
    "metric": "cosine",
    "dimension": 1536,
    "status": {
      "ready": true,
      "state": "Ready"
    },
    "host": "example-serverless-dedicated-index-bhnyigt.svc.aped-4627-b74a.pinecone.io",
    "spec": {
      "serverless": {
        "region": "us-east-1",
        "cloud": "aws",
        "read_capacity": {
          "mode": "Dedicated",
          "dedicated": {
            "node_type": "b1",
            "scaling": "Manual",
            "manual": {
              "shards": 1,
              "replicas": 2
            }
          },
          "status": {
            "state": "Scaling",
            "current_shards": 1,
            "current_replicas": 1
          }
        }
      }
    },
    "deletion_protection": "enabled",
    "tags": {
      "tag0": "value0",
      "tag1": "value1"
    }
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml get /indexes/{index_name}
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /indexes/{index_name}:
    get:
      tags:
        - Manage Indexes
      summary: Describe an index
      description: Get a description of an index.
      operationId: describe_index
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: index_name
          description: The name of the index to be described.
          required: true
          schema:
            type: string
          example: test-index
          style: simple
      responses:
        '200':
          description: Configuration information and deployment status of the index.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/IndexModel'
              examples:
                movie-recommendations-serverless:
                  summary: A serverless index
                  value:
                    dimension: 1536
                    host: movie-recommendations-c01b5b5.svc.us-east1-gcp.pinecone.io
                    metric: cosine
                    name: movie-recommendations
                    spec:
                      serverless:
                        cloud: aws
                        read_capacity:
                          mode: OnDemand
                          status:
                            state: Ready
                        region: us-east-1
                        schema:
                          fields:
                            genre:
                              filterable: true
                            title:
                              filterable: true
                    status:
                      ready: false
                      state: Initializing
                    vector_type: dense
                movie-recommendations-pod:
                  summary: A pod-based index
                  value:
                    dimension: 1536
                    host: movie-recommendations-c01b5b5.svc.us-east1-gcp.pinecone.io
                    metric: cosine
                    name: movie-recommendations
                    spec:
                      pod:
                        environment: us-east-1-aws
                        metadata_config:
                          indexed:
                            - genre
                            - title
                            - imdb_rating
                        pod_type: p1.x1
                        pods: 1
                        replicas: 1
                        shards: 1
                    status:
                      ready: false
                      state: Initializing
                    vector_type: dense
          links:
            UpsertVector:
              operationId: upsert
              server:
                url: $response.body#/host
            UpdateVector:
              operationId: update
              server:
                url: $response.body#/host
            QueryVector:
              operationId: query
              server:
                url: $response.body#/host
            FetchVector:
              operationId: fetch
              server:
                url: $response.body#/host
            DeleteOneVector:
              operationId: delete1
              server:
                url: $response.body#/host
            DeleteVector:
              operationId: delete
              server:
                url: $response.body#/host
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Index not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-not-found:
                  summary: Index not found
                  value:
                    error:
                      code: NOT_FOUND
                      message: Index example-index not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    IndexModel:
      description: >-
        The IndexModel describes the configuration and status of a Pinecone
        index.
      type: object
      properties:
        name:
          example: example-index
          description: >
            The name of the index. Resource name must be 1-45 characters long,
            start and end with an alphanumeric character, and consist only of
            lower case alphanumeric characters or '-'.
          type: string
          minLength: 1
          maxLength: 45
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If the 'vector_type' is
            'sparse', the metric must be 'dotproduct'. If the `vector_type` is
            `dense`, the metric defaults to 'cosine'.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        host:
          example: semantic-search-c01b5b5.svc.us-west1-gcp.pinecone.io
          description: The URL address where the index is hosted.
          type: string
        private_host:
          example: semantic-search-c01b5b5.svc.private.us-west1-gcp.pinecone.io
          description: The private endpoint URL of an index.
          type: string
        deletion_protection:
          $ref: '#/components/schemas/DeletionProtection'
        tags:
          $ref: '#/components/schemas/IndexTags'
        embed:
          $ref: '#/components/schemas/ModelIndexEmbed'
        spec:
          example:
            pod:
              environment: us-east-1-aws
              metadata_config:
                indexed:
                  - genre
                  - title
                  - imdb_rating
              pod_type: p1.x1
              pods: 1
              replicas: 1
              shards: 1
          description: The spec object defines how the index should be deployed.
          type: object
          oneOf:
            - title: Serverless
              type: object
              properties:
                serverless:
                  $ref: '#/components/schemas/ServerlessSpecResponse'
              required:
                - serverless
              additionalProperties: false
            - title: Pod-based
              type: object
              properties:
                pod:
                  $ref: '#/components/schemas/PodSpec'
              required:
                - pod
              additionalProperties: false
            - title: BYOC
              type: object
              properties:
                byoc:
                  $ref: '#/components/schemas/ByocSpecResponse'
              required:
                - byoc
              additionalProperties: false
        status:
          example:
            ready: true
            state: ScalingUpPodSize
          description: The current status of the index
          type: object
          properties:
            ready:
              description: Whether the index is ready for use
              type: boolean
            state:
              description: >-
                The state of the index.

                Possible values: `Initializing`, `InitializationFailed`,
                `ScalingUp`, `ScalingDown`, `ScalingUpPodSize`,
                `ScalingDownPodSize`, `Terminating`, `Ready`, or `Disabled`.
              x-enum:
                - Initializing
                - InitializationFailed
                - ScalingUp
                - ScalingDown
                - ScalingUpPodSize
                - ScalingDownPodSize
                - Terminating
                - Ready
                - Disabled
              type: string
          required:
            - ready
            - state
        vector_type:
          description: >-
            The index vector type. You can use 'dense' or 'sparse'. If 'dense',
            the vector dimension must be specified.  If 'sparse', the vector
            dimension should not be specified.
          default: dense
          type: string
      required:
        - name
        - metric
        - status
        - spec
        - host
        - vector_type
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    DeletionProtection:
      description: >-
        Whether [deletion
        protection](http://docs.pinecone.io/guides/manage-data/manage-indexes#configure-deletion-protection)
        is enabled/disabled for the index.

        Possible values: `disabled` or `enabled`.
      default: disabled
      x-enum:
        - disabled
        - enabled
      type: string
    IndexTags:
      example:
        tag0: val0
        tag1: val1
      description: >-
        Custom user tags added to an index. Keys must be 80 characters or less.
        Values must be 120 characters or less. Keys must be alphanumeric, '_',
        or '-'.  Values must be alphanumeric, ';', '@', '_', '-', '.', '+', or '
        '. To unset a key, set the value to be an empty string.
      type: object
      additionalProperties:
        type: string
    ModelIndexEmbed:
      example:
        field_map:
          text: your-text-field
        metric: cosine
        model: multilingual-e5-large
        read_parameters:
          input_type: query
          truncate: NONE
        write_parameters:
          input_type: passage
      description: The embedding model and document fields mapped to embedding inputs.
      type: object
      properties:
        model:
          example: multilingual-e5-large
          description: The name of the embedding model used to create the index.
          type: string
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If not specified, the metric
            will be defaulted according to the model. Cannot be updated once
            set.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        vector_type:
          description: >-
            The index vector type. You can use 'dense' or 'sparse'. If 'dense',
            the vector dimension must be specified.  If 'sparse', the vector
            dimension should not be specified.
          default: dense
          type: string
        field_map:
          example:
            text: your-text-field
          description: >-
            Identifies the name of the text field from your document model that
            is embedded.
          type: object
        read_parameters:
          description: The read parameters for the embedding model.
          type: object
        write_parameters:
          description: The write parameters for the embedding model.
          type: object
      required:
        - model
    ServerlessSpecResponse:
      description: Configuration of a serverless index.
      type: object
      properties:
        cloud:
          example: aws
          description: |-
            The public cloud where you would like your index hosted.
            Possible values: `gcp`, `aws`, or `azure`.
          x-enum:
            - gcp
            - aws
            - azure
          type: string
        region:
          example: us-east-1
          description: The region where you would like your index to be created.
          type: string
        read_capacity:
          $ref: '#/components/schemas/ReadCapacityResponse'
        source_collection:
          example: movie-embeddings
          description: The name of the collection to be used as the source for the index.
          type: string
        schema:
          $ref: '#/components/schemas/MetadataSchema'
      required:
        - cloud
        - region
        - read_capacity
    PodSpec:
      example:
        environment: us-east1-gcp
        metadata_config:
          indexed:
            - genre
            - title
            - imdb_rating
        pod_type: p1.x1
        pods: 1
        replicas: 1
        shards: 1
        source_collection: movie-embeddings
      title: Pod-based
      description: Configuration needed to deploy a pod-based index.
      type: object
      properties:
        environment:
          example: us-east1-gcp
          description: The environment where the index is hosted.
          type: string
        replicas:
          description: >-
            The number of replicas. Replicas duplicate your index. They provide
            higher availability and throughput. Replicas can be scaled up or
            down as your needs change.
          default: 1
          type: integer
          format: int32
          minimum: 1
        shards:
          description: >-
            The number of shards. Shards split your data across multiple pods so
            you can fit more data into an index.
          default: 1
          type: integer
          format: int32
          minimum: 1
        pod_type:
          description: >-
            The type of pod to use. One of `s1`, `p1`, or `p2` appended with `.`
            and one of `x1`, `x2`, `x4`, or `x8`.
          default: p1.x1
          type: string
        pods:
          example: 1
          description: >-
            The number of pods to be used in the index. This should be equal to
            `shards` x `replicas`.'
          default: 1
          type: integer
          minimum: 1
        metadata_config:
          example:
            indexed:
              - genre
              - title
              - imdb_rating
          description: >-
            Configuration for the behavior of Pinecone's internal metadata
            index. By default, all metadata is indexed; when `metadata_config`
            is present, only specified metadata fields are indexed. These
            configurations are only valid for use with pod-based indexes.
          type: object
          properties:
            indexed:
              description: >-
                By default, all metadata is indexed; to change this behavior,
                use this property to specify an array of metadata fields that
                should be indexed.
              type: array
              items:
                type: string
        source_collection:
          example: movie-embeddings
          description: The name of the collection to be used as the source for the index.
          type: string
      required:
        - environment
        - pod_type
    ByocSpecResponse:
      title: BYOC
      description: Configuration of a BYOC index.
      type: object
      properties:
        environment:
          example: aws-us-east-1-b921
          description: The environment where the index is hosted.
          type: string
        read_capacity:
          $ref: '#/components/schemas/ReadCapacityResponse'
        schema:
          $ref: '#/components/schemas/MetadataSchema'
      required:
        - environment
        - read_capacity
    ReadCapacityResponse:
      description: Response containing read capacity configuration
      discriminator:
        propertyName: mode
        mapping:
          OnDemand: '#/components/schemas/ReadCapacityOnDemandSpecResponse'
          Dedicated: '#/components/schemas/ReadCapacityDedicatedSpecResponse'
      oneOf:
        - $ref: '#/components/schemas/ReadCapacityOnDemandSpecResponse'
        - $ref: '#/components/schemas/ReadCapacityDedicatedSpecResponse'
    MetadataSchema:
      example:
        fields:
          description:
            filterable: true
          genre:
            filterable: true
          year:
            filterable: true
      description: >-
        Schema for the behavior of Pinecone's internal metadata index. By
        default, all metadata is indexed; when `schema` is present, only fields
        which are present in the `fields` object with a `filterable: true` are
        indexed. Note that `filterable: false` is not currently supported.
      type: object
      properties:
        fields:
          description: >-
            A map of metadata field names to their configuration. The field name
            must be a valid metadata field name. The field name must be unique.
          type: object
          additionalProperties:
            type: object
            properties:
              filterable:
                description: >-
                  Whether the field is filterable. If true, the field is indexed
                  and can be used in filters. Only true values are allowed.
                type: boolean
      required:
        - fields
    ReadCapacityOnDemandSpecResponse:
      example:
        mode: OnDemand
        status:
          state: Ready
      title: On-demand
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
        status:
          $ref: '#/components/schemas/ReadCapacityStatus'
      required:
        - mode
        - status
      additionalProperties: false
    ReadCapacityDedicatedSpecResponse:
      example:
        dedicated:
          manual:
            replicas: 2
            shards: 2
          node_type: t1
          scaling: Manual
        mode: Dedicated
        status:
          current_replicas: 2
          current_shards: 2
          state: Ready
      title: Dedicated
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
        dedicated:
          $ref: '#/components/schemas/ReadCapacityDedicatedConfig'
        status:
          $ref: '#/components/schemas/ReadCapacityStatus'
      required:
        - mode
        - dedicated
        - status
      additionalProperties: false
    ReadCapacityStatus:
      description: >-
        The current status of factors affecting the read capacity of a
        serverless index
      type: object
      properties:
        state:
          description: >-
            The `state` describes the overall status of factors relating to the
            read capacity of an index. 


            Available values:

            - `Ready` is the state most of the time

            - `Scaling` if the number of replicas or shards has been recently
            updated by calling the [configure index
            endpoint](https://docs.pinecone.io/reference/api/2025-10/control-plane/configure_index)

            - `Migrating` if the index is being migrated to a new `node_type`

            - `Error` if there is an error with the read capacity configuration.
            In that case, see `error_message` for more details.
          type: string
        current_replicas:
          description: >-
            The number of replicas. Each replica has dedicated  compute
            resources and data storage. Increasing this number  will increase
            the total throughput of the index.
          type: integer
          format: int32
        current_shards:
          description: >-
            The number of shards. Each shard has dedicated storage.  Increasing
            shards alleiviates index fullness. 
          type: integer
          format: int32
        error_message:
          description: >-
            An optional error message indicating any issues with your read
            capacity configuration
          type: string
      required:
        - state
    ReadCapacityDedicatedConfig:
      description: >-
        Configuration for dedicated read capacity. See  [this
        guide](https://docs.pinecone.io/guides/index-data/dedicated-read-nodes)
        for more details on  how to configure dedicated read capacity.
      type: object
      properties:
        node_type:
          description: >-
            The type of machines to use. Available options: `b1` and `t1`. `t1`
            includes increased processing power and memory.
          type: string
        scaling:
          description: The type of scaling strategy to use.
          type: string
        manual:
          $ref: '#/components/schemas/ScalingConfigManual'
    ScalingConfigManual:
      description: The config to use for manual read capacity scaling.
      type: object
      properties:
        replicas:
          description: >-
            The number of replicas to use. Replicas duplicate the compute
            resources and data of an index, allowing higher query throughput and
            availability. Setting replicas to 0 disables the index but can be
            used to reduce costs while usage is paused.
          type: integer
          format: int32
          minimum: 0
        shards:
          description: >-
            The number of shards to use. Shards determine the storage capacity
            of an index, with each shard providing 250 GB of storage.
          type: integer
          format: int32
          minimum: 1
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/inference/describe_model.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Describe a model

> Get a description of a model hosted by Pinecone. 

You can use hosted models as an integrated part of Pinecone operations or for standalone embedding and reranking. For more details, see [Vector embedding](https://docs.pinecone.io/guides/index-data/indexing-overview#vector-embedding) and [Rerank results](https://docs.pinecone.io/guides/search/rerank-results).

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://api.pinecone.io/models/llama-text-embed-v2" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "model": "llama-text-embed-v2",
    "short_description": "A high performance dense embedding model optimized for multilingual and cross-lingual text question-answering retrieval with support for long documents (up to 2048 tokens) and dynamic embedding size (Matryoshka Embeddings).",
    "type": "embed",
    "vector_type": "dense",
    "default_dimension": 1024,
    "modality": "text",
    "max_sequence_length": 2048,
    "max_batch_size": 96,
    "provider_name": "NVIDIA",
    "supported_metrics": [
      "Cosine",
      "DotProduct"
    ],
    "supported_dimensions": [
      384,
      512,
      768,
      1024,
      2048
    ],
    "supported_parameters": [
      {
        "parameter": "input_type",
        "required": true,
        "type": "one_of",
        "value_type": "string",
        "allowed_values": [
          "query",
          "passage"
        ]
      },
      {
        "parameter": "truncate",
        "required": false,
        "default": "END",
        "type": "one_of",
        "value_type": "string",
        "allowed_values": [
          "END",
          "NONE",
          "START"
        ]
      },
      {
        "parameter": "dimension",
        "required": false,
        "default": 1024,
        "type": "one_of",
        "value_type": "integer",
        "allowed_values": [
          384,
          512,
          768,
          1024,
          2048
        ]
      }
    ]
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/inference_2025-10.oas.yaml get /models/{model_name}
openapi: 3.0.3
info:
  title: Pinecone Inference API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Inference
    description: Model inference
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /models/{model_name}:
    get:
      tags:
        - Inference
      summary: Describe a model
      description: >-
        Get a description of a model hosted by Pinecone. 


        You can use hosted models as an integrated part of Pinecone operations
        or for standalone embedding and reranking. For more details, see [Vector
        embedding](https://docs.pinecone.io/guides/index-data/indexing-overview#vector-embedding)
        and [Rerank
        results](https://docs.pinecone.io/guides/search/rerank-results).
      operationId: get_model
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: model_name
          description: The name of the model to look up.
          required: true
          schema:
            type: string
          example: multilingual-e5-large
          style: simple
      responses:
        '200':
          description: The model details.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ModelInfo'
              examples:
                embedding-model:
                  summary: An embedding model.
                  value:
                    default_dimension: 256
                    max_batch_size: 96
                    max_sequence_length: 512
                    modality: text
                    model: example-embedding-model
                    provider_name: Embedding Model Provider
                    short_description: An example embedding model.
                    supported_dimensions:
                      - 256
                      - 512
                    supported_metrics:
                      - cosine
                      - euclidean
                    supported_parameters:
                      - allowed_values:
                          - value1
                          - value2
                        parameter: example_required_param
                        required: true
                        type: one_of
                        value_type: string
                      - allowed_values:
                          - value1
                          - value2
                        default: value1
                        parameter: example_param_with_default
                        required: false
                        type: one_of
                        value_type: string
                      - default: 5
                        max: 10
                        min: 0
                        parameter: example_numeric_range
                        required: false
                        type: numeric_range
                        value_type: integer
                    type: embed
                    vector_type: dense
                rerank-model:
                  summary: A reranking model.
                  value:
                    max_batch_size: 100
                    max_sequence_length: 1024
                    modality: text
                    model: example-reranking-model
                    provider_name: Reranking Model Provider
                    short_description: An example reranking model.
                    supported_parameters:
                      - default: true
                        parameter: example_any_value
                        required: false
                        type: any
                        value_type: boolean
                    type: rerank
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Model not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                model-not-found:
                  summary: Model not found
                  value:
                    error:
                      code: NOT_FOUND
                      message: Model example-model not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    ModelInfo:
      description: >-
        Represents the model configuration including model type, supported
        parameters, and other model details.
      type: object
      properties:
        model:
          example: multilingual-e5-large
          description: The name of the model.
          type: string
        short_description:
          example: multilingual-e5-large
          description: A summary of the model.
          type: string
        type:
          example: embed
          description: The type of model (e.g. 'embed' or 'rerank').
          type: string
        vector_type:
          description: Whether the embedding model produces 'dense' or 'sparse' embeddings.
          type: string
        default_dimension:
          example: 1024
          description: >-
            The default embedding model dimension (applies to dense embedding
            models only).
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        modality:
          example: text
          description: The modality of the model (e.g. 'text').
          type: string
        max_sequence_length:
          example: 512
          description: The maximum tokens per sequence supported by the model.
          type: integer
          format: int32
          minimum: 1
        max_batch_size:
          example: 96
          description: The maximum batch size (number of sequences) supported by the model.
          type: integer
          format: int32
          minimum: 1
        provider_name:
          example: NVIDIA
          description: The name of the provider of the model.
          type: string
        supported_dimensions:
          description: >-
            The list of supported dimensions for the model (applies to dense
            embedding models only).
          type: array
          items:
            example: 1024
            type: integer
            format: int32
            minimum: 1
            maximum: 20000
        supported_metrics:
          $ref: '#/components/schemas/ModelInfoSupportedMetrics'
        supported_parameters:
          description: List of parameters supported by the model.
          type: array
          items:
            $ref: '#/components/schemas/ModelInfoSupportedParameter'
      required:
        - model
        - short_description
        - type
        - supported_parameters
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, or `FORBIDDEN`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable error message describing the error.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    ModelInfoSupportedMetrics:
      description: The distance metrics supported by the model for similarity search.
      type: array
      items:
        $ref: '#/components/schemas/ModelInfoMetric'
    ModelInfoSupportedParameter:
      description: >-
        Describes a parameter supported by the model, including parameter value
        constraints.
      type: object
      properties:
        parameter:
          example: input_type
          description: The name of the parameter.
          type: string
        type:
          example: one_of
          description: >-
            The parameter type e.g. 'one_of', 'numeric_range', or 'any'.


            If the type is 'one_of', then 'allowed_values' will be set, and the
            value specified must be one of the allowed values. 'one_of' is only
            compatible with value_type 'string' or 'integer'.


            If 'numeric_range', then 'min' and 'max' will be set, then the value
            specified must adhere to the value_type and must fall within the
            `[min, max]` range (inclusive).


            If 'any' then any value is allowed, as long as it adheres to the
            value_type.
          type: string
        value_type:
          example: string
          description: >-
            The type of value the parameter accepts, e.g. 'string', 'integer',
            'float', or 'boolean'.
          type: string
        required:
          example: true
          description: Whether the parameter is required (true) or optional (false).
          type: boolean
        allowed_values:
          description: The allowed parameter values when the type is 'one_of'.
          type: array
          items:
            anyOf:
              - type: string
              - type: integer
        min:
          example: 1
          description: >-
            The minimum allowed value (inclusive) when the type is
            'numeric_range'.
          type: number
        max:
          example: 1
          description: >-
            The maximum allowed value (inclusive) when the type is
            'numeric_range'.
          type: number
        default:
          example: END
          description: The default value for the parameter when a parameter is optional.
          anyOf:
            - type: string
            - type: integer
              format: int32
            - type: number
              format: float
            - type: boolean
      required:
        - parameter
        - type
        - value_type
        - required
    ModelInfoMetric:
      description: >-
        A distance metric that the embedding model supports for similarity
        searches.

        Possible values: `cosine`, `euclidean`, or `dotproduct`.
      x-enum:
        - cosine
        - euclidean
        - dotproduct
      type: string
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/describe_restore_job.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Describe a restore job

> Get a description of a restore job.

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  JOB_ID="9857add2-99d4-4399-870e-aa7f15d8d326"

  curl "https://api.pinecone.io/restore-jobs/$JOB_ID" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H 'accept: application/json'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "restore_job_id": "9857add2-99d4-4399-870e-aa7f15d8d326",
    "backup_id": "94a63aeb-efae-4f7a-b059-75d32c27ca57",
    "target_index_name": "restored-index",
    "target_index_id": "0d8aed24-adf8-4b77-8e10-fd674309dc85",
    "status": "Completed",
    "created_at": "2025-04-25T18:14:05.227526Z",
    "completed_at": "2025-04-25T18:14:11.074618Z",
    "percent_complete": 100
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml get /restore-jobs/{job_id}
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /restore-jobs/{job_id}:
    get:
      tags:
        - Manage Indexes
      summary: Describe a restore job
      description: Get a description of a restore job.
      operationId: describe_restore_job
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: job_id
          description: The ID of the restore job to describe.
          required: true
          schema:
            type: string
          example: 670e8400-e29b-41d4-a716-446655440000
          style: simple
      responses:
        '200':
          description: Configuration information and deployment status of the restore job.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/RestoreJobModel'
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Restore job not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                restore-job-not-found:
                  summary: Restore job not found
                  value:
                    error:
                      code: NOT_FOUND
                      message: >-
                        Restore job 670e8400-e29b-41d4-a716-446655440002 not
                        found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    RestoreJobModel:
      description: The RestoreJobModel describes the status of a restore job.
      type: object
      properties:
        restore_job_id:
          example: 670e8400-e29b-41d4-a716-446655440001
          description: Unique identifier for the restore job
          type: string
        backup_id:
          example: 670e8400-e29b-41d4-a716-446655440000
          description: Backup used for the restore
          type: string
        target_index_name:
          example: sample-index
          description: Name of the index into which data is being restored
          type: string
        target_index_id:
          example: 670e8400-e29b-41d4-a716-446655440002
          description: ID of the index
          type: string
        status:
          example: Completed
          description: Status of the restore job
          type: string
        created_at:
          example: '2025-02-04T13:00:00.000Z'
          description: Timestamp when the restore job started
          type: string
          format: date-time
        completed_at:
          example: '2025-02-04T14:00:00.000Z'
          description: Timestamp when the restore job finished
          type: string
          format: date-time
        percent_complete:
          example: 42.2
          description: The progress made by the restore job out of 100
          type: number
          format: float
          minimum: 0
          maximum: 100
      required:
        - restore_job_id
        - backup_id
        - target_index_name
        - target_index_id
        - status
        - created_at
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/describeindexstats.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Get index stats

> Return statistics about the contents of an index, including the vector count per namespace, the number of dimensions, and the index fullness.

Serverless indexes scale automatically as needed, so index fullness is relevant only for pod-based indexes.

<RequestExample>
  ```shell curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="YOUR_INDEX_HOST"

  curl -X POST "https://$INDEX_HOST/describe_index_stats" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  // EXAMPLE RESPONSE 1: Serverless index (on-demand)
  {
    "namespaces": {
      "example-namespace": {
        "vectorCount": 10000
      }
    },
    "indexFullness": 0,
    "totalVectorCount": 10000,
    "dimension": 1024,
    "metric": "cosine",
    "vectorType": "dense"
  }

  // EXAMPLE RESPONSE 2: Serverless index (dedicated)
  {
    "namespaces": {
      "example-namespace": {
        "vectorCount": 10000
      }
    },
    "indexFullness": 0.000309539,
    "totalVectorCount": 10000,
    "dimension": 1536,
    "metric": "cosine",
    "vectorType": "dense"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml post /describe_index_stats
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /describe_index_stats:
    post:
      tags:
        - Vector Operations
      summary: Get index stats
      description: >-
        Return statistics about the contents of an index, including the vector
        count per namespace, the number of dimensions, and the index fullness.


        Serverless indexes scale automatically as needed, so index fullness is
        relevant only for pod-based indexes.
      operationId: describeIndexStats
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/DescribeIndexStatsRequest'
        required: true
      responses:
        '200':
          description: A successful response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/IndexDescription'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    DescribeIndexStatsRequest:
      description: The request for the `describe_index_stats` operation.
      type: object
      properties:
        filter:
          description: >-
            If this parameter is present, the operation only returns statistics
            for vectors that satisfy the filter. See [Understanding
            metadata](https://docs.pinecone.io/guides/index-data/indexing-overview#metadata).


            Serverless indexes do not support filtering `describe_index_stats`
            by metadata.
          type: object
    IndexDescription:
      example:
        dimension: 1024
        index_fullness: 0.4
        namespaces:
          '':
            vectorCount: 50000
          example-namespace-2:
            vectorCount: 30000
        totalVectorCount: 80000
      description: The response for the `describe_index_stats` operation.
      type: object
      properties:
        namespaces:
          description: >-
            A mapping for each namespace in the index from the namespace name to
            a summary of its contents. If a metadata filter expression is
            present, the summary will reflect only vectors matching that
            expression.
          type: object
          additionalProperties:
            $ref: '#/components/schemas/NamespaceSummary'
        dimension:
          example: 1024
          description: >-
            The dimension of the indexed vectors. Not specified if `sparse`
            index.
          type: integer
          format: int64
        indexFullness:
          example: 0.4
          description: >-
            The fullness of the index, regardless of whether a metadata filter
            expression was passed. The granularity of this metric is 10%.


            Serverless indexes scale automatically as needed, so index fullness 
            is relevant only for pod-based indexes.


            The index fullness result may be inaccurate during pod resizing; to
            get the status of a pod resizing process, use
            [`describe_index`](https://docs.pinecone.io/reference/api/2024-10/control-plane/describe_index).
          type: number
          format: float
        totalVectorCount:
          example: 80000
          description: >-
            The total number of vectors in the index, regardless of whether a
            metadata filter expression was passed
          type: integer
          format: int64
        metric:
          example: cosine
          description: The metric used to measure similarity.
          type: string
        vectorType:
          example: dense
          description: The type of vectors stored in the index.
          type: string
        memory_fullness:
          description: The amount of memory used by a dedicated index
          type: number
          format: float
        storage_fullness:
          description: The amount of storage used by a dedicated index
          type: number
          format: float
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    NamespaceSummary:
      description: A summary of the contents of a namespace.
      type: object
      properties:
        vectorCount:
          example: 50000
          description: >-
            The number of vectors stored in this namespace. Note that updates to
            this field may lag behind updates to the underlying index and
            corresponding query results, etc.
          type: integer
          format: int64
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/describenamespace.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Describe a namespace

> Describe a namespace in a serverless index, including the total number of vectors in the namespace.

For guidance and examples, see [Manage namespaces](https://docs.pinecone.io/guides/manage-data/manage-namespaces).

**Note:** This operation is not supported for pod-based indexes.

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="YOUR_INDEX_HOST"
  NAMESPACE="YOUR_NAMESPACE"  # To target the default namespace, use "__default__".

  curl "https://$INDEX_HOST/namespaces/$NAMESPACE" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "name": "example-namespace",
    "record_count": 20000
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml get /namespaces/{namespace}
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /namespaces/{namespace}:
    get:
      tags:
        - Namespace Operations
      summary: Describe a namespace
      description: >-
        Describe a namespace in a serverless index, including the total number
        of vectors in the namespace.


        For guidance and examples, see [Manage
        namespaces](https://docs.pinecone.io/guides/manage-data/manage-namespaces).


        **Note:** This operation is not supported for pod-based indexes.
      operationId: describeNamespace
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: namespace
          description: The namespace to describe.
          required: true
          schema:
            type: string
          style: simple
      responses:
        '200':
          description: A description of a namespace.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/NamespaceDescription'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    NamespaceDescription:
      description: A description of a namespace, including the name and record count.
      type: object
      properties:
        name:
          example: example-namespace
          description: The name of the namespace.
          type: string
        record_count:
          example: 20000
          description: The total amount of records within the namespace.
          type: integer
          format: int64
        schema:
          example:
            fields:
              description:
                filterable: true
              genre:
                filterable: true
              year:
                filterable: true
          description: >-
            Schema for the behavior of Pinecone's internal metadata index. By
            default, all metadata is indexed; when `schema` is present, only
            fields which are present in the `fields` object with a `filterable:
            true` are indexed. Note that `filterable: false` is not currently
            supported.
          type: object
          properties:
            fields:
              description: >-
                A map of metadata field names to their configuration. The field
                name must be a valid metadata field name. The field name must be
                unique.
              type: object
              additionalProperties:
                type: object
                properties:
                  filterable:
                    description: >-
                      Whether the field is filterable. If true, the field is
                      indexed and can be used in filters. Only true values are
                      allowed.
                    type: boolean
          required:
            - fields
        indexed_fields:
          description: A list of all indexed metatadata fields in the namespace
          type: object
          properties:
            fields:
              type: array
              items:
                type: string
              example:
                - genre
                - year
                - author
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/troubleshooting/differences-between-lexical-semantic-search.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Differences between Lexical and Semantic Search regarding relevancy

When it comes to searching for information in a large corpus of text, there are two main approaches that search engines use: keyword or lexical search and vector semantic similarity search. While both methods aim to retrieve relevant documents, they use different techniques to do so.

Keyword or lexical search relies on matching exact words or phrases that appear in a query with those in the documents. This approach is relatively simple and fast, but it has limitations. For example, it may not be able to handle misspellings, synonyms, or polysemy (when a word has multiple meanings). In addition, it does not take into account the context or meaning of the words, which can lead to irrelevant results.

On the other hand, vector semantic similarity search uses natural language processing (NLP) techniques to analyze the meaning of words and their relationships. It represents words as vectors in a high-dimensional space, where the distance between vectors indicates their semantic similarity. This approach can handle misspellings, synonyms, and polysemy, and it can also capture more subtle relationships between words, such as antonyms, hypernyms, and meronyms. As a result, it can produce more accurate and relevant results.

However, there is a caveat to using vector semantic similarity search. It requires a large amount of data to train the NLP models, which can be computationally expensive and time-consuming. As a result, it may not be as effective for short documents or queries that do not contain enough context to determine the meaning of the words. In such cases, a simple keyword or lexical search may be more suitable and effective.

In fact, in some cases, a short document may actually show higher in a vector space for a given query, even if it is not as relevant as a longer document. This is because short documents typically have fewer words, which means that their word vectors are more likely to be closer to the query vector in the high-dimensional space. As a result, they may have a higher cosine similarity score than longer documents, even if they do not contain as much information or context. This phenomenon is known as the "curse of dimensionality" and it can affect the performance of vector semantic similarity search in certain scenarios.

In conclusion, both keyword or lexical search and vector semantic similarity search have their strengths and weaknesses. Depending on the nature of the corpus, the type of queries, and the computational resources available, one approach may be more appropriate than the other. It is important to understand the differences between the two methods and use them judiciously to achieve the best results.

---

# Source: https://docs.pinecone.io/reference/dotnet-sdk.md

# .NET SDK

<Tip>
  See the [.NET SDK documentation](https://github.com/pinecone-io/pinecone-dotnet-client/blob/main/README.md) for full installation instructions and usage examples.

  To make a feature request or report an issue, please [file an issue](https://github.com/pinecone-io/pinecone-dotnet-client/issues).
</Tip>

## Requirements

To use this Python .NET SDK, ensure that your project is targeting one of the following:

* .NET Standard 2.0+
* .NET Core 3.0+
* .NET Framework 4.6.2+
* .NET 6.0+

## SDK versions

SDK versions are pinned to specific [API versions](/reference/api/versioning). When a new API version is released, a new version of the SDK is also released.

The mappings between API versions and .NET SDK versions are as follows:

| API version        | SDK version |
| :----------------- | :---------- |
| `2025-04` (latest) | v4.x        |
| `2025-01`          | v3.x        |
| `2024-10`          | v2.x        |
| `2024-07`          | v1.x        |
| `2024-04`          | v0.x        |

When a new stable API version is released, you should upgrade your SDK to the latest version to ensure compatibility with the latest API changes.

## Install

To add the latest version of the [.NET SDK](https://github.com/pinecone-io/pinecone-dotnet-client) to your project, run the following command:

<CodeGroup>
  ```shell .NET Core CLI theme={null}
  dotnet add package Pinecone.Client 
  ```

  ```shell NuGet CLI theme={null}
  nuget install Pinecone.Client
  ```
</CodeGroup>

To add a specific version of the [.NET SDK](https://github.com/pinecone-io/pinecone-dotnet-client) to your project, run the following command:

<CodeGroup>
  ```shell .NET Core CLI theme={null}
  dotnet add package Pinecone.Client --version <version>
  ```

  ```shell NuGet CLI theme={null}
  nuget install Pinecone.Client -Version <version>
  ```
</CodeGroup>

To check your SDK version, run the following command:

<CodeGroup>
  ```shell .NET Core CLI theme={null}
  dotnet list package
  ```

  ```shell NuGet CLI theme={null}
  nuget list 
  ```
</CodeGroup>

## Upgrade

<Warning>
  Before upgrading to `v4.0.0`, update all relevant code to account for the breaking changes explained [here](/release-notes/2025#2025-05-14-2).
</Warning>

If you are already using `Pinecone.Client` in your project, upgrade to the latest version as follows:

<CodeGroup>
  ```shell .NET Core CLI theme={null}
  dotnet add package Pinecone.Client 
  ```

  ```shell NuGet CLI theme={null}
  nuget install Pinecone.Client
  ```
</CodeGroup>

## Initialize

Once installed, you can import the SDK and then use an [API key](/guides/production/security-overview#api-keys) to initialize a client instance:

```csharp C# theme={null}
using Pinecone;

var pinecone = new PineconeClient("YOUR_API_KEY");
```

## Proxy configuration

If your network setup requires you to interact with Pinecone through a proxy, configure the HTTP client as follows:

```csharp  theme={null}
using System.Net;
using Pinecone;

var pinecone = new PineconeClient("PINECONE_API_KEY", new ClientOptions
{
    HttpClient = new HttpClient(new HttpClientHandler
    {
        Proxy = new WebProxy("PROXY_HOST:PROXY_PORT")
    })
});
```

If you're building your HTTP client using the [HTTP client factory](https://learn.microsoft.com/en-us/dotnet/core/extensions/httpclient-factory#configure-the-httpmessagehandler), use the `ConfigurePrimaryHttpMessageHandler` method to configure the proxy:

```csharp  theme={null}
   .ConfigurePrimaryHttpMessageHandler(() => new HttpClientHandler
       {
           Proxy = new WebProxy("PROXY_HOST:PROXY_PORT")
       });
```

---

# Source: https://docs.pinecone.io/guides/organizations/manage-billing/downgrade-billing-plan.md

# Source: https://docs.pinecone.io/guides/assistant/admin/downgrade-billing-plan.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Downgrade your plan

> Downgrade from a paid plan to the free Starter plan.

<Note>
  To change your billing plan, you must be an [organization owner or billing admin](/guides/organizations/understanding-organizations#organization-roles).
</Note>

## Requirements

Before you can downgrade, your organization must be under the [Starter plan quotas](/reference/api/database-limits):

* No more than 5 indexes, all serverless and in the `us-east-1` region of AWS
  * If you have serverless indexes in a region other than `us-east-1`, [create a new serverless index](/guides/index-data/create-an-index#create-a-serverless-index) in `us-east-1`, [re-upsert your data](/guides/index-data/upsert-data) into the new index, and [delete the old index](/guides/manage-data/manage-indexes#delete-an-index).
  * If you have more than 5 serverless indexes, [delete indexes](/guides/manage-data/manage-indexes#delete-an-index) until you have 5 or fewer.
  * If you have pod-based indexes, [delete them](/guides/manage-data/manage-indexes#delete-an-index).
* No more than 1 project
  * If you have more than 1 project, [delete all but 1 project](/guides/projects/manage-projects#delete-a-project).
  * Before you can delete a project, you must [delete all indexes](/guides/manage-data/manage-indexes#delete-an-index) and [delete all collections](/guides/manage-data/back-up-an-index#delete-a-collection) in the project.
* No more than 2 GB of data across all of your serverless indexes
  * If you are storing more than 2 GB of data, [delete records](/guides/manage-data/delete-data) until you're storing less than 2 GB.
* No more than 100 namespaces per serverless index
  * If any serverless index has more than 100 namespaces, [delete namespaces](/guides/manage-data/delete-data#delete-all-records-from-a-namespace) until it has 100 or fewer remaining.
* No more than 3 [assistants](/guides/assistant/overview)
  * If you have more than 3 assistants, [delete assistants](/guides/assistant/manage-assistants#delete-an-assistant) until you have 3 or fewer.
* No more than 10 files per assistant
  * If you have more than 10 files uploaded to an assistant, [delete files](/guides/assistant/manage-files#delete-a-file) until the assistant has 10 or fewer.
* No more than 1 GB of assistant storage
  * If you have more than 1 GB of assistant storage, [delete files](https://docs.pinecone.io/guides/assistant/manage-files#delete-a-file) until you're storing less than 1 GB.
* No more than 2 users
* No collections or backups (these are automatically deleted as part of the downgrade process)

## Downgrade to the Starter plan

The downgrade process is different depending on how you are paying for Pinecone.

<Warning>
  It is important to start the downgrade process in the Pinecone console, as described below. When you do so, Pinecone checks that you are under the [Starter plan quotas](#requirements) before allowing you to downgrade. In contrast, if you start the downgrade process in one of the cloud marketplaces, Pinecone cannot check that you are under these quotas before allowing you to downgrade. If you are over the quotas, Pinecone will deactivate your account, and you will need to [contact support](https://www.pinecone.io/contact/support/).
</Warning>

<Tabs>
  <Tab title="Credit card">
    If you are paying with a credit card, downgrade as follows:

    1. In the Pinecone console, go to [**Settings > Billing > Plans**](https://app.pinecone.io/organizations/-/settings/billing/plans).
    2. Click **Downgrade** in the **Starter** plan section.

    Your billing will end immediately. However, you will receive a final invoice for any charges accrued in the current month.
  </Tab>

  <Tab title="Google Cloud Marketplace">
    If you are paying through the Google Cloud Marketplace, downgrade as follows:

    1. In the Pinecone console, go to [**Settings > Billing > Plans**](https://app.pinecone.io/organizations/-/settings/billing/plans).
    2. In the **Starter** section, click **Downgrade**.
    3. Click **Confirm downgrade**.
    4. On the **Continue your downgrade on the GCP marketplace** modal, click **Continue to marketplace**. This takes you to your orders page in Google Cloud Marketplace.
    5. [Cancel the order](https://cloud.google.com/marketplace/docs/manage-billing#saas-products) for your Pinecone subscription.

       <Tip>
         If you don't see the order, check that the correct billing account is selected.
       </Tip>

    Your billing will end immediately. However, you will receive a final invoice for any charges accrued in the current month.
  </Tab>

  <Tab title="AWS Marketplace">
    If you are paying through the AWS Marketplace, downgrade as follows:

    1. In the Pinecone console, go to [**Settings > Billing > Plans**](https://app.pinecone.io/organizations/-/settings/billing/plans).
    2. In the **Starter** section, click **Downgrade**.
    3. Click **Confirm downgrade**.
    4. On the **Continue your downgrade on the AWS marketplace** modal, click **Continue to marketplace**. This takes you to the [Manage subscriptions](https://console.aws.amazon.com/marketplace) page in the AWS Marketplace.
    5. [Cancel the subscription](https://docs.aws.amazon.com/marketplace/latest/buyerguide/cancel-subscription.html#cancel-saas-subscription) to Pinecone.

    Your billing will end immediately. However, you will receive a final invoice for any charges accrued in the current month.
  </Tab>

  <Tab title="Microsoft Marketplace">
    If you are paying through the Microsoft Marketplace, downgrade as follows:

    1. In the Pinecone console, go to [**Settings > Billing > Plans**](https://app.pinecone.io/organizations/-/settings/billing/plans).
    2. In the **Starter** section, click **Downgrade**.
    3. Click **Confirm downgrade**.
    4. On the **Continue your downgrade on Microsoft marketplace** modal, click **Continue to marketplace**.
    5. On the **SaaS** page, click your subscription to Pinecone.
    6. Click **Cancel subscription**.
    7. Confirm the cancellation.

    Your billing will end immediately. However, you will receive a final invoice for any charges accrued in the current month.
  </Tab>
</Tabs>

---

# Source: https://docs.pinecone.io/guides/organizations/manage-billing/download-usage-report.md

# Source: https://docs.pinecone.io/guides/assistant/admin/download-usage-report.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Download a usage report

> Export organization usage and cost reports.

<Note>
  To view usage and costs across your Pinecone organization, you must be an [organization owner](/guides/organizations/understanding-organizations#organization-owners). Also, this feature is available only to organizations on the Standard or Enterprise plans.
</Note>

The **Usage** dashboard in the Pinecone console gives you a detailed report of usage and costs across your organization, broken down by each billable SKU or aggregated by project or service. You can view the report in the console or download it as a CSV file for more detailed analysis.

1. Go to [**Settings > Usage**](https://app.pinecone.io/organizations/-/settings/usage) in the Pinecone console.
2. Select the time range to report on. This defaults to the last 30 days.
3. Select the scope for your report:
   * **SKU:** The usage and cost for each billable SKU, for example, read units per cloud region, storage size per cloud region, or tokens per embedding model.
   * **Project:** The aggregated cost for each project in your organization.
   * **Service:** The aggregated cost for each service your organization uses, for example, database (includes serverless back up and restore), assistants, inference (embedding and reranking), and collections.
4. Choose the specific SKUs, projects, or services you want to report on. This defaults to all.
5. To download the report as a CSV file, click **Download**.

   <Tip>
     The CSV download provides more granular detail than the console view, including breakdowns by individual index as well as project and index tags.
   </Tip>

Dates are shown in UTC to match billing invoices. Cost data is delayed up to three days from the actual usage date.

---

# Source: https://docs.pinecone.io/troubleshooting/embedding-values-changed-when-upserted.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Embedding values changed when upserted

There are two distinct cases in which you might notice that the values of your embeddings appear different in Pinecone than the floats you upserted.

If you use a pod-based index with `p2` pods, we use quantization to enable the [Pinecone Graph Algorithm.](https://www.pinecone.io/blog/hnsw-not-enough/#:~:text=built%20vector%20databases.-,The%20Pinecone%20Graph%20Algorithm,-In%20order%20to) This is utilized only in `p2` indexes and powers faster query paths and greater QPS capacities.

For all serverless and other pod-based indexes, you may see slightly different values in Pinecone for high-precision floats. What’s happening here is a result of the fact that fractions can’t be accurately represented in fixed amounts of memory. Different numbers might be mapped to the same bit representation, according to a standard known as [IEEE 754](https://en.wikipedia.org/wiki/IEEE_754). More specifically, at Pinecone, we use Rust’s `f32` type, which is more commonly known as a `float` in Java, C, C++, and Python.

If you take these numbers and look at their physical memory representation, you’ll see that each float maps to the same representation before and after we upsert the vector to Pinecone. Our team built a [small demonstration in Rust](https://play.rust-lang.org/?version=stable\&mode=debug\&edition=2021\&gist=3584c20894714c5cba47127a036678fa) that you can use to explore some examples. We've also included a sample result in the attached screenshot.

This behavior is common across every system in the world, and the general trend in machine learning has been to reduce accuracy even more (see Google’s [bfloat16](https://en.wikipedia.org/wiki/Bfloat16_floating-point_format), which is now also standardized).

<img src="https://mintlify.s3.us-west-1.amazonaws.com/pinecone/troubleshooting/images/float_memory_representation.png" alt="A screenshot of the Rust demonstration, showing that two different floats have the same memory representation in bits." />

---

# Source: https://docs.pinecone.io/troubleshooting/error-cannot-import-name-pinecone.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Error: Cannot import name 'Pinecone' from 'pinecone'

## Problem

When using an older version of the [Python SDK](https://github.com/pinecone-io/pinecone-python-client) (earlier than 3.0.0), trying to import the `Pinecone` class raises the following error:

```console console theme={null}
ImportError: cannot import name 'Pinecone' from 'pinecone'
```

## Solution

Upgrade the SDK version and try again:

```Shell Shell theme={null}
# If you're interacting with Pinecone via HTTP requests, use:
pip install pinecone --upgrade
```

```Shell Shell theme={null}
# If you're interacting with Pinecone via gRPC, use:
pip install "pinecone[grpc]" --upgrade
```

---

# Source: https://docs.pinecone.io/guides/production/error-handling.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Error handling

> Handle errors with retry logic and best practices.

## Understand error types

Pinecone uses [conventional HTTP response codes](/reference/api/errors) to indicate the success or failure of API requests:

* **2xx codes** indicate success
* **4xx codes** indicate client errors (issues with your request)
* **5xx codes** indicate server errors (issues with Pinecone's servers)

### Client errors (4xx)

Client errors indicate problems with your request. These errors typically require changes to your code or configuration:

* **400 - Invalid Argument**: Your request contains invalid parameters. Check your request format and parameters.
* **401 - Unauthenticated**: Your API key is missing or invalid. Verify your [API key](/guides/projects/manage-api-keys).
* **402 - Payment Required**: Your account has a payment issue. Check your billing status in the [console](https://app.pinecone.io).
* **403 - Forbidden**: You've exceeded a [quota](/reference/api/database-limits#object-limits) or hit [deletion protection](/guides/manage-data/manage-indexes#configure-deletion-protection).
* **404 - Not Found**: The requested resource doesn't exist. Verify the resource name and that it hasn't been deleted.
* **409 - Already Exists**: You're trying to create a resource that already exists.
* **429 - Too Many Requests**: You're being [rate-limited](/reference/api/database-limits#rate-limits). Implement [backoff and retry logic](#implement-retry-logic).

### Server errors (5xx)

Server errors indicate temporary issues with Pinecone's infrastructure:

* **500 - Unknown**: An internal server error occurred.
* **502 - Bad Gateway**: The API gateway received an invalid response from a backend service.
* **503 - Unavailable**: The service is currently unavailable.
* **504 - Gateway Timeout**: The API gateway did not receive a timely response from the backend server. This can occur due to slow requests or backend processing delays.

**Best practice for 5xx errors**: [Implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic). These errors are typically transient.

## Capture errors

Each SDK provides error handling mechanisms specific to the language:

### Python SDK

The Python SDK raises exceptions that you can catch and handle:

```python  theme={null}
from pinecone import Pinecone
from pinecone.exceptions import PineconeException

pc = Pinecone(api_key="YOUR_API_KEY")
index = pc.Index("your-index")

try:
    index.upsert(
        vectors=[
            {"id": "vec1", "values": [0.1, 0.2, 0.3]}
        ]
    )
except PineconeException as e:
    # Handle Pinecone-specific errors
    print(f"Pinecone error: {e}")
except Exception as e:
    # Handle other errors
    print(f"Unexpected error: {e}")
```

See the [Python SDK documentation](https://sdk.pinecone.io/python/) for more details on exception handling.

### Node.js SDK

The Node.js SDK uses standard JavaScript error handling:

```javascript  theme={null}
const { Pinecone } = require('@pinecone-database/pinecone');

const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

try {
  const index = pc.index('your-index');
  await index.upsert([
    { id: 'vec1', values: [0.1, 0.2, 0.3] }
  ]);
} catch (error) {
  console.error('Error upserting data:', error);
  // Handle the error appropriately
}
```

See the [Node.js SDK documentation](https://sdk.pinecone.io/typescript/) for more information.

### Other SDKs

For SDK-specific error handling patterns, see the documentation for your language:

* [Go SDK](/reference/sdks/go/overview)
* [Java SDK](/reference/sdks/java/overview)
* [.NET SDK](/reference/sdks/dotnet/overview)

## Implement retry logic

For transient errors (5xx codes and 429 rate limiting), implement retry logic. Start with basic retries for simple use cases, or use exponential backoff for production systems.

### Basic retry logic

For simple use cases, start with a basic retry loop with fixed delays:

```python  theme={null}
import time
from pinecone.exceptions import PineconeException

def simple_retry(func, max_retries=3, delay=2):
    """
    Retry a function with a fixed delay between attempts.

    Args:
        func: Function to retry
        max_retries: Maximum number of retry attempts
        delay: Delay in seconds between retries
    """
    for attempt in range(max_retries):
        try:
            return func()
        except PineconeException as e:
            if attempt == max_retries - 1:
                raise  # Last attempt, re-raise the exception

            print(f"Attempt {attempt + 1} failed, retrying in {delay}s...")
            time.sleep(delay)

# Usage
try:
    simple_retry(lambda: index.upsert(vectors))
except Exception as e:
    print(f"Failed after {max_retries} attempts: {e}")
```

This basic approach works well for occasional transient errors, but for production systems with higher traffic, use exponential backoff instead.

### Exponential backoff

Exponential backoff progressively increases the wait time between retries to avoid overwhelming the service:

```python  theme={null}
import time
import random

def exponential_backoff_retry(func, max_retries=5, base_delay=1, max_delay=60):
    """
    Retry a function with exponential backoff.

    Args:
        func: Function to retry
        max_retries: Maximum number of retry attempts
        base_delay: Initial delay in seconds
        max_delay: Maximum delay between retries
    """
    for attempt in range(max_retries):
        try:
            return func()
        except PineconeException as e:
            if attempt == max_retries - 1:
                raise  # Last attempt, re-raise the exception

            # Get status code if available
            status_code = getattr(e, 'status', None)

            # Only retry on 5xx errors or 429 (rate limiting)
            if status_code and (status_code >= 500 or status_code == 429):
                # Calculate delay with exponential backoff and jitter
                delay = min(base_delay * (2 ** attempt), max_delay)
                jitter = random.uniform(0, delay * 0.1)  # Add 10% jitter
                wait_time = delay + jitter

                print(f"Retry attempt {attempt + 1}/{max_retries} after {wait_time:.2f}s")
                time.sleep(wait_time)
            else:
                # Don't retry client errors (4xx except 429)
                raise

# Usage
try:
    exponential_backoff_retry(lambda: index.upsert(vectors))
except Exception as e:
    print(f"Failed after retries: {e}")
```

### Key retry principles

1. **Add jitter**: Random variation in retry timing helps avoid thundering herd problems.
2. **Set max retries**: Prevent infinite retry loops.
3. **Cap delay time**: Don't wait indefinitely between retries.
4. **Don't retry client errors**: 4xx errors (except 429) won't resolve with retries.
5. **Log retry attempts**: Track retry behavior for monitoring and debugging.

## Handle rate limits (429)

When you receive a 429 error, you're being rate-limited. See [Rate limits](/reference/api/database-limits#rate-limits) for current limits.

Rate limits help protect your applications and maintain the health of the serverless infrastructure. **Most limits can be adjusted upon request**—if you need higher limits to scale, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket) with details about your use case.

<Note>
  Indexes built on [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) are not subject to read unit limits for query, fetch, and list operations. For sizing and capacity planning guidance, see the [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) guide.
</Note>

**Best practices**:

* Implement exponential backoff as described above.
* Proactively [monitor request metrics](/guides/production/monitoring) and reduce the request rate if you're approaching limits.
* Use [batching](/guides/index-data/upsert-data#upsert-in-batches) to reduce the number of requests.
* For high-throughput needs, see [Increase throughput](/guides/optimize/increase-throughput).

## Getting support

If you've implemented error handling and retry logic but continue to experience issues:

1. Review [How to work with Support](/troubleshooting/how-to-work-with-support) for best practices.
2. Gather the following information:
   * Index name and project name
   * Error messages and stack traces
   * Timestamp of errors
   * Request/response examples (without sensitive data)
   * Whether the issue is reproducible
3. [Contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

Ensure your [plan tier](https://www.pinecone.io/pricing/) provides the support SLA you need for production workloads.

## See also

* [API error codes](/reference/api/errors)
* [Database limits](/reference/api/database-limits)
* [Assistant limits](/reference/api/assistant/assistant-limits)
* [Monitoring](/guides/production/monitoring)
* [Production checklist](/guides/production/production-checklist)

---

# Source: https://docs.pinecone.io/troubleshooting/error-handshake-read-failed.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Error: Handshake read failed when connecting

## Problem

When trying to connect to Pinecone server, some users may receive an error message that says `Handshake read failed` and their connection attempt fails. This error can prevent them from running queries against their Pinecone indexes.

## Solution

If you encounter this error message, it means that your computer is not properly connecting with the Pinecone server. The error is often due to a misconfiguration of your Pinecone client or API key. Here is a recommended solution:

1. Make sure your firewall is not blocking any traffic and your internet connection is working fine. If you are unsure about how to do this, please consult your IT team.
2. Check that you have set up the Pinecone client and API key correctly. Double-check that you have followed the instructions in our [documentation](/guides/get-started/quickstart) correctly.
3. If you are still having issues, try creating a new index on Pinecone and populating it with data by running another script on your computer. This will verify that your computer can access the Pinecone servers for some tasks.
4. If the error persists, you may need to check your code for any misconfigurations. Make sure you are setting up your Pinecone client correctly and passing the right parameters when running queries against your indexes.
5. If you are still unable to resolve the issue, you can reach out to Pinecone support for assistance. They will be able to help you diagnose and resolve the issue.

## Conclusion

If you encounter the `Handshake read failed` error when trying to connect to Pinecone server, there are several steps you can take to resolve the issue. First, double-check that you have set up the Pinecone client and API key correctly. Then, check for any misconfigurations in your code. If the error persists, [contact Pinecone Support](/troubleshooting/contact-support) for assistance.

---

# Source: https://docs.pinecone.io/reference/api/errors.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Errors

Pinecone uses conventional HTTP response codes to indicate the success or failure of an API request. In general, codes in the `2xx` range indicate success, codes in the `4xx` range indicate an error that failed given the information provided, and codes in the `5xx` range indicate an error with Pinecone's servers.

For guidance on handling errors in production, see [Error handling](/guides/production/error-handling).

## 200 - OK

The request succeeded.

## 201 - CREATED

The request succeeded and a new resource was created.

## 202 - NO CONTENT

The request succeeded, but there is no content to return.

## 400 - INVALID ARGUMENT

The request failed due to an invalid argument.

## 401 - UNAUTHENTICATED

The request failed due to a missing or invalid [API key](/guides/projects/understanding-projects#api-keys).

## 402 - PAYMENT REQUIRED

The request failed due to delinquent payment.

## 403 - FORBIDDEN

The request failed due to an exceeded [quota](/reference/api/database-limits#object-limits) or [index deletion protection](/guides/manage-data/manage-indexes#configure-deletion-protection).

## 404 - NOT FOUND

The request failed because the resource was not found.

## 409 - ALREADY EXISTS

The request failed because the resource already exists.

## 412 - FAILED PRECONDITIONS

The request failed due to preconditions not being met. |

## 422 - UNPROCESSABLE ENTITY

The request failed because the server was unable to process the contained instructions.

## 429 - TOO MANY REQUESTS

The request was [rate-limited](/reference/api/database-limits#rate-limits). [Implement retry logic with exponential backoff](/guides/production/error-handling#handle-rate-limits-429) to handle this error.

## 500 - UNKNOWN

An internal server error occurred. [Implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic) to handle transient errors.

## 502 - BAD GATEWAY

The API gateway received an invalid response from a backend service. This is typically a temporary error. [Implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic) to handle transient errors.

## 503 - UNAVAILABLE

The server is currently unavailable. [Implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic) to handle transient errors.

## 504 - GATEWAY TIMEOUT

The API gateway did not receive a timely response from the backend server. This can occur due to slow requests or backend processing delays. [Implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic) to handle transient errors.

---

# Source: https://docs.pinecone.io/integrations/estuary.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Estuary

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

[Estuary](https://estuary.dev/) builds real-time data pipelines that focus on moving data from sources to destinations with millisecond latency. It supports integrations with hundreds of systems including databases, warehouses, SaaS products and streaming solutions.

The Pinecone connector for Estuary enables users to source from these systems and push data to Pinecone, for an always up-to-date view. It incrementally updates source data to ensure that minimal credits are used when reaching out to get embeddings from providers prior to pushing them to Pinecone.

Estuary's Pinecone connector enables a variety of use cases like enabling LLM-based search across your organizations data and building intelligent recommendation systems. It can be set up in 5 minutes, without any engineering effort to maximize efficiency.

<PrimarySecondaryCTA primaryHref={"https://docs.estuary.dev/reference/Connectors/materialization-connectors/pinecone/"} primaryLabel={"Get started"} />

---

# Source: https://docs.pinecone.io/guides/assistant/evaluate-answers.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Evaluate answers

> Measure assistant response quality with LLM-based evaluation.

This page shows you how to [evaluate responses](/guides/assistant/evaluation-overview) from an assistant or other RAG systems using the `metrics_alignment` operation.

You can [evaluate a response](/reference/api/latest/assistant/metrics_alignment) from an assistant, as in the following example:

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant
  # pip install requests

  import requests
  from pinecone_plugins.assistant.models.chat import Message

  payload = {
      "question": "What are the capital cities of France, England and Spain?", # Question to ask the assistant.
      "answer": "Paris is the capital city of France and Barcelona of Spain", # Answer from the assistant.
      "ground_truth_answer": "Paris is the capital city of France, London of England and Madrid of Spain." # Expected answer to evaluate the assistant's response.
  }

  headers = {
      "Api-Key": "YOUR_API_KEY",
      "Content-Type": "application/json"
  }

  url = "https://prod-1-data.ke.pinecone.io/assistant/evaluation/metrics/alignment"

  response = requests.request("POST", url, json=payload, headers=headers)

  print(response.text)
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://prod-1-data.ke.pinecone.io/assistant/evaluation/metrics/alignment" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
    "question": "What are the capital cities of France, England and Spain?",
    "answer": "Paris is the capital city of France and Barcelona of Spain",
    "ground_truth_answer": "Paris is the capital city of France, London of England and Madrid of Spain"
  }'
  ```
</CodeGroup>

```json Response theme={null}
{
  "metrics": {
    "correctness": 0.5,
    "completeness": 0.3333,
    "alignment": 0.4
  },
  "reasoning": {
    "evaluated_facts": [
      {
        "fact": {
          "content": "Paris is the capital city of France."
        },
        "entailment": "entailed"
      },
      {
        "fact": {
          "content": "London is the capital city of England."
        },
        "entailment": "neutral"
      },
      {
        "fact": {
          "content": "Madrid is the capital city of Spain."
        },
        "entailment": "contradicted"
      }
    ]
  },
  "usage": {
    "prompt_tokens": 1223,
    "completion_tokens": 51,
    "total_tokens": 1274
  }
}
```

---

# Source: https://docs.pinecone.io/guides/assistant/evaluation-overview.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Evaluation overview

> Learn about evaluating the correctness and completeness of assistant responses.

You can [evaluate the correctness and completeness of a response](/guides/assistant/evaluate-answers) from an assistant or RAG system.

## Use cases

Response evaluation is useful when performing tasks like the following:

* Understanding how well the Pinecone Assistant captures the facts of the ground truth answer.
* Comparing the Pinecone Assistant's answers to those of another RAG system.
* Comparing the answers of your own RAG system to those of the Pinecone Assistant or another RAG system.

## SDK support

You can [evaluate responses](/reference/api/latest/assistant/metrics_alignment) directly or through the [Pinecone Python SDK](/reference/sdks/python/overview).

## Request

The request body requires the following fields:

| Field                 | Description                                           |
| --------------------- | ----------------------------------------------------- |
| `question`            | The question asked to the RAG system.                 |
| `answer`              | The answer provided by the assistant being evaluated. |
| `ground_truth_answer` | The expected answer.                                  |

For example:

```json  theme={null}
{
  "question": "What are the capital cities of France, England and Spain?",
  "answer": "Paris is the capital city of France and Barcelona of Spain",
  "ground_truth_answer": "Paris is the capital city of France, London of England and Madrid of Spain."
}
```

## Response

### Metrics

Calculated scores between `0` to `1` are returned for the following metrics:

| Metric         | Description                                                                  |
| -------------- | ---------------------------------------------------------------------------- |
| `correctness`  | Correctness of the RAG system's answer compared to the ground truth answer.  |
| `completeness` | Completeness of the RAG system's answer compared to the ground truth answer. |
| `alignment`    | A combined score of the correctness and completeness scores.                 |

```json  theme={null}
{
  "metrics": {
    "correctness": 0.5,
    "completeness": 0.333,
    "alignment": 0.398,
  }
},

...
```

### Reasoning

The response includes explanations for the reasoning behind each metric's score. This includes a list of evaluated facts with their entailment status:

| Status         | Description                                                                |
| -------------- | -------------------------------------------------------------------------- |
| `entailed`     | The fact is supported by the ground truth answer.                          |
| `contradicted` | The fact contradicts the ground truth answer.                              |
| `neutral`      | The fact is neither supported nor contradicted by the ground truth answer. |

```json  theme={null}
...

  "reasoning":{
    "evaluated_facts": [
      {
        "fact": {"content": "Paris is the capital of France"},
        "entailment": "entailed",
      },
      {
        "fact": {"content": "London is the capital of England"},
        "entailment": "neutral"
      },
      {
        "fact": {"content": "Madrid is the capital of Spain"},
        "entailment": "contradicted",
      }
    ]
  },

...
```

### Usage

The response includes the number of tokens used to calculate the metrics. This includes the number of tokens used for the prompt and completion.

```json  theme={null}
...

  "usage": {
    "prompt_tokens": 22,
    "completion_tokens": 33,
    "total_tokens": 55
  }
}
```

## Pricing

Cost is calculated by [token usage](#usage). See [Pricing](https://www.pinecone.io/pricing/) for up-to-date pricing information.

Response evaluation is only available for [Standard and Enterprise plans](https://www.pinecone.io/pricing/).

---

# Source: https://docs.pinecone.io/troubleshooting/export-indexes.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Export indexes

Pinecone does not support an export function. It is on our roadmap for the future, however.

In the meantime, we recommend keeping a copy of your source data in case you need to move from one project to another, in which case you'll need to reindex the data.

For backup purposes, we recommend that you take periodic backups. Please see [Back up indexes](/guides/manage-data/back-up-an-index) in our documentation for more details on doing so.

---

# Source: https://docs.pinecone.io/release-notes/feature-availability.md

# Source: https://docs.pinecone.io/assistant-release-notes/feature-availability.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Feature availability

This page defines the different availability phases of a feature in Pinecone.

The availability phases are used to communicate the maturity and stability of a feature. The availability phases are:

* **Early access**: In active development and may change at any time. Intended for user feedback only. In some cases, users must be granted explicit access to the API by Pinecone.

* **Public preview**: Unlikely to change between public preview and general availability. Not recommended for production usage. Available to all users.

* **Limited availability**: Available to select customers in a subset of regions and providers for production usage.

* **General availability**: Will not change on short notice. Recommended for production usage. Officially [supported by Pinecone](/troubleshooting/pinecone-support-slas) for non-production and production usage.

* **Deprecated**: Still supported, but no longer under active development, except for critical security fixes. Existing usage will continue to function, but migration following the upgrade guide is strongly recommended. Will be removed in the future at an announced date.

* **End of life (EOL)**: Removed, and no longer supported or available.

<Note>
  A feature is in **general availability** unless explicitly marked otherwise.
</Note>

---

# Source: https://docs.pinecone.io/guides/manage-data/fetch-data.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Fetch records

> Retrieve complete records by ID or metadata filter.

<Tip>
  You can fetch data using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/indexes/-/browser).
</Tip>

## Fetch records by ID

To fetch records from a namespace based on their IDs, use the `fetch` operation with the following parameters:

* `namespace`: The [namespace](/guides/index-data/indexing-overview#namespaces) containing the records to fetch. To use the default namespace, set this to `"__default__"`.
* `ids`: The IDs of the records to fetch. Maximum of 1000.

<Note>
  For on-demand indexes, since vector values are retrieved from object storage, fetch operations may have increased latency. If you only need metadata or IDs, consider using the [`query`](/reference/api/latest/data-plane/query) operation with `include_values` set to `false` instead. See [Decrease latency](/guides/optimize/decrease-latency#avoid-including-vector-values-when-not-needed) for more details.
</Note>

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  index.fetch(ids=["id-1", "id-2"], namespace="example-namespace")
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  const fetchResult = await index.namespace('example-namespace').fetch(['id-1', 'id-2']);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import io.pinecone.proto.FetchResponse;

  import java.util.Arrays;
  import java.util.List;

  public class FetchExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");

          List<String> ids = Arrays.asList("id-1", "id-2");
          FetchResponse fetchResponse = index.fetch(ids, "example-namespace");
          System.out.println(fetchResponse);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      res, err := idxConnection.FetchVectors(ctx, []string{"id-1", "id-2"})
      if err != nil {
          log.Fatalf("Failed to fetch vectors: %v", err)
      } else {
          fmt.Printf(prettifyStruct(res))
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var fetchResponse = await index.FetchAsync(new FetchRequest {
      Ids = new List<string> { "id-1", "id-2" },
      Namespace = "example-namespace"
  });
  ```

  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X GET "https://$INDEX_HOST/vectors/fetch?ids=id-1&ids=id-2&namespace=example-namespace" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

The response looks like this:

<CodeGroup>
  ```Python Python theme={null}
  {'namespace': 'example-namespace',
   'usage': {'readUnits': 1},
   'vectors': {'id-1': {'id': 'id-1',
                        'values': [0.568879, 0.632687092, 0.856837332, ...]},
               'id-2': {'id': 'id-2',
                        'values': [0.00891787093, 0.581895, 0.315718859, ...]}}}
  ```

  ```JavaScript JavaScript theme={null}
  {'namespace': 'example-namespace',
   'usage': {'readUnits': 1},
   'records': {'id-1': {'id': 'id-1',
                        'values': [0.568879, 0.632687092, 0.856837332, ...]},
               'id-2': {'id': 'id-2',
                        'values': [0.00891787093, 0.581895, 0.315718859, ...]}}}
  ```

  ```java Java theme={null}
  namespace: "example-namespace"
  vectors {
    key: "id-1"
    value {
      id: "id-1"
      values: 0.568879
      values: 0.632687092
      values: 0.856837332
      ...
    }
  }
  vectors {
    key: "id-2"
    value {
      id: "id-2"
      values: 0.00891787093
      values: 0.581895
      values: 0.315718859
      ...
    }
  }
  usage {
    read_units: 1
  }
  ```

  ```go Go theme={null}
  {
    "vectors": {
      "id-1": {
        "id": "id-1",
        "values": [
          -0.0089730695,
          -0.020010853,
          -0.0042787646,
          ...
        ]
      },
      "id-2": {
        "id": "id-2",
        "values": [
          -0.005380766,
          0.00215196,
          -0.014833462,
          ...
        ]
      }
    },
    "usage": {
      "read_units": 1
    }
  }
  ```

  ```csharp C# theme={null}
  {
    "vectors": {
      "id-1": {
        "id": "id-1",
        "values": [
          -0.0089730695,
          -0.020010853,
          -0.0042787646,
          ...
        ],
        "sparseValues": null,
        "metadata": null
      },
      "vec1": {
        "id": "id-2",
        "values": [
          -0.005380766,
          0.00215196,
          -0.014833462,
          ...
        ],
        "sparseValues": null,
        "metadata": null
      }
    },
    "namespace": "example-namespace",
    "usage": {
      "readUnits": 1
    }
  ```

  ```json curl theme={null}
  {
    "vectors": {
      "id-1": {
        "id": "id-1",
        "values": [0.568879, 0.632687092, 0.856837332, ...]
      },
      "id-2": {
        "id": "id-2",
        "values": [0.00891787093, 0.581895, 0.315718859, ...]
      }
    },
    "namespace": "example-namespace",
    "usage": {"readUnits": 1},
  }
  ```
</CodeGroup>

## Fetch records by metadata

<Warning>
  This feature is in [public preview](/release-notes/feature-availability).
</Warning>

To fetch records from a namespace based on their metadata values, use the `fetch_by_metadata` operation with the following parameters:

| Parameter         | Required | Description                                                                                                                                                                                                                                                                                                                                         |
| :---------------- | :------- | :-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `filter`          | Yes      | A [metadata filter expression](/guides/index-data/indexing-overview#metadata-filter-expressions) describing the records to fetch. Must be present and non-empty.                                                                                                                                                                                    |
| `limit`           | No       | The number of matching records to return. Defaults to 100 if not specified. Maximum of 10,000.                                                                                                                                                                                                                                                      |
| `namespace`       | No       | The [namespace](/guides/index-data/indexing-overview#namespaces) containing the records to fetch. If omitted or set to an empty string, defaults to the default namespace. To explicitly use the default namespace, set this to `"__default__"`.                                                                                                    |
| `paginationToken` | No       | The `next` token value from the `pagination` object found in a previous response. Include this value to fetch the next page of results, or omit it to start from the beginning. Must be used with the same `namespace` and `filter` parameters that generated it — using an existing token with different parameters will return incorrect results. |

For example, the following code fetches 2 records with a `genre` field set to `Action/Adventure` from the default namespace:

```shell curl theme={null}
# To get the unique host for an index,
# see https://docs.pinecone.io/guides/manage-data/target-an-index
PINECONE_API_KEY="YOUR_API_KEY"
INDEX_HOST="INDEX_HOST"

curl -X POST "https://$INDEX_HOST/vectors/fetch_by_metadata" \
  -H "Api-Key: $PINECONE_API_KEY" \
  -H "Content-Type: application/json" \
  -H "X-Pinecone-Api-Version: 2025-10" \
  -d '{
    "namespace": "__default__",
    "filter": {"genre": {"$eq": "Action/Adventure"}},
    "limit": 2
  }'
```

The response looks like this:

```json curl theme={null}
{
  "vectors": {
    "0": {
      "id": "0",
      "values": [
        0.0234527588, 0.0291595459 ...
      ],
      "metadata": {
        "box-office": 2923706026,
        "genre": "Action/Adventure",
        "summary": "On the alien world of Pandora, paraplegic Marine Jake Sully uses an avatar to walk again and becomes torn between his mission and protecting the planet's indigenous Na'vi people. The film stars Sam Worthington, Zoe Saldana, and Sigourney Weaver.",
        "title": "Avatar",
        "year": 2009
      }
    },
    "1": {
      "id": "1",
      "values": [
        0.0397644043, 0.013053894, ...
      ],
      "metadata": {
        "box-office": 2799439100,
        "genre": "Action/Adventure",
        "summary": "In the aftermath of Thanos wiping out half of the universe, the remaining Avengers assemble once more to undo the chaos, leading to a time-traveling adventure. Stars Robert Downey Jr., Chris Evans, and Scarlett Johansson.",
        "title": "Avengers: Endgame",
        "year": 2019
      }
    }
  },
  "namespace": "__default__",
  "usage": {
    "readUnits": 1
  },
  "pagination": {
    "next": "Tm90aGluZyB0byBzZWUgaGVyZQo="
  }
}
```

To fetch the next page of results, set `paginationToken` to the value of `next`. For example:

```shell curl theme={null}
curl -X POST "https://$INDEX_HOST/vectors/fetch_by_metadata" \
  -H "Api-Key: $PINECONE_API_KEY" \
  -H "Content-Type: application/json" \
  -H "X-Pinecone-Api-Version: 2025-10" \
  -d '{
    "namespace": "__default__",
    "filter": {"genre": {"$eq": "Action/Adventure"}},
    "limit": 2,
    "paginationToken": "Tm90aGluZyB0byBzZWUgaGVyZQo="
  }'
```

When there are more results available, the response includes a `pagination` object with a `next` token. When there are no more results, the response does not include a `pagination` object.

## Fetch limits

**Fetch by ID limits:**

| Metric              | Limit                             |
| :------------------ | :-------------------------------- |
| Max IDs per request | 1000 IDs                          |
| Max request size    | N/A                               |
| Max request rate    | 100 requests per second per index |

**Fetch by metadata limits:**

| Metric                   | Limit                                |
| :----------------------- | :----------------------------------- |
| Max records per response | 10,000 records                       |
| Max response size        | 4MB                                  |
| Max response rate        | 10 requests per second per namespace |

## Data freshness

Pinecone is eventually consistent, so there can be a slight delay before new or changed records are visible to queries. You can view index stats to [check data freshness](/guides/index-data/check-data-freshness).

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/fetch.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Fetch vectors

> Look up and return vectors by ID from a single namespace. The returned vectors include the vector data and/or metadata.

For on-demand indexes, since vector values are retrieved from object storage, fetch operations may have increased latency. If you only need metadata or IDs, consider using the query operation with `includeValues` set to `false` instead.

For guidance and examples, see [Fetch data](https://docs.pinecone.io/guides/manage-data/fetch-data).

<RequestExample>
  ```shell curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X GET "https://$INDEX_HOST/vectors/fetch?ids=id-1&ids=id-2&namespace=example-namespace" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "vectors": {
      "id-1": {
        "id": "id-1",
        "values": [0.568879, 0.632687092, 0.856837332, ...]
      },
      "id-2": {
        "id": "id-2",
        "values": [0.00891787093, 0.581895, 0.315718859, ...]
      }
    },
    "namespace": "example-namespace",
    "usage": {"readUnits": 1},
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml get /vectors/fetch
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /vectors/fetch:
    get:
      tags:
        - Vector Operations
      summary: Fetch vectors
      description: >-
        Look up and return vectors by ID from a single namespace. The returned
        vectors include the vector data and/or metadata.


        For on-demand indexes, since vector values are retrieved from object
        storage, fetch operations may have increased latency. If you only need
        metadata or IDs, consider using the query operation with `includeValues`
        set to `false` instead.


        For guidance and examples, see [Fetch
        data](https://docs.pinecone.io/guides/manage-data/fetch-data).
      operationId: fetchVectors
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: query
          name: ids
          description: The vector IDs to fetch. Does not accept values containing spaces.
          required: true
          schema:
            type: array
            items:
              type: string
          explode: true
          style: form
        - in: query
          name: namespace
          description: >-
            The namespace to fetch vectors from. If not provided, the default
            namespace is used.
          schema:
            type: string
          style: form
      responses:
        '200':
          description: A successful response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/FetchResponse'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    FetchResponse:
      example:
        namespace: example-namespace
        usage:
          readUnits: 1
        vectors:
          id-1:
            id: id-1
            values:
              - 1
              - 1.5
          id-2:
            id: id-2
            values:
              - 2
              - 1
      description: The response for the `fetch` operation.
      type: object
      properties:
        vectors:
          title: >-
            The fetched vectors, in the form of a map between the fetched ids
            and the fetched vectors
          type: object
          additionalProperties:
            $ref: '#/components/schemas/Vector'
        namespace:
          example: example-namespace
          description: The namespace of the vectors.
          type: string
        usage:
          $ref: '#/components/schemas/Usage'
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    Vector:
      type: object
      properties:
        id:
          example: example-vector-1
          description: This is the vector's unique id.
          type: string
          required:
            - id
          minLength: 1
          maxLength: 512
        values:
          example:
            - 0.1
            - 0.2
            - 0.3
            - 0.4
            - 0.5
            - 0.6
            - 0.7
            - 0.8
          description: This is the vector data included in the request.
          type: array
          required:
            - values
          items:
            type: number
            format: float
          minLength: 1
          maxLength: 20000
        sparseValues:
          $ref: '#/components/schemas/SparseValues'
        metadata:
          example:
            genre: documentary
            year: 2019
          description: This is the metadata included in the request.
          type: object
      required:
        - id
    Usage:
      type: object
      properties:
        readUnits:
          example: 5
          description: The number of read units consumed by this operation.
          type: integer
          format: int64
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
    SparseValues:
      description: >-
        Vector sparse data. Represented as a list of indices and a list of 
        corresponded values, which must be with the same length.
      type: object
      properties:
        indices:
          example:
            - 1
            - 312
            - 822
            - 14
            - 980
          description: The indices of the sparse data.
          type: array
          required:
            - indices
          items:
            type: integer
            format: int64
          minLength: 1
          maxLength: 1000
        values:
          example:
            - 0.1
            - 0.2
            - 0.3
            - 0.4
            - 0.5
          description: >-
            The corresponding values of the sparse data, which must be with the
            same length as the indices.
          type: array
          required:
            - values
          items:
            type: number
            format: float
          minLength: 1
          maxLength: 1000
      required:
        - indices
        - values
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/admin/fetch_api_key.md

# Source: https://docs.pinecone.io/reference/api/2025-10/admin-assistant/fetch_api_key.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Get API key details

> Get the details of an API key, excluding the API key secret.

<RequestExample>
  ```bash curl theme={null}
  PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"
  PINECONE_API_KEY_ID="3fa85f64-5717-4562-b3fc-2c963f66afa6"

  curl -X GET "https://api.pinecone.io/admin/api-keys/$PINECONE_API_KEY_ID" \
  	-H "Authorization: Bearer $PINECONE_ACCESS_TOKEN" \
      -H "accept: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
    "name": "string",
    "project_id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
    "roles": [
      "ProjectEditor"
    ]
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/admin_2025-10.oas.yaml get /admin/api-keys/{api_key_id}
openapi: 3.0.3
info:
  title: Pinecone Admin API
  description: |
    Provides an API for managing a Pinecone organization and its resources.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - BearerAuth: []
tags:
  - name: API Keys
    description: Actions that manage API Keys.
  - name: Organizations
    description: Actions that manage organizations.
  - name: Projects
    description: Actions that manage projects.
paths:
  /admin/api-keys/{api_key_id}:
    get:
      tags:
        - API Keys
      summary: Get API key details
      description: Get the details of an API key, excluding the API key secret.
      operationId: fetch_api_key
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: api_key_id
          description: API key ID
          required: true
          schema:
            type: string
            format: uuid
          style: simple
      responses:
        '200':
          description: The details of the API key, excluding the API key secret.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/APIKey'
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Not found
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
        4XX:
          description: Unexpected error on request.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
components:
  schemas:
    APIKey:
      description: The details of an API key, without the secret.
      type: object
      properties:
        id:
          description: The unique ID of the API key.
          type: string
          format: uuid
        name:
          description: The name of the API key.
          type: string
        project_id:
          description: The ID of the project containing the API key.
          type: string
          format: uuid
        roles:
          description: The roles assigned to the API key.
          type: array
          items:
            example: ProjectEditor
            description: >-
              A role that can be assigned to an API key.

              Possible values: `ProjectEditor`, `ProjectViewer`,
              `ControlPlaneEditor`, `ControlPlaneViewer`, `DataPlaneEditor`, or
              `DataPlaneViewer`.
            x-enum:
              - ProjectEditor
              - ProjectViewer
              - ControlPlaneEditor
              - ControlPlaneViewer
              - DataPlaneEditor
              - DataPlaneViewer
            type: string
      required:
        - id
        - name
        - project_id
        - roles
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, or `UNPROCESSABLE_ENTITY`.        
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    BearerAuth:
      type: http
      scheme: bearer
      description: >
        An [access
        token](https://docs.pinecone.io/guides/organizations/manage-service-accounts#retrieve-an-access-token)
        must be provided in the `Authorization` header using the `Bearer`
        scheme.

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/fetch_by_metadata.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Fetch vectors by metadata

> Look up and return vectors by metadata filter from a single namespace. The returned vectors include the vector data and/or metadata.
For guidance and examples, see [Fetch data](https://docs.pinecone.io/guides/manage-data/fetch-data).

<RequestExample>
  ```shell curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X POST "https://$INDEX_HOST/vectors/fetch_by_metadata" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "namespace": "__default__",
      "filter": {"genre": {"$eq": "Action/Adventure"}},
      "limit": 2
    }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "vectors": {
      "0": {
        "id": "0",
        "values": [
          0.0234527588, 0.0291595459 ...
        ],
        "metadata": {
          "box-office": 2923706026,
          "genre": "Action/Adventure",
          "summary": "On the alien world of Pandora, paraplegic Marine Jake Sully uses an avatar to walk again and becomes torn between his mission and protecting the planet's indigenous Na'vi people. The film stars Sam Worthington, Zoe Saldana, and Sigourney Weaver.",
          "title": "Avatar",
          "year": 2009
        }
      },
      "1": {
        "id": "1",
        "values": [
          0.0397644043, 0.013053894, ...
        ],
        "metadata": {
          "box-office": 2799439100,
          "genre": "Action/Adventure",
          "summary": "In the aftermath of Thanos wiping out half of the universe, the remaining Avengers assemble once more to undo the chaos, leading to a time-traveling adventure. Stars Robert Downey Jr., Chris Evans, and Scarlett Johansson.",
          "title": "Avengers: Endgame",
          "year": 2019
        }
      }
    },
    "namespace": "__default__",
    "usage": {
      "readUnits": 1
    },
    "pagination": {
      "next": "Tm90aGluZyB0byBzZWUgaGVyZQo="
    }
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml post /vectors/fetch_by_metadata
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /vectors/fetch_by_metadata:
    post:
      tags:
        - Vector Operations
      summary: Fetch vectors by metadata
      description: >-
        Look up and return vectors by metadata filter from a single namespace.
        The returned vectors include the vector data and/or metadata.

        For guidance and examples, see [Fetch
        data](https://docs.pinecone.io/guides/manage-data/fetch-data).
      operationId: fetch_vectors_by_metadata
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/FetchByMetadataRequest'
        required: true
      responses:
        '200':
          description: A successful response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/FetchByMetadataResponse'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    FetchByMetadataRequest:
      description: The request for the `fetch_by_metadata` operation.
      type: object
      properties:
        namespace:
          example: example-namespace
          description: The namespace to fetch vectors from.
          type: string
        filter:
          example:
            genre:
              $in:
                - comedy
                - documentary
                - drama
            year:
              $eq: 2019
          description: >-
            Metadata filter expression to select vectors. See [Understanding
            metadata](https://docs.pinecone.io/guides/index-data/indexing-overview#metadata).
          type: object
        limit:
          example: 12
          description: Max number of vectors to return.
          default: 100
          type: integer
          format: int64
          minimum: 1
        paginationToken:
          example: Tm90aGluZyB0byBzZWUgaGVyZQo=
          description: Pagination token to continue a previous listing operation.
          type: string
    FetchByMetadataResponse:
      example:
        namespace: example-namespace
        pagination:
          next: Tm90aGluZyB0byBzZWUgaGVyZQo=
        usage:
          readUnits: 5
        vectors:
          id-1:
            id: id-1
            metadata:
              genre: documentary
              year: 2019
            values:
              - 1
              - 1.5
          id-2:
            id: id-2
            metadata:
              genre: comedy
              year: 2019
            values:
              - 2
              - 1
      description: The response for the `fetch_by_metadata` operation.
      type: object
      properties:
        vectors:
          description: >-
            The fetched vectors, in the form of a map between the fetched ids
            and the fetched vectors
          type: object
          additionalProperties:
            $ref: '#/components/schemas/Vector'
        namespace:
          example: example-namespace
          description: The namespace of the vectors.
          default: ''
          type: string
        usage:
          $ref: '#/components/schemas/Usage'
        pagination:
          $ref: '#/components/schemas/Pagination'
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    Vector:
      type: object
      properties:
        id:
          example: example-vector-1
          description: This is the vector's unique id.
          type: string
          required:
            - id
          minLength: 1
          maxLength: 512
        values:
          example:
            - 0.1
            - 0.2
            - 0.3
            - 0.4
            - 0.5
            - 0.6
            - 0.7
            - 0.8
          description: This is the vector data included in the request.
          type: array
          required:
            - values
          items:
            type: number
            format: float
          minLength: 1
          maxLength: 20000
        sparseValues:
          $ref: '#/components/schemas/SparseValues'
        metadata:
          example:
            genre: documentary
            year: 2019
          description: This is the metadata included in the request.
          type: object
      required:
        - id
    Usage:
      type: object
      properties:
        readUnits:
          example: 5
          description: The number of read units consumed by this operation.
          type: integer
          format: int64
    Pagination:
      type: object
      properties:
        next:
          example: Tm90aGluZyB0byBzZWUgaGVyZQo=
          type: string
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
    SparseValues:
      description: >-
        Vector sparse data. Represented as a list of indices and a list of 
        corresponded values, which must be with the same length.
      type: object
      properties:
        indices:
          example:
            - 1
            - 312
            - 822
            - 14
            - 980
          description: The indices of the sparse data.
          type: array
          required:
            - indices
          items:
            type: integer
            format: int64
          minLength: 1
          maxLength: 1000
        values:
          example:
            - 0.1
            - 0.2
            - 0.3
            - 0.4
            - 0.5
          description: >-
            The corresponding values of the sparse data, which must be with the
            same length as the indices.
          type: array
          required:
            - values
          items:
            type: number
            format: float
          minLength: 1
          maxLength: 1000
      required:
        - indices
        - values
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/admin/fetch_project.md

# Source: https://docs.pinecone.io/reference/api/2025-10/admin-assistant/fetch_project.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Get project details

> Get details about a project.

<RequestExample>
  ```bash curl theme={null}
  PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"
  PROJECT_ID="3fa85f64-5717-4562-b3fc-2c963f66afa6"

  curl -X GET "https://api.pinecone.io/admin/projects/$PROJECT_ID" \
      -H "Authorization: Bearer $PINECONE_ACCESS_TOKEN" \
  	-H "X-Pinecone-Api-Version: 2025-10" \
      -H "accept: application/json"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
    "name": "example-project",
    "max_pods": 0,
    "force_encryption_with_cmek": false,
    "organization_id": "string",
    "created_at": "2025-03-17T00:30:23.262Z"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/admin_2025-10.oas.yaml get /admin/projects/{project_id}
openapi: 3.0.3
info:
  title: Pinecone Admin API
  description: |
    Provides an API for managing a Pinecone organization and its resources.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - BearerAuth: []
tags:
  - name: API Keys
    description: Actions that manage API Keys.
  - name: Organizations
    description: Actions that manage organizations.
  - name: Projects
    description: Actions that manage projects.
paths:
  /admin/projects/{project_id}:
    get:
      tags:
        - Projects
      summary: Get project details
      description: Get details about a project.
      operationId: fetch_project
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: project_id
          description: Project ID
          required: true
          schema:
            type: string
            format: uuid
          style: simple
      responses:
        '200':
          description: The details of a project.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Project'
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Not found
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
        4XX:
          description: Unexpected error on request.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
components:
  schemas:
    Project:
      description: The details of a project.
      type: object
      properties:
        id:
          description: The unique ID of the project.
          type: string
          format: uuid
        name:
          description: The name of the project.
          type: string
          minLength: 1
          maxLength: 512
        max_pods:
          description: The maximum number of Pods that can be created in the project.
          type: integer
        force_encryption_with_cmek:
          description: >-
            Whether to force encryption with a customer-managed encryption key
            (CMEK).
          type: boolean
        organization_id:
          description: The unique ID of the organization that the project belongs to.
          type: string
        created_at:
          description: The date and time when the project was created.
          type: string
          format: date-time
      required:
        - id
        - name
        - max_pods
        - force_encryption_with_cmek
        - organization_id
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, or `UNPROCESSABLE_ENTITY`.        
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    BearerAuth:
      type: http
      scheme: bearer
      description: >
        An [access
        token](https://docs.pinecone.io/guides/organizations/manage-service-accounts#retrieve-an-access-token)
        must be provided in the `Authorization` header using the `Bearer`
        scheme.

````

---

# Source: https://docs.pinecone.io/guides/assistant/files-overview.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Files in Pinecone Assistant

> Understand supported file types and metadata in Pinecone Assistant.

export const word_0 = "files"

Before you can chat with the assistant, you need to [upload files](/guides/assistant/manage-files#upload-a-local-file). The files provide your assistant with context and information to reference when generating responses. Files are not shared across assistants.

### Supported file types

Pinecone Assistant supports the following file types:

* DOCX (.docx)
* JSON (.json)
* Markdown (.md)
* PDF (.pdf)
* Text (.txt)

<Note>
  For PDF files, assistants support [multimodal context](/guides/assistant/multimodal), allowing them to analyze and gather context from images. This feature is in [public preview](/release-notes/feature-availability).
</Note>

For information about file size and storage limits, see [Pricing and limits](/guides/assistant/pricing-and-limits).

### File storage

Files are uploaded to Google Cloud Storage (`us-central1` region) and to your organization's Pinecone vector database. The assistant processes the files, so data is not sent outside of blob storage or Pinecone.

Some API responses include a `signed_url` field, which provides temporary, read-only access to one of the assistant's files. The URL is [signed](https://cloud.google.com/storage/docs/access-control/signed-urls) and hard to guess, but publicly accessible, so treat it as sensitive. `signed_url` links expire in one hour.

### File metadata

You can [upload a file with metadata](/guides/assistant/manage-files#upload-a-file-with-metadata), which allows you to store additional information about the file as key-value pairs.

<Warning>
  File metadata can be set only when the file is uploaded. You cannot update metadata after the file is uploaded.
</Warning>

File metadata can be used for the following purposes:

* [Filtering chat responses](/guides/assistant/chat-with-assistant#filter-chat-with-metadata): Specify filters on assistant responses so only files that match the metadata filter are referenced in the response. Chat requests without metadata filters do not consider metadata.
* [Viewing a filtered list of files](/guides/assistant/manage-files#view-a-filtered-list-of-files): Use metadata filters to list files in an assistant that match specific criteria.

#### Supported metadata size and format

Pinecone Assistant supports 16 KB of metadata per file.

* Metadata fields must be key-value pairs in a flat JSON object. Nested JSON objects are not supported.
* Keys must be strings and must not start with a `$`.
* Values must be one of the following data types:
  * String
  * Integer (converted to a 64-bit floating point by Pinecone)
  * Floating point
  * Boolean (`true`, `false`)
  * List of strings
* Null metadata values aren't supported. Instead of setting a key to `null`, remove the key from the metadata payload.

**Examples**

<CodeGroup>
  ```json Valid metadata theme={null}
  {
    "document_id": "document1",
    "document_title": "Introduction to Vector Databases",
    "chunk_number": 1,
    "chunk_text": "First chunk of the document content...",
    "is_public": true,
    "tags": ["beginner", "database", "vector-db"],
    "scores": ["85", "92"]
  }
  ```

  ```json Invalid metadata theme={null}
  {
    "document": {       // Nested JSON objects are not supported
      "document_id": "document1",
      "document_title": "Introduction to Vector Databases",
    },
    "$chunk_number": 1, // Keys must not start with a `$`
    "chunk_text": null, // Null values are not supported
    "is_public": true,
    "tags": ["beginner", "database", "vector-db"],
    "scores": [85, 92]  // Lists of non-strings are not supported
  }
  ```
</CodeGroup>

#### Metadata query language

Pinecone's filtering language supports the following operators:

| Operator  | Function                                                                                                                           | Supported types         |
| :-------- | :--------------------------------------------------------------------------------------------------------------------------------- | :---------------------- |
| `$eq`     | Matches {word_0} with metadata values that are equal to a specified value. Example: `{"genre": {"$eq": "documentary"}}`            | Number, string, boolean |
| `$ne`     | Matches {word_0} with metadata values that are not equal to a specified value. Example: `{"genre": {"$ne": "drama"}}`              | Number, string, boolean |
| `$gt`     | Matches {word_0} with metadata values that are greater than a specified value. Example: `{"year": {"$gt": 2019}}`                  | Number                  |
| `$gte`    | Matches {word_0} with metadata values that are greater than or equal to a specified value. Example:`{"year": {"$gte": 2020}}`      | Number                  |
| `$lt`     | Matches {word_0} with metadata values that are less than a specified value. Example: `{"year": {"$lt": 2020}}`                     | Number                  |
| `$lte`    | Matches {word_0} with metadata values that are less than or equal to a specified value. Example: `{"year": {"$lte": 2020}}`        | Number                  |
| `$in`     | Matches {word_0} with metadata values that are in a specified array. Example: `{"genre": {"$in": ["comedy", "documentary"]}}`      | String, number          |
| `$nin`    | Matches {word_0} with metadata values that are not in a specified array. Example: `{"genre": {"$nin": ["comedy", "documentary"]}}` | String, number          |
| `$exists` | Matches {word_0} with the specified metadata field. Example: `{"genre": {"$exists": true}}`                                        | Number, string, boolean |
| `$and`    | Joins query clauses with a logical `AND`. Example: `{"$and": [{"genre": {"$eq": "drama"}}, {"year": {"$gte": 2020}}]}`             | -                       |
| `$or`     | Joins query clauses with a logical `OR`. Example: `{"$or": [{"genre": {"$eq": "drama"}}, {"year": {"$gte": 2020}}]}`               | -                       |

<Note>
  Only `$and` and `$or` are allowed at the top level of the query expression.
</Note>

<Note>
  Each `$in` or `$nin` operator accepts a maximum of 10,000 values. Exceeding this limit will cause the request to fail. For more information, see [Metadata filter limits](/reference/api/database-limits#metadata-filter-limits).
</Note>

For example, the following has a `"genre"` metadata field with a list of strings:

```JSON JSON theme={null}
{ "genre": ["comedy", "documentary"] }
```

This means `"genre"` takes on both values, and requests with the following filters will match:

```JSON JSON theme={null}
{"genre":"comedy"}

{"genre": {"$in":["documentary","action"]}}

{"$and": [{"genre": "comedy"}, {"genre":"documentary"}]}
```

However, requests with the following filter will **not** match:

```JSON JSON theme={null}
{ "$and": [{ "genre": "comedy" }, { "genre": "drama" }] }
```

Additionally, requests with the following filters will **not** match because they are invalid. They will result in a compilation error:

```json JSON theme={null}
# INVALID QUERY:
{"genre": ["comedy", "documentary"]}
```

```json JSON theme={null}
# INVALID QUERY:
{"genre": {"$eq": ["comedy", "documentary"]}}
```

---

# Source: https://docs.pinecone.io/guides/search/filter-by-metadata.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Filter by metadata

> Narrow search results with metadata filtering.

export const word_0 = "vectors"

Every [record](/guides/get-started/concepts#record) in an index must contain an ID and a dense or sparse vector. In addition, you can include [metadata key-value pairs](/guides/index-data/indexing-overview#metadata) to store related information or context. When you search the index, you can then include a metadata filter to limit the search to records matching the filter expression.

## Search with a metadata filter

The following code searches for the 3 records that are most semantically similar to a query and that have a `category` metadata field with the value `digestive system`.

<Tabs>
  <Tab title="Search with text">
    <Note>
      Searching with text is supported only for [indexes with integrated embedding](/guides/index-data/indexing-overview#integrated-embedding).
    </Note>

    <CodeGroup>
      ```python Python theme={null}
      from pinecone import Pinecone

      pc = Pinecone(api_key="YOUR_API_KEY")

      # To get the unique host for an index, 
      # see https://docs.pinecone.io/guides/manage-data/target-an-index
      index = pc.Index(host="INDEX_HOST")

      filtered_results = index.search(
          namespace="example-namespace", 
          query={
              "inputs": {"text": "Disease prevention"}, 
              "top_k": 3,
              "filter": {"category": "digestive system"},
          },
          fields=["category", "chunk_text"]
      )

      print(filtered_results)
      ```

      ```javascript JavaScript theme={null}
      import { Pinecone } from '@pinecone-database/pinecone'

      const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      const namespace = pc.index("INDEX_NAME", "INDEX_HOST").namespace("example-namespace");

      const response = await namespace.searchRecords({
        query: {
          topK: 3,
          inputs: { text: "Disease prevention" },
          filter: { category: "digestive system" }
        },
        fields: ['chunk_text', 'category']
      });

      console.log(response);
      ```

      ```java Java theme={null}
      import io.pinecone.clients.Index;
      import io.pinecone.configs.PineconeConfig;
      import io.pinecone.configs.PineconeConnection;
      import org.openapitools.db_data.client.ApiException;
      import org.openapitools.db_data.client.model.SearchRecordsResponse;

      import java.util.*;

      public class SearchText {
          public static void main(String[] args) throws ApiException {
              PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
              // To get the unique host for an index, 
              // see https://docs.pinecone.io/guides/manage-data/target-an-index
              config.setHost("INDEX_HOST");
              PineconeConnection connection = new PineconeConnection(config);

              Index index = new Index(config, connection, "integrated-dense-java");

              String query = "Disease prevention";
              List<String> fields = new ArrayList<>();
              fields.add("category");
              fields.add("chunk_text");

              Map<String, Object> filter = new HashMap<>();
              filter.put("category", "digestive system");

              // Search the dense index
              SearchRecordsResponse recordsResponse = index.searchRecordsByText(query,  "example-namespace", fields, 3, filter, null);

              // Print the results
              System.out.println(recordsResponse);
          }
      }
      ```

      ```go Go theme={null}
      package main

      import (
          "context"
          "encoding/json"
          "fmt"
          "log"

          "github.com/pinecone-io/go-pinecone/v4/pinecone"
      )

      func prettifyStruct(obj interface{}) string {
        	bytes, _ := json.MarshalIndent(obj, "", "  ")
          return string(bytes)
      }

      func main() {
          ctx := context.Background()

          pc, err := pinecone.NewClient(pinecone.NewClientParams{
              ApiKey: "YOUR_API_KEY",
          })
          if err != nil {
              log.Fatalf("Failed to create Client: %v", err)
          }

          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
          if err != nil {
              log.Fatalf("Failed to create IndexConnection for Host: %v", err)
          } 

          metadataMap := map[string]interface{}{
              "category": map[string]interface{}{
                  "$eq": "digestive system",
              },
          }
          res, err := idxConnection.SearchRecords(ctx, &pinecone.SearchRecordsRequest{
              Query: pinecone.SearchRecordsQuery{
                  TopK: 3,
                  Inputs: &map[string]interface{}{
                      "text": "Disease prevention",
                  },
                  Filter: &metadataMap,
              },
              Fields: &[]string{"chunk_text", "category"},
          })
          if err != nil {
              log.Fatalf("Failed to search records: %v", err)
          }
          fmt.Printf(prettifyStruct(res))
      }
      ```

      ```csharp C# theme={null}
      using Pinecone;

      var pinecone = new PineconeClient("YOUR_API_KEY");

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      var index = pinecone.Index(host: "INDEX_HOST");

      var response = await index.SearchRecordsAsync(
          "example-namespace",
          new SearchRecordsRequest
          {
              Query = new SearchRecordsRequestQuery
              {
                  TopK = 4,
                  Inputs = new Dictionary<string, object?> { { "text", "Disease prevention" } },
                  Filter = new Dictionary<string, object?>
                  {
                      ["category"] = new Dictionary<string, object?>
                      {
                          ["$eq"] = "digestive system"
                      }
                  }
              },
              Fields = ["category", "chunk_text"],
          }
      );

      Console.WriteLine(response);
      ```

      ```shell curl theme={null}
      INDEX_HOST="INDEX_HOST"
      NAMESPACE="YOUR_NAMESPACE"
      PINECONE_API_KEY="YOUR_API_KEY"

      curl "https://$INDEX_HOST/records/namespaces/$NAMESPACE/search" \
        -H "Accept: application/json" \
        -H "Content-Type: application/json" \
        -H "Api-Key: $PINECONE_API_KEY" \
        -H "X-Pinecone-Api-Version: unstable" \
        -d '{
              "query": {
                  "inputs": {"text": "Disease prevention"},
                  "top_k": 3,
                  "filter": {"category": "digestive system"}
              },
              "fields": ["category", "chunk_text"]
           }'
      ```
    </CodeGroup>
  </Tab>

  <Tab title="Search with a vector">
    <CodeGroup>
      ```Python Python theme={null}
      from pinecone.grpc import PineconeGRPC as Pinecone

      pc = Pinecone(api_key="YOUR_API_KEY")

      # To get the unique host for an index, 
      # see https://docs.pinecone.io/guides/manage-data/target-an-index
      index = pc.Index(host="INDEX_HOST")

      index.query(
          namespace="example-namespace",
          vector=[0.0236663818359375,-0.032989501953125, ..., -0.01041412353515625,0.0086669921875], 
          top_k=3,
          filter={
              "category": {"$eq": "digestive system"}
          },
          include_metadata=True,
          include_values=False
      )
      ```

      ```JavaScript JavaScript theme={null}
      import { Pinecone } from '@pinecone-database/pinecone'

      const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      const index = pc.index("INDEX_NAME", "INDEX_HOST")

      const queryResponse = await index.namespace('example-namespace').query({
          vector: [0.0236663818359375,-0.032989501953125,...,-0.01041412353515625,0.0086669921875],
          topK: 3,
          filter: {
              "category": { "$eq": "digestive system" }
          }
          includeValues: false,
          includeMetadata: true,
      });
      ```

      ```java Java theme={null}
      import com.google.protobuf.Struct;
      import com.google.protobuf.Value;
      import io.pinecone.clients.Index;
      import io.pinecone.configs.PineconeConfig;
      import io.pinecone.configs.PineconeConnection;
      import io.pinecone.unsigned_indices_model.QueryResponseWithUnsignedIndices;

      import java.util.Arrays;
      import java.util.List;

      public class QueryExample {
          public static void main(String[] args) {
              PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
              // To get the unique host for an index, 
              // see https://docs.pinecone.io/guides/manage-data/target-an-index
              config.setHost("INDEX_HOST");
              PineconeConnection connection = new PineconeConnection(config);
              Index index = new Index(connection, "INDEX_NAME");
              List<Float> query = Arrays.asList(0.0236663818359375f, -0.032989501953125f, ..., -0.01041412353515625f, 0.0086669921875f);
              Struct filter = Struct.newBuilder()
                      .putFields("category", Value.newBuilder()
                              .setStructValue(Struct.newBuilder()
                                      .putFields("$eq", Value.newBuilder()
                                              .setStringValue("digestive system")
                                              .build()))
                              .build())
                      .build();

              QueryResponseWithUnsignedIndices queryResponse = index.query(1, query, null, null, null, "example-namespace", filter, false, true);
              System.out.println(queryResponse);
          }
      }
      ```

      ```go Go theme={null}
      package main

      import (
          "context"
          "encoding/json"
          "fmt"
          "log"

          "github.com/pinecone-io/go-pinecone/v4/pinecone"
      )

      func prettifyStruct(obj interface{}) string {
      	bytes, _ := json.MarshalIndent(obj, "", "  ")
      	return string(bytes)
      }

      func main() {
          ctx := context.Background()

          pc, err := pinecone.NewClient(pinecone.NewClientParams{
              ApiKey: "YOUR_API_KEY",
          })
          if err != nil {
              log.Fatalf("Failed to create Client: %v", err)
          }

          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
          if err != nil {
              log.Fatalf("Failed to create IndexConnection for Host: %v", err)
        	}

          queryVector := []float32{0.0236663818359375,-0.032989501953125,...,-0.01041412353515625,0.0086669921875}

          metadataMap := map[string]interface{}{
              "category": map[string]interface{}{
                  "$eq": "digestive system",
              }
          }

          metadataFilter, err := structpb.NewStruct(metadataMap)
          if err != nil {
              log.Fatalf("Failed to create metadata map: %v", err)
          }

          res, err := idxConnection.QueryByVectorValues(ctx, &pinecone.QueryByVectorValuesRequest{
              Vector:          queryVector,
              TopK:            3,
              MetadataFilter: metadataFilter,
              IncludeValues:   false,
              includeMetadata: true,
          })
          if err != nil {
              log.Fatalf("Error encountered when querying by vector: %v", err)
          } else {
              fmt.Printf(prettifyStruct(res))
          }
      }
      ```

      ```csharp C# theme={null}
      using Pinecone;

      var pinecone = new PineconeClient("YOUR_API_KEY");

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      var index = pinecone.Index(host: "INDEX_HOST");

      var queryResponse = await index.QueryAsync(new QueryRequest {
          Vector = new[] { 0.0236663818359375f ,-0.032989501953125f, ..., -0.01041412353515625f, 0.0086669921875f },
          Namespace = "example-namespace",
          TopK = 3,
          Filter = new Metadata
          {
              ["category"] =
                  new Metadata
                  {
                      ["$eq"] = "digestive system",
                  }
          },
          IncludeMetadata = true,
      });

      Console.WriteLine(queryResponse);
      ```

      ```bash curl theme={null}
      # To get the unique host for an index,
      # see https://docs.pinecone.io/guides/manage-data/target-an-index
      PINECONE_API_KEY="YOUR_API_KEY"
      INDEX_HOST="INDEX_HOST"

      curl "https://$INDEX_HOST/query" \
        -H "Api-Key: $PINECONE_API_KEY" \
        -H 'Content-Type: application/json' \
        -H "X-Pinecone-Api-Version: 2025-10" \
        -d '{
              "vector": [0.0236663818359375,-0.032989501953125,...,-0.01041412353515625,0.0086669921875],
              "namespace": "example-namespace",
              "topK": 3,
              "filter": {"category": {"$eq": "digestive system"}},
              "includeMetadata": true,
              "includeValues": false
          }'
      ```
    </CodeGroup>
  </Tab>
</Tabs>

## Metadata filter expressions

Pinecone's filtering language supports the following operators:

| Operator  | Function                                                                                                                           | Supported types         |
| :-------- | :--------------------------------------------------------------------------------------------------------------------------------- | :---------------------- |
| `$eq`     | Matches {word_0} with metadata values that are equal to a specified value. Example: `{"genre": {"$eq": "documentary"}}`            | Number, string, boolean |
| `$ne`     | Matches {word_0} with metadata values that are not equal to a specified value. Example: `{"genre": {"$ne": "drama"}}`              | Number, string, boolean |
| `$gt`     | Matches {word_0} with metadata values that are greater than a specified value. Example: `{"year": {"$gt": 2019}}`                  | Number                  |
| `$gte`    | Matches {word_0} with metadata values that are greater than or equal to a specified value. Example:`{"year": {"$gte": 2020}}`      | Number                  |
| `$lt`     | Matches {word_0} with metadata values that are less than a specified value. Example: `{"year": {"$lt": 2020}}`                     | Number                  |
| `$lte`    | Matches {word_0} with metadata values that are less than or equal to a specified value. Example: `{"year": {"$lte": 2020}}`        | Number                  |
| `$in`     | Matches {word_0} with metadata values that are in a specified array. Example: `{"genre": {"$in": ["comedy", "documentary"]}}`      | String, number          |
| `$nin`    | Matches {word_0} with metadata values that are not in a specified array. Example: `{"genre": {"$nin": ["comedy", "documentary"]}}` | String, number          |
| `$exists` | Matches {word_0} with the specified metadata field. Example: `{"genre": {"$exists": true}}`                                        | Number, string, boolean |
| `$and`    | Joins query clauses with a logical `AND`. Example: `{"$and": [{"genre": {"$eq": "drama"}}, {"year": {"$gte": 2020}}]}`             | -                       |
| `$or`     | Joins query clauses with a logical `OR`. Example: `{"$or": [{"genre": {"$eq": "drama"}}, {"year": {"$gte": 2020}}]}`               | -                       |

<Note>
  Only `$and` and `$or` are allowed at the top level of the query expression.
</Note>

<Note>
  Each `$in` or `$nin` operator accepts a maximum of 10,000 values. Exceeding this limit will cause the request to fail. For more information, see [Metadata filter limits](/reference/api/database-limits#metadata-filter-limits).
</Note>

For example, the following has a `"genre"` metadata field with a list of strings:

```JSON JSON theme={null}
{ "genre": ["comedy", "documentary"] }
```

This means `"genre"` takes on both values, and requests with the following filters will match:

```JSON JSON theme={null}
{"genre":"comedy"}

{"genre": {"$in":["documentary","action"]}}

{"$and": [{"genre": "comedy"}, {"genre":"documentary"}]}
```

However, requests with the following filter will **not** match:

```JSON JSON theme={null}
{ "$and": [{ "genre": "comedy" }, { "genre": "drama" }] }
```

Additionally, requests with the following filters will **not** match because they are invalid. They will result in a compilation error:

```json JSON theme={null}
# INVALID QUERY:
{"genre": ["comedy", "documentary"]}
```

```json JSON theme={null}
# INVALID QUERY:
{"genre": {"$eq": ["comedy", "documentary"]}}
```

---

# Source: https://docs.pinecone.io/integrations/fleak.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Fleak

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Fleak simplifies the process of building, deploying, and managing data workflows. As a low-code platform, Fleak lets users create and deploy complex workflows using SQL and pre-configured processing nodes. The platform facilitates seamless data processing, microservice interactions, and the inference of large language models (LLMs) within a single, intuitive environment. Fleak makes advanced technology accessible and manageable for Pinecone users without requiring extensive coding hours or infrastructure knowledge.

The platform provides serverless, autoscaling HTTP API endpoints, ensuring that workflows are robust, reliable, and scalable to meet the complex data needs of enterprises. This setup allows businesses to automate and enhance their operations, driving productivity and innovation through powerful, user-friendly tools.

By integrating Pinecone into Fleak's platform, users can access and leverage their vector data to enrich their workflows without additional engineering overhead, enabling seamless data-driven decision-making, advanced analytics, and the integration of AI-driven insights.

<PrimarySecondaryCTA primaryHref={"https://docs.fleak.ai/1.0/getting-started/pinecone-integration"} primaryLabel={"Get started"} />

---

# Source: https://docs.pinecone.io/integrations/flowise.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# FlowiseAI

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Flowise is a low-code LLM apps development platform. It supports integrations with dozens of systems, including databases and chat models.

The Pinecone integration with Flowise allows users to build RAG apps, including upserting and querying documents.

<PrimarySecondaryCTA primaryHref={"https://docs.flowiseai.com/integrations/langchain/vector-stores/pinecone"} primaryLabel={"Get started"} />

---

# Source: https://docs.pinecone.io/integrations/gathr.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Gathr

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

[Gathr](https://www.gathr.one/) is the world's first and only "data to outcome" platform. Leading enterprises use Gathr to build and operationalize data and AI-driven solutions at scale.

Gathr unifies data engineering, machine learning, generative AI, actionable analytics, and process automation on a single platform. With Gen AI capabilities and no-code rapid application development, Gathr significantly boosts productivity for all. The unified experience fosters seamless handoff and collaboration between teams, accelerating the journey from prototype to production.

Users have achieved success with Gathr, from ingesting petabyte-scale data in real time to orchestrating thousands of complex data processing pipelines in months and delivering actionable insights and xOps solutions to multiply business impact. Additionally, Gathr helps enterprises architect Gen AI solutions for use cases like document summarization, sentiment analysis, next best action, insider threat detection, predictive maintenance, custom chatbots, and more.

Gathr Gen AI Fabric is designed to build enterprise-grade Gen AI solutions end-to-end on a unified platform. It offers production-ready building blocks for creating Gen AI solutions, out-of-the-box Gen AI solution templates, and GathrIQ, a data-to-outcome copilot.

One of the building blocks is integration with Vector DB and Knowledge Graphs. Gathr supports reading and writing from Pinecone using a built-in, ready-to-use connector to support use cases requiring knowledge graphs.

<PrimarySecondaryCTA primaryHref={"https://docs.gathr.one/gathr-saas/docs/components/pinecone/"} primaryLabel={"Get started"} />

---

# Source: https://docs.pinecone.io/reference/api/2025-10/inference/generate-embeddings.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Generate vectors

> Generate vector embeddings for input data. This endpoint uses Pinecone's [hosted embedding models](https://docs.pinecone.io/guides/index-data/create-an-index#embedding-models).

<RequestExample>
  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl https://api.pinecone.io/embed \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
        "model": "llama-text-embed-v2",
        "parameters": {
          "input_type": "passage",
          "truncate": "END"
        },
        "inputs": [
          {"text": "Apple is a popular fruit known for its sweetness and crisp texture."},
          {"text": "The tech company Apple is known for its innovative products like the iPhone."},
          {"text": "Many people enjoy eating apples as a healthy snack."},
          {"text": "Apple Inc. has revolutionized the tech industry with its sleek designs and user-friendly interfaces."},
          {"text": "An apple a day keeps the doctor away, as the saying goes."},
          {"text": "Apple Computer Company was founded on April 1, 1976, by Steve Jobs, Steve Wozniak, and Ronald Wayne as a partnership."}
        ]
    }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "data": [
      {
        "values": [
          0.04925537109375,
          -0.01313018798828125,
          -0.0112762451171875,
          ...
        ]
      }, 
      ...
    ],
    "model": "llama-text-embed-v2",
    "usage": {
      "total_tokens": 130
    }
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/inference_2025-10.oas.yaml post /embed
openapi: 3.0.3
info:
  title: Pinecone Inference API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Inference
    description: Model inference
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /embed:
    post:
      tags:
        - Inference
      summary: Generate vectors
      description: >-
        Generate vector embeddings for input data. This endpoint uses Pinecone's
        [hosted embedding
        models](https://docs.pinecone.io/guides/index-data/create-an-index#embedding-models).
      operationId: embed
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        description: Generate embeddings for inputs.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/EmbedRequest'
      responses:
        '200':
          description: OK
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/EmbeddingsList'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-metric-validation-error:
                  summary: Validation error
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Bad request. The request body included invalid request
                        parameters.
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    EmbedRequest:
      type: object
      properties:
        model:
          example: multilingual-e5-large
          description: >-
            The
            [model](https://docs.pinecone.io/guides/index-data/create-an-index#embedding-models)
            to use for embedding generation.
          type: string
        parameters:
          example:
            input_type: passage
            truncate: END
          description: >-
            Additional model-specific parameters. Refer to the [model
            guide](https://docs.pinecone.io/guides/index-data/create-an-index#embedding-models)
            for available model parameters.
          type: object
          additionalProperties: true
        inputs:
          description: List of inputs to generate embeddings for.
          type: array
          items:
            type: object
            properties:
              text:
                example: The quick brown fox jumps over the lazy dog.
                description: The text input to generate embeddings for.
                type: string
      required:
        - model
        - inputs
    EmbeddingsList:
      description: Embeddings generated for the input.
      type: object
      properties:
        model:
          example: multilingual-e5-large
          description: The model used to generate the embeddings
          type: string
        vector_type:
          example: dense
          description: >-
            Indicates whether the response data contains 'dense' or 'sparse'
            embeddings.
          type: string
        data:
          description: The embeddings generated for the inputs.
          type: array
          items:
            $ref: '#/components/schemas/Embedding'
        usage:
          description: Usage statistics for the model inference.
          type: object
          properties:
            total_tokens:
              example: 205
              description: Total number of tokens consumed across all inputs.
              type: integer
              format: int32
              minimum: 0
      required:
        - model
        - vector_type
        - data
        - usage
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, or `FORBIDDEN`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable error message describing the error.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    Embedding:
      description: Embedding of a single input
      discriminator:
        propertyName: vector_type
        mapping:
          dense: '#/components/schemas/DenseEmbedding'
          sparse: '#/components/schemas/SparseEmbedding'
      type: object
      oneOf:
        - $ref: '#/components/schemas/DenseEmbedding'
        - $ref: '#/components/schemas/SparseEmbedding'
    DenseEmbedding:
      title: Dense embedding
      description: A dense embedding of a single input
      type: object
      properties:
        values:
          example:
            - 0.1
            - 0.2
            - 0.3
          description: The dense embedding values.
          type: array
          items:
            type: number
            format: float
        vector_type:
          $ref: '#/components/schemas/VectorType'
      required:
        - values
        - vector_type
    SparseEmbedding:
      title: Sparse embedding
      description: A sparse embedding of a single input
      type: object
      properties:
        sparse_values:
          example:
            - 0.1
            - 0.2
            - 0.3
          description: The sparse embedding values.
          type: array
          items:
            type: number
            format: float
        sparse_indices:
          example:
            - 10
            - 3
            - 156
          description: The sparse embedding indices.
          type: array
          items:
            type: integer
            format: int64
            minimum: 0
            maximum: 4294967295
        sparse_tokens:
          example:
            - quick
            - brown
            - fox
          description: The normalized tokens used to create the sparse embedding.
          type: array
          items:
            type: string
        vector_type:
          $ref: '#/components/schemas/VectorType'
      required:
        - sparse_values
        - sparse_indices
        - vector_type
    VectorType:
      description: Indicates whether this is a 'dense' or 'sparse' embedding.
      type: string
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/integrations/genkit.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Genkit

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

The [Genkit](https://firebase.google.com/docs/genkit) Pinecone plugin empowers developers to reduce the complexity of integrating AI components through simple indexers, embedders and retrievers abstractions. Through the Genkit Pinecone plugin, developers can integrate AI models with their own custom logic and data to build AI features optimized for their businesses.

Additionally, developers can analyze unstructured text, generate creative content, select tasks, and send results back to their app as structured type-safe objects.
The plugin provides a common format for content that supports combinations of text, data, and other media. Developers can use Genkit for models that perform any generative task (such as image generation), not just LLMs.

<PrimarySecondaryCTA primaryLabel="Get started" primaryHref="https://firebase.google.com/docs/genkit/plugins/pinecone" />

---

# Source: https://docs.pinecone.io/reference/api/2025-10/admin/get_token.md

# Source: https://docs.pinecone.io/reference/api/2025-10/admin-assistant/get_token.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Create an access token

> Obtain an access token for a service account using the OAuth2 client credentials flow. An access token is needed to authorize requests to the Pinecone Admin API.
The host domain for OAuth endpoints is `login.pinecone.io`.


<RequestExample>
  ```bash curl theme={null}
  curl "https://login.pinecone.io/oauth/token" \ # Note: Base URL is login.pinecone.io
  	-H "Content-Type: application/json" \
  	-d '{
  		"grant_type": "client_credentials",
  		"client_id": "YOUR_CLIENT_ID",
  		"client_secret": "YOUR_CLIENT_SECRET",
  		"audience": "https://api.pinecone.io/"
  	}'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
      "access_token":"YOUR_ACCESS_TOKEN",
      "expires_in":86400,
      "token_type":"Bearer"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/oauth_2025-10.oas.yaml post /oauth/token
openapi: 3.0.3
info:
  title: Pinecone OAuth API
  description: |
    Provides an API for authenticating with Pinecone.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://login.pinecone.io
security: []
tags:
  - name: OAuth
    description: Authentication using the OAuth2 protocol.
paths:
  /oauth/token:
    post:
      tags:
        - OAuth
      summary: Create an access token
      description: >
        Obtain an access token for a service account using the OAuth2 client
        credentials flow. An access token is needed to authorize requests to the
        Pinecone Admin API.

        The host domain for OAuth endpoints is `login.pinecone.io`.
      operationId: get_token
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        description: |
          A request to exchange client credentials for an access token.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TokenRequest'
          application/x-www-form-urlencoded:
            schema:
              $ref: '#/components/schemas/TokenRequest'
        required: true
      responses:
        '200':
          description: A response that contains the access token.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TokenResponse'
        '400':
          description: Invalid request.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '401':
          description: Unauthorized.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '403':
          description: Forbidden.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '429':
          description: Too many requests.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '501':
          description: Not implemented.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '503':
          description: Service unavailable.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
components:
  schemas:
    TokenRequest:
      example:
        audience: https://api.pinecone.io/
        client_id: I1r8m4i6jX9JTFYk0t3q85HWzciEgcA5
        client_secret: EriX...j2ci
        grant_type: client_credentials
      description: A request to obtain an access token.
      type: object
      properties:
        client_id:
          description: |
            The service account's client ID.
          type: string
        client_secret:
          description: |
            The service account's client secret.
          type: string
        grant_type:
          description: |
            The type of grant to use.
          x-enum:
            - client_credentials
          type: string
        audience:
          description: |
            The audience for the token.
          x-enum:
            - https://api.pinecone.io/
          type: string
      required:
        - client_id
        - client_secret
        - grant_type
        - audience
    TokenResponse:
      example:
        access_token: eyJz93a...k4laUWw
        expires_in: 1800
        token_type: Bearer
      description: A response that contains the access token.
      type: object
      properties:
        access_token:
          description: The access token.
          type: string
        token_type:
          description: |-
            The type of token.
            Possible values: `Bearer`.
          x-enum:
            - Bearer
          type: string
        expires_in:
          description: The number of seconds until the token expires.
          type: integer
      required:
        - access_token
        - token_type
        - expires_in
    ErrorResponse:
      example:
        error: access_denied
        error_description: Unauthorized
      description: A response indicating that an error occurred.
      type: object
      properties:
        error:
          description: A code identifying the error that occurred.
          type: string
        error_description:
          description: A human-readable description of the error.
          type: string

````

---

# Source: https://docs.pinecone.io/integrations/github-copilot.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# GitHub Copilot

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Access the Pinecone Copilot Extension through our GitHub Marketplace listing. The Pinecone Copilot Extension serves as a seamless bridge between you and your Pinecone data-- providing product information, coding assistance, troubleshooting capabilities and streamlining the debugging process.

This extension offers personalized recommendations right to your fingertips, enabling you to swiftly retrieve relevant data and collaborate effectively with Copilot.

<PrimarySecondaryCTA
  primaryLabel={"Get started"}
  primaryHref={
  "https://github.com/marketplace/pinecone-ai"
}
/>

---

# Source: https://docs.pinecone.io/reference/go-sdk.md

# Go SDK

<Tip>
  See the [Go SDK documentation](https://github.com/pinecone-io/go-pinecone/blob/main/README.md) for full installation instructions and usage examples.

  To make a feature request or report an issue, please [file an issue](https://github.com/pinecone-io/go-pinecone/issues).
</Tip>

## Requirements

The Pinecone Go SDK requires a Go version with [modules](https://go.dev/wiki/Modules) support.

## SDK versions

SDK versions are pinned to specific [API versions](/reference/api/versioning). When a new API version is released, a new version of the SDK is also released.

The mappings between API versions and Go SDK versions are as follows:

| API version | SDK version |
| :---------- | :---------- |
| `2025-04`   | v4.x        |
| `2025-01`   | v3.x        |
| `2024-10`   | v2.x        |
| `2024-07`   | v1.x        |
| `2024-04`   | v0.x        |

When a new stable API version is released, you should upgrade your SDK to the latest version to ensure compatibility with the latest API changes.

## Install

To install the latest version of the [Go SDK](https://github.com/pinecone-io/go-pinecone), add a dependency to the current module:

```shell  theme={null}
go get github.com/pinecone-io/go-pinecone/v4/pinecone
```

To install a specific version of the Go SDK, run the following command:

```shell  theme={null}
go get github.com/pinecone-io/go-pinecone/v4/pinecone@<version>
```

To check your SDK version, run the following command:

```shell  theme={null}
go list -u -m all | grep go-pinecone
```

## Upgrade

<Warning>
  Before upgrading to `v3.0.0` or later, update all relevant code to account for the breaking changes explained [here](/release-notes/2025#2025-02-07-4).
</Warning>

If you already have the Go SDK, upgrade to the latest version as follows:

```shell  theme={null}
go get -u github.com/pinecone-io/go-pinecone/v4/pinecone@latest
```

## Initialize

Once installed, you can import the SDK and then use an [API key](/guides/production/security-overview#api-keys) to initialize a client instance:

```Go  theme={null}
package main

import (
    "context"
    "log"

    "github.com/pinecone-io/go-pinecone/v4/pinecone"
)

func main() {
    ctx := context.Background()

    pc, err := pinecone.NewClient(pinecone.NewClientParams{
        ApiKey: "YOUR_API_KEY",
    })
    if err != nil {
        log.Fatalf("Failed to create Client: %v", err)
    }
} 
```

---

# Source: https://docs.pinecone.io/integrations/google-cloud-marketplace.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Google Cloud Marketplace

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Access Pinecone through our Google Cloud Marketplace listing. Google Cloud Marketplace allows you to manage Pinecone and other third-party software from a centralized location, and simplifies software licensing and procurement with flexible pricing options and multiple deployment methods.

You can set up pay-as-you-go billing for a Pinecone organization through the Google Cloud Marketplace.

<PrimarySecondaryCTA primaryLabel={"Get started"} primaryHref={"https://console.cloud.google.com/marketplace/product/pinecone-public/pinecone"} />

---

# Source: https://docs.pinecone.io/integrations/haystack.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Haystack

> Using Haystack and Pinecone to keep your NLP-driven apps up-to-date

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Haystack is the open source Python framework by Deepset for building custom apps with large language models (LLMs). It lets you quickly try out the latest models in natural language processing (NLP) while being flexible and easy to use. Their community of users and builders has helped shape Haystack into what it is today: a complete framework for building production-ready NLP apps.

Haystack and Pinecone integration can be used to keep your NLP-driven apps up-to-date with Haystack's indexing pipelines that help you prepare and maintain your data.

<PrimarySecondaryCTA secondaryHref={"#setup-guide"} secondaryLabel={"View setup guide"} />

## Setup guide

In this guide we will see how to integrate Pinecone and the popular [Haystack library](https://github.com/deepset-ai/haystack) for *Question-Answering*.

### Install Haystack

We start by installing the latest version of Haystack with all dependencies required for the `PineconeDocumentStore`.

```Python Python theme={null}
pip install -U farm-haystack>=1.3.0 pinecone[grpc] datasets
```

### Initialize the PineconeDocumentStore

We initialize a `PineconeDocumentStore` by providing an API key and environment name. [Create an account](https://app.pinecone.io) to get your free API key.

```Python Python theme={null}
from haystack.document_stores import PineconeDocumentStore

document_store = PineconeDocumentStore(
    api_key='<YOUR_API_KEY>',
    index='haystack-extractive-qa',
    similarity="cosine",
    embedding_dim=384
)
```

```
INFO - haystack.document_stores.pinecone -  Index statistics: name: haystack-extractive-qa, embedding dimensions: 384, record count: 0
```

### Prepare data

Before adding data to the document store, we must download and convert data into the Document format that Haystack uses.

We will use the SQuAD dataset available from Hugging Face Datasets.

```Python Python theme={null}
from datasets import load_dataset

# load the squad dataset
data = load_dataset("squad", split="train")
```

Next, we remove duplicates and unecessary columns.

```Python Python theme={null}
# convert to a pandas dataframe
df = data.to_pandas()
# select only title and context column
df = df[["title", "context"]]
# drop rows containing duplicate context passages
df = df.drop_duplicates(subset="context")
df.head()
```

| title | context                     |                                                   |
| ----- | --------------------------- | ------------------------------------------------- |
| 0     | University\_of\_Notre\_Dame | Architecturally, the school has a Catholic cha... |
| 5     | University\_of\_Notre\_Dame | As at most other universities, Notre Dame's st... |
| 10    | University\_of\_Notre\_Dame | The university is the major seat of the Congre... |
| 15    | University\_of\_Notre\_Dame | The College of Engineering was established in ... |
| 20    | University\_of\_Notre\_Dame | All of Notre Dame's undergraduate students are... |

Then convert these records into the Document format.

```Python Python theme={null}
from haystack import Document

docs = []
for d in df.iterrows():
    d = d[1]
    # create haystack document object with text content and doc metadata
    doc = Document(
        content=d["context"],
        meta={
            "title": d["title"],
            'context': d['context']
        }
    )
    docs.append(doc)
```

This `Document` format contains two fields; *'content'* for the text content or paragraphs, and *'meta'* where we can place any additional information that can later be used to apply metadata filtering in our search.

Now we upsert the documents to Pinecone.

```Python Python theme={null}
# upsert the data document to pinecone index
document_store.write_documents(docs)
```

### Initialize retriever

The next step is to create embeddings from these documents. We will use Haystacks `EmbeddingRetriever` with a SentenceTransformer model (`multi-qa-MiniLM-L6-cos-v1`) which has been designed for question-answering.

```Python Python theme={null}
from haystack.retriever.dense import EmbeddingRetriever

retriever = EmbeddingRetriever(
    document_store=document_store,
    embedding_model="multi-qa-MiniLM-L6-cos-v1",
    model_format="sentence_transformers"
)
```

Then we run the `PineconeDocumentStore.update_embeddings` method with the `retriever` provided as an argument. GPU acceleration can greatly reduce the time required for this step.

```Python Python theme={null}
document_store.update_embeddings(
    retriever,
    batch_size=16
)
```

### Inspect documents and embeddings

We can get documents by their ID with the `PineconeDocumentStore.get_documents_by_id` method.

```Python Python theme={null}
d = document_store.get_documents_by_id(ids=['49091c797d2236e73fab510b1e9c7f6b'], return_embedding=True)[0]
```

From here we return can view document content with `d.content` and the document embedding with `d.embedding`.

### Initialize an extractive QA pipeline

An `ExtractiveQAPipeline` contains three key components by default:

* a document store (`PineconeDocumentStore`)
* a retriever model
* a reader model

We use the `deepset/electra-base-squad2` model from the HuggingFace model hub as our reader model.

```Python Python theme={null}
from haystack.nodes import FARMReader

reader = FARMReader(
    model_name_or_path='deepset/electra-base-squad2', 
    use_gpu=True
)
```

We are now ready to initialize the `ExtractiveQAPipeline`.

```Python Python theme={null}
from haystack.pipelines import ExtractiveQAPipeline

pipe = ExtractiveQAPipeline(reader, retriever)
```

### Ask Questions

Using our QA pipeline we can begin querying with `pipe.run`.

```Python Python theme={null}
from haystack.utils import print_answers

query = "What was Albert Einstein famous for?"
# get the answer
answer = pipe.run(
    query=query,
    params={
        "Retriever": {"top_k": 1},
    }
)
# print the answer(s)
print_answers(answer)
```

```
Inferencing Samples: 100%|██████████| 1/1 [00:00<00:00,  3.53 Batches/s]

Query: What was Albert Einstein famous for?
Answers:
[   <Answer {
    'answer': 'his theories of special relativity and general relativity', 'type': 'extractive', 'score': 0.993550717830658,
    'context': 'Albert Einstein is known for his theories of special relativity and general relativity. He also made important contributions to statistical mechanics,',
    'offsets_in_document': [{'start': 29, 'end': 86}],
    'offsets_in_context': [{'start': 29, 'end': 86}], 
    'document_id': '23357c05e3e46bacea556705de1ea6a5',
    'meta': {
        'context': 'Albert Einstein is known for his theories of special relativity and general relativity. He also made important contributions to statistical mechanics, especially his mathematical treatment of Brownian motion, his resolution of the paradox of specific heats, and his connection of fluctuations and dissipation. Despite his reservations about its interpretation, Einstein also made contributions to quantum mechanics and, indirectly, quantum field theory, primarily through his theoretical studies of the photon.', 'title': 'Modern_history'
    }
}>]
```

```Python Python theme={null}
query = "How much oil is Egypt producing in a day?"
# get the answer
answer = pipe.run(
    query=query,
    params={
        "Retriever": {"top_k": 1},
    }
)
# print the answer(s)
print_answers(answer)
```

```
Inferencing Samples: 100%|██████████| 1/1 [00:00<00:00,  3.81 Batches/s]

Query: How much oil is Egypt producing in a day?
Answers:
[   <Answer {
    'answer': '691,000 bbl/d', 'type': 'extractive', 'score': 0.9999906420707703,
    'context': 'Egypt was producing 691,000 bbl/d of oil and 2,141.05 Tcf of natural gas (in 2013), which makes Egypt as the largest oil producer not member of the Or',
    'offsets_in_document': [{'start': 20, 'end': 33}],
    'offsets_in_context': [{'start': 20, 'end': 33}],
    'document_id': '57ed9720050a17237e323da5e3969a9b',
    'meta': {
        'context': 'Egypt was producing 691,000 bbl/d of oil and 2,141.05 Tcf of natural gas (in 2013), which makes Egypt as the largest oil producer not member of the Organization of the Petroleum Exporting Countries (OPEC) and the second-largest dry natural gas producer in Africa. In 2013, Egypt was the largest consumer of oil and natural gas in Africa, as more than 20% of total oil consumption and more than 40% of total dry natural gas consumption in Africa. Also, Egypt possesses the largest oil refinery capacity in Africa 726,000 bbl/d (in 2012). Egypt is currently planning to build its first nuclear power plant in El Dabaa city, northern Egypt.', 'title': 'Egypt'
    }
}>]
```

```Python Python theme={null}
query = "What are the first names of the youtube founders?"
# get the answer
answer = pipe.run(
    query=query,
    params={
        "Retriever": {"top_k": 1},
    }
)
# print the answer(s)
print_answers(answer)
```

```
Inferencing Samples: 100%|██████████| 1/1 [00:00<00:00,  3.83 Batches/s]

Query: What are the first names of the youtube founders?
Answers:
[   <Answer {
    'answer': 'Hurley and Chen', 'type': 'extractive', 'score': 0.9998972713947296,
    'context': 'According to a story that has often been repeated in the media, Hurley and Chen developed the idea for YouTube during the early months of 2005, after ',
    'offsets_in_document': [{'start': 64, 'end': 79}],
    'offsets_in_context': [{'start': 64, 'end': 79}],
    'document_id': 'bd1cbd61ab617d840c5f295e21e80092',
    'meta': {
        'context': 'According to a story that has often been repeated in the media, Hurley and Chen developed the idea for YouTube during the early months of 2005, after they had experienced difficulty sharing videos that had been shot at a dinner party at Chen\'s apartment in San Francisco. Karim did not attend the party and denied that it had occurred, but Chen commented that the idea that YouTube was founded after a dinner party "was probably very strengthened by marketing ideas around creating a story that was very digestible".', 'title': 'YouTube'
    }
}>]
```

We can return multiple answers by setting the `top_k` parameter.

```Python Python theme={null}
query = "Who was the first person to step foot on the moon?"
# get the answer
answer = pipe.run(
    query=query,
    params={
        "Retriever": {"top_k": 3},
    }
)
# print the answer(s)
print_answers(answer)
```

```
Inferencing Samples: 100%|██████████| 1/1 [00:00<00:00,  3.71 Batches/s]
Inferencing Samples: 100%|██████████| 1/1 [00:00<00:00,  3.78 Batches/s]
Inferencing Samples: 100%|██████████| 1/1 [00:00<00:00,  3.88 Batches/s]

Query: Who was the first person to step foot on the moon?
Answers:
[   <Answer {
    'answer': 'Armstrong', 'type': 'extractive', 'score': 0.9998227059841156, 
    'context': 'The trip to the Moon took just over three days. After achieving orbit, Armstrong and Aldrin transferred into the Lunar Module, named Eagle, and after ', 
    'offsets_in_document': [{'start': 71, 'end': 80}], 
    'offsets_in_context': [{'start': 71, 'end': 80}], 
    'document_id': 'f74e1bf667e68d72e45437a7895df921', 
    'meta': {
        'context': 'The trip to the Moon took just over three days. After achieving orbit, Armstrong and Aldrin transferred into the Lunar Module, named Eagle, and after a landing gear inspection by Collins remaining in the Command/Service Module Columbia, began their descent. After overcoming several computer overload alarms caused by an antenna switch left in the wrong position, and a slight downrange error, Armstrong took over manual flight control at about 180 meters (590 ft), and guided the Lunar Module to a safe landing spot at 20:18:04 UTC, July 20, 1969 (3:17:04 pm CDT). The first humans on the Moon would wait another six hours before they ventured out of their craft. At 02:56 UTC, July 21 (9:56 pm CDT July 20), Armstrong became the first human to set foot on the Moon.', 'title': 'Space_Race'
        }
    }>, <Answer {
    'answer': 'Frank Borman', 'type': 'extractive', 'score': 0.7770257890224457, 
    'context': 'On December 21, 1968, Frank Borman, James Lovell, and William Anders became the first humans to ride the Saturn V rocket into space on Apollo 8. They ', 
    'offsets_in_document': [{'start': 22, 'end': 34}], 
    'offsets_in_context': [{'start': 22, 'end': 34}], 
    'document_id': '2bc046ba90d94fe201ccde9d20552200', 
    'meta': {
        'context': "On December 21, 1968, Frank Borman, James Lovell, and William Anders became the first humans to ride the Saturn V rocket into space on Apollo 8. They also became the first to leave low-Earth orbit and go to another celestial body, and entered lunar orbit on December 24. They made ten orbits in twenty hours, and transmitted one of the most watched TV broadcasts in history, with their Christmas Eve program from lunar orbit, that concluded with a reading from the biblical Book of Genesis. Two and a half hours after the broadcast, they fired their engine to perform the first trans-Earth injection to leave lunar orbit and return to the Earth. Apollo 8 safely landed in the Pacific ocean on December 27, in NASA's first dawn splashdown and recovery.", 'title': 'Space_Race'
        }
    }>, <Answer {
    'answer': 'Aldrin', 'type': 'extractive', 'score': 0.6680101901292801, 
    'context': ' were, "That\'s one small step for [a] man, one giant leap for mankind." Aldrin joined him on the surface almost 20 minutes later. Altogether, they spe', 
    'offsets_in_document': [{'start': 240, 'end': 246}], 
    'offsets_in_context': [{'start': 72, 'end': 78}], 
    'document_id': 'ae1c366b1eaf5fc9d32a8d81f76bd795', 
    'meta': {
        'context': 'The first step was witnessed by at least one-fifth of the population of Earth, or about 723 million people. His first words when he stepped off the LM\'s landing footpad were, "That\'s one small step for [a] man, one giant leap for mankind." Aldrin joined him on the surface almost 20 minutes later. Altogether, they spent just under two and one-quarter hours outside their craft. The next day, they performed the first launch from another celestial body, and rendezvoused back with Columbia.', 'title': 'Space_Race'
        }
    }>
]
```

---

# Source: https://docs.pinecone.io/troubleshooting/how-to-work-with-support.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# How to work with Support

There are several best practices for working with Pinecone Support that can lead to faster resolutions and more relevant recommendations. Please note that Pinecone Support is reserved for users in organizations on the Standard or Enterprise plan. First-response SLAs only apply to tickets created by users in an organization subscribed to a [support plan](https://www.pinecone.io/pricing/?plans=support). To upgrade your support plan, go to [Manage your support plan](https://app.pinecone.io/organizations/-/settings/support/plans) in the console and select your desired plan.

## Utilize Pinecone AI Support

Our [support chatbot](https://app.pinecone.io/organizations/-/settings/support) is knowledgeable of our documentation, troubleshooting articles, website and more. Many of your questions can be answered immediately using this resource. We also review all interactions with the support chatbot and constantly make improvements.

## Use the email associated with your Pinecone account

We map your account information to the tier of your organization to assign appropriate SLAs. If you open tickets using an email not associated with your Pinecone account, we will close your request and suggest alternative contact methods.

## Create tickets using the support portal

Instead of creating tickets via email, use the [Help center](https://app.pinecone.io/organizations/-/settings/support) in the Pinecone console to create tickets. The form allows you to provide helpful information such as severity and category. Furthermore, the conversation format will be much more digestible in the portal, especially when involving code snippets and other attachments.

## Select an appropriate severity

Pinecone Support reserves the right to change the ticket severity after our initial response and assessment of the case. Note that a Sev-1 ticket indicates that your production environment is completely unavailable, and a Sev-2 ticket indicates that your production environment has degraded performance. If your issue does not involve a production-level usage or application, please refrain from opening Sev-1 or Sev-2 tickets.

## Provide the exact names of impacted indexes and projects

When opening a ticket that involves specific resources in your organization, please specify the name of the impacted index(es) and project(s).

## Provide as detailed a description as possible

Please include code snippets, version specifications, and the full stack trace of error messages you encounter. Whenever possible, please include screenshots or screen recordings. The more information you provide, the more likely we can effectively assist you in our first response, and you can return to building with Pinecone.

---

# Source: https://docs.pinecone.io/integrations/hugging-face-inference-endpoints.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Hugging Face Inference Endpoints

> Using Hugging Face Inference Endpoints and Pinecone to generate and index high-quality vector embeddings

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Hugging Face Inference Endpoints offers a secure production solution to easily deploy any Hugging Face Transformers, Sentence-Transformers and Diffusion models from the Hub on dedicated and autoscaling infrastructure managed by Hugging Face.

Coupled with Pinecone, you can use Hugging Face to generate and index high-quality vector embeddings with ease.

<PrimarySecondaryCTA secondaryHref={"#setup-guide"} secondaryLabel={"View setup guide"} />

## Setup guide

Hugging Face Inference Endpoints allows access to straightforward model inference. Coupled with Pinecone we can generate and index high-quality vector embeddings with ease.

Let's get started by initializing an Inference Endpoint for generating vector embeddings.

### Create an endpoint

We start by heading over to the [Hugging Face Inference Endpoints homepage](https://ui.endpoints.huggingface.co/endpoints) and signing up for an account if needed. After, we should find ourselves on this page:

<img src="https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-0.png?fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=193e39d5e65bb87b3b1b1b0d429180b6" alt="endpoints 0" data-og-width="1622" width="1622" data-og-height="936" height="936" data-path="images/hf-endpoints-0.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-0.png?w=280&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=a93797264f78198d1dee10c2a1fe820a 280w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-0.png?w=560&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=8f4799ed2133a717e22e78aa2cf0f245 560w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-0.png?w=840&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=7afce26d262b3f4a30933214e6274ca0 840w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-0.png?w=1100&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=d3e1e36538d3684c13b371cb1f2ede96 1100w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-0.png?w=1650&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=02472896f27429a4f7655faf8abb1c3f 1650w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-0.png?w=2500&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=df1732e3534af7e1ffb4e55dc6b33245 2500w" />

We click on **Create new endpoint**, choose a model repository (eg name of the model), endpoint name (this can be anything), and select a cloud environment. Before moving on it is *very important* that we set the **Task** to **Sentence Embeddings** (found within the *Advanced configuration* settings).

<img src="https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-1.png?fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=322d411596affb620805d391d824c863" alt="endpoints 1" data-og-width="1622" width="1622" data-og-height="1196" height="1196" data-path="images/hf-endpoints-1.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-1.png?w=280&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=094ffc9c6babf9f1fa1060d214063d5d 280w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-1.png?w=560&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=3a794a6a789b7a2082b2147f1d79cc57 560w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-1.png?w=840&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=0556363ec9cb927bc76433c4f7525072 840w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-1.png?w=1100&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=2191bb241e7042f790baac1d5323b18d 1100w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-1.png?w=1650&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=3f1f41f28c98a2c1944d15802cc3ea01 1650w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-1.png?w=2500&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=7ab0e0991b83c12202d8986e8182a292 2500w" />

<img src="https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-2.png?fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=db8b7952c2d42d939b6764cb39f62a7d" alt="endpoints 2" data-og-width="1548" width="1548" data-og-height="354" height="354" data-path="images/hf-endpoints-2.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-2.png?w=280&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=8b8ddd85cddf0b44fdfb12c5d828e200 280w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-2.png?w=560&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=3cf09f91cc295046fd8cdeef7f95f617 560w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-2.png?w=840&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=59fed1c0984ab4fcdb9ed8cd0a14512d 840w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-2.png?w=1100&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=67b3dfd3362eecf6dd0cacaa5988ed8c 1100w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-2.png?w=1650&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=aa7577163d22caa6dd4032a926acd9f4 1650w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-2.png?w=2500&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=d9ce21fa370f79c7d5acd8a9651aa3e7 2500w" />

Other important options include the *Instance Type*, by default this uses CPU which is cheaper but also slower. For faster processing we need a GPU instance. And finally, we set our privacy setting near the end of the page.

After setting our options we can click **Create Endpoint** at the bottom of the page. This action should take use to the next page where we will see the current status of our endpoint.

<img src="https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-3.png?fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=925a8437da312a756edf0a1bdbac16e0" alt="endpoints 3" data-og-width="1548" width="1548" data-og-height="112" height="112" data-path="images/hf-endpoints-3.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-3.png?w=280&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=a08a7a0a590037b0f959fd94f5563006 280w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-3.png?w=560&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=bd7b7970fbd20dae84dc2116dd6647f5 560w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-3.png?w=840&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=6202139573fbda7e1b2cbe71c0ddd7a9 840w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-3.png?w=1100&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=2542fa57778a6a81d564e1e39bf5ba6f 1100w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-3.png?w=1650&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=3ff523c988657e3dbdfcc8bc374a374d 1650w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-3.png?w=2500&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=37ba16c888d5f1dfc1c0ad60480ae57e 2500w" />

Once the status has moved from **Building** to **Running** (this can take some time), we're ready to begin creating embeddings with it.

## Create embeddings

Each endpoint is given an **Endpoint URL**, it can be found on the endpoint **Overview** page. We need to assign this endpoint URL to the `endpoint_url` variable.

<img src="https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-4.png?fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=af3da5b0476bec94258c3cc711fd8b47" alt="endpoints 4" data-og-width="1404" width="1404" data-og-height="222" height="222" data-path="images/hf-endpoints-4.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-4.png?w=280&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=26b6d9ed0620c27e4982f99b6427f3df 280w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-4.png?w=560&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=281c535c7fe5e8fa90f1cb267fd922a6 560w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-4.png?w=840&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=0a037b4ca662219475cacc73d1243262 840w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-4.png?w=1100&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=7df385446234a0f87c199b3748fe6f0b 1100w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-4.png?w=1650&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=eda7a0dd9a7424ab38a6eeb49c5256af 1650w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-4.png?w=2500&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=2bc2d005700830aa2660cfd6afa723a4 2500w" />

```Python Python theme={null}
endpoint = "<ENDPOINT_URL>"
```

We will also need the organization API token, we find this via the organization settings on Hugging Face (`https://huggingface.co/organizations/<ORG_NAME>/settings/profile`). This is assigned to the `api_org` variable.

<img src="https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-5.png?fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=41b16cbb97439082bbd4a4c7e76c662d" alt="endpoints 5" data-og-width="2230" width="2230" data-og-height="1120" height="1120" data-path="images/hf-endpoints-5.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-5.png?w=280&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=81ba23ffdce70a192ee85f22d4cb525f 280w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-5.png?w=560&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=be1170a4cc6f68a9ec56252dfcff73b0 560w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-5.png?w=840&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=e64873c28eb2937436cde6d91a33990a 840w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-5.png?w=1100&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=40ccf67ba795afd1a2b3d1ca7ec39efe 1100w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-5.png?w=1650&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=1d9f9593f03a500ae85698d07e86c5cf 1650w, https://mintcdn.com/pinecone/-4J8tNwjqgTl1iBy/images/hf-endpoints-5.png?w=2500&fit=max&auto=format&n=-4J8tNwjqgTl1iBy&q=85&s=e154eaafeb63a69100d2a73e55edd2ce 2500w" />

```Python Python theme={null}
api_org = "<API_ORG_TOKEN>"
```

Now we're ready to create embeddings via Inference Endpoints. Let's start with a toy example.

```Python Python theme={null}
import requests

# add the api org token to the headers
headers = {
    'Authorization': f'Bearer {api_org}'
}
# we add sentences to embed like so
json_data = {"inputs": ["a happy dog", "a sad dog"]}
# make the request
res = requests.post(
    endpoint,
    headers=headers,
    json=json_data
)
```

We should see a `200` response.

```Python Python theme={null}
res
```

```
<Response [200]>
```

Inside the response we should find two embeddings...

```Python Python theme={null}
len(res.json()['embeddings'])
```

```
2
```

We can also see the dimensionality of our embeddings like so:

```Python Python theme={null}
dim = len(res.json()['embeddings'][0])
dim
```

```
768
```

We will need more than two items to search through, so let's download a larger dataset. For this we will use Hugging Face datasets.

```Python Python theme={null}
from datasets import load_dataset

snli = load_dataset("snli", split='train')
snli
```

```
Downloading: 100%|██████████| 1.93k/1.93k [00:00<00:00, 992kB/s]
Downloading: 100%|██████████| 1.26M/1.26M [00:00<00:00, 31.2MB/s]
Downloading: 100%|██████████| 65.9M/65.9M [00:01<00:00, 57.9MB/s]
Downloading: 100%|██████████| 1.26M/1.26M [00:00<00:00, 43.6MB/s]

Dataset({
    features: ['premise', 'hypothesis', 'label'],
    num_rows: 550152
})
```

SNLI contains 550K sentence pairs, many of these include duplicate items so we will take just one set of these (the *hypothesis*) and deduplicate them.

```Python  theme={null}
passages = list(set(snli['hypothesis']))
len(passages)
```

```
480042
```

We will drop to 50K sentences so that the example is quick to run, if you have time, feel free to keep the full 480K.

```Python Python theme={null}
passages = passages[:50_000]
```

## Create a Pinecone index

With our endpoint and dataset ready, all that we're missing is a vector database. For this, we need to initialize our connection to Pinecone, this requires a [free API key](https://app.pinecone.io/).

```Python Python theme={null}
import pinecone

# initialize connection to pinecone (get API key at app.pinecone.io)
pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")

```

Now we create a new index called `'hf-endpoints'`, the name isn't important *but* the `dimension` must align to our endpoint model output dimensionality (we found this in `dim` above) and the model metric (typically `cosine` is okay, but not for all models).

```Python Python theme={null}
index_name = 'hf-endpoints'

# check if the hf-endpoints index exists
if index_name not in pinecone.list_indexes():
    # create the index if it does not exist
    pinecone.create_index(
        index_name,
        dimension=dim,
        metric="cosine"
    )

# connect to hf-endpoints index we created
index = pinecone.Index(index_name)
```

## Create and index embeddings

Now we have all of our components ready; endpoints, dataset, and Pinecone. Let's go ahead and create our dataset embeddings and index them within Pinecone.

```Python Python theme={null}
from tqdm.auto import tqdm

# we will use batches of 64
batch_size = 64

for i in tqdm(range(0, len(passages), batch_size)):
    # find end of batch
    i_end = min(i+batch_size, len(passages))
    # extract batch
    batch = passages[i:i_end]
    # generate embeddings for batch via endpoints
    res = requests.post(
        endpoint,
        headers=headers,
        json={"inputs": batch}
    )
    emb = res.json()['embeddings']
    # get metadata (just the original text)
    meta = [{'text': text} for text in batch]
    # create IDs
    ids = [str(x) for x in range(i, i_end)]
    # add all to upsert list
    to_upsert = list(zip(ids, emb, meta))
    # upsert/insert these records to pinecone
    _ = index.upsert(vectors=to_upsert)

# check that we have all vectors in index
index.describe_index_stats()
```

```
100%|██████████| 782/782 [11:02<00:00,  1.18it/s]

{'dimension': 768,
 'index_fullness': 0.1,
 'namespaces': {'': {'vector_count': 50000}},
 'total_vector_count': 50000}
```

With everything indexed we can begin querying. We will take a few examples from the *premise* column of the dataset.

```Python Python theme={null}
query = snli['premise'][0]
print(f"Query: {query}")
# encode with HF endpoints
res = requests.post(endpoint, headers=headers, json={"inputs": query})
xq = res.json()['embeddings']
# query and return top 5
xc = index.query(xq, top_k=5, include_metadata=True)
# iterate through results and print text
print("Answers:")
for match in xc['matches']:
    print(match['metadata']['text'])
```

```
Query: A person on a horse jumps over a broken down airplane.
Answers:
The horse jumps over a toy airplane.
a lady rides a horse over a plane shaped obstacle
A person getting onto a horse.
person rides horse
A woman riding a horse jumps over a bar.
```

These look good, let's try a couple more examples.

```Python Python theme={null}
query = snli['premise'][100]
print(f"Query: {query}")
# encode with HF endpoints
res = requests.post(endpoint, headers=headers, json={"inputs": query})
xq = res.json()['embeddings']
# query and return top 5
xc = index.query(xq, top_k=5, include_metadata=True)
# iterate through results and print text
print("Answers:")
for match in xc['matches']:
    print(match['metadata']['text'])
```

```
Query: A woman is walking across the street eating a banana, while a man is following with his briefcase.
Answers:
A woman eats a banana and walks across a street, and there is a man trailing behind her.
A woman eats a banana split.
A woman is carrying two small watermelons and a purse while walking down the street.
The woman walked across the street.
A woman walking on the street with a monkey on her back.
```

And one more...

```Python Python theme={null}
query = snli['premise'][200]
print(f"Query: {query}")
# encode with HF endpoints
res = requests.post(endpoint, headers=headers, json={"inputs": query})
xq = res.json()['embeddings']
# query and return top 5
xc = index.query(xq, top_k=5, include_metadata=True)
# iterate through results and print text
print("Answers:")
for match in xc['matches']:
    print(match['metadata']['text'])
```

```
Query: People on bicycles waiting at an intersection.
Answers:
A pair of people on bikes are waiting at a stoplight.
Bike riders wait to cross the street.
people on bicycles
Group of bike riders stopped in the street.
There are bicycles outside.
```

All of these results look excellent. If you are not planning on running your endpoint and vector DB beyond this tutorial, you can shut down both.

## Clean up

Shut down the endpoint by navigating to the Inference Endpoints **Overview** page and selecting **Delete endpoint**. Delete the Pinecone index with:

```Python Python theme={null}
pinecone.delete_index(index_name)
```

<Note>Once the index is deleted, you cannot use it again.</Note>

---

# Source: https://docs.pinecone.io/guides/search/hybrid-search.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Hybrid search

> Combine semantic and lexical search for better results.

[Semantic search](/guides/search/semantic-search) and [lexical search](/guides/search/lexical-search) are powerful information retrieval techniques, but each has notable limitations. For example:

* Semantic search can miss results based on exact keyword matches, especially in scenarios involving domain-specific terminology.
* Lexical search can miss results based on relationships, such as synonyms and paraphrases.

To work around these limitations, you can use hybrid search, which combines semantic and lexical search.

## Hybrid search approaches

There are two ways to perform hybrid search in Pinecone:

* [Use a single hybrid index](#use-a-single-hybrid-index). This is the **recommended** approach for most use cases because it provides a simpler architecture with less operational overhead.
* [Use separate dense and sparse indexes](#use-separate-dense-and-sparse-indexes). This approach provides more flexibility but requires managing two indexes and maintaining linkages between vectors.

The following table summarizes the key differences between the two approaches:

| Approach                          | Pros                                                                                                                                                                                                                                                                                       | Cons                                                                                                                                                                                                                                                    |
| :-------------------------------- | :----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | :------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| Single hybrid index               | <ul><li>You make requests to only a single index.</li><li>The linkage between dense and sparse vectors is implicit.</li><li>Simpler architecture with less operational overhead.</li></ul>                                                                                                 | <ul><li>You can't do sparse-only queries.</li><li>You can't use integrated embedding and reranking.</li></ul>                                                                                                                                           |
| Separate dense and sparse indexes | <ul><li>You can start with dense for semantic search and add sparse for lexical search later.</li><li>You can do sparse-only queries.</li><li>You can rerank at multiple levels (for each index and for merged results).</li><li>You can use integrated embedding and reranking.</li></ul> | <ul><li>You need to manage and make requests to two separate indexes.</li><li>You need to maintain the linkage between sparse and dense vectors in different indexes.</li><li>More complex architecture with additional operational overhead.</li></ul> |

## Choosing the right approach

**For most use cases, a single hybrid index is recommended.**

* This approach provides a simpler architecture with less operational overhead. You make requests to a single index rather than managing and querying two separate indexes.
* The linkage between dense and sparse vectors is implicit, eliminating the need to maintain explicit linkages between vectors in different indexes.
* You can perform hybrid queries with a single request, reducing latency and complexity compared to querying separate indexes and merging results.

**Consider using separate indexes only when:**

* You need to do sparse-only queries.
* You want to use Pinecone's integrated sparse model ([`pinecone-sparse-english-v0`](/models/pinecone-sparse-english-v0)), which only works with sparse-only indexes.
* You need complete independence in reranking results from each index.
* You require the flexibility to manage dense and sparse indexes independently.

<Note>
  If you're unsure which approach to use, start with a single hybrid index. You can always split indexes later if you have specific requirements that justify the additional operational complexity.
</Note>

## Use a single hybrid index

To perform hybrid search with a single hybrid index, follow these steps:

<Steps>
  <Step title="Create a hybrid index">
    To store both dense and sparse vectors in a single index, use the [`create_index`](/reference/api/latest/control-plane/create_index) operation, setting the `vector_type` to `dense` and the `metric` to `dotproduct`. This is the only combination that supports dense/sparse search on a single index.

    ```python Python theme={null}
    from pinecone.grpc import PineconeGRPC as Pinecone
    from pinecone import ServerlessSpec

    pc = Pinecone(api_key="YOUR_API_KEY")

    index_name = "hybrid-index"

    if not pc.has_index(index_name):
        pc.create_index(
            name=index_name,
            vector_type="dense",
            dimension=1024,
            metric="dotproduct",
            spec=ServerlessSpec(
                cloud="aws",
                region="us-east-1"
            )
        )
    ```
  </Step>

  <Step title="Generate vectors">
    Use Pinecone's [hosted embedding models](/guides/index-data/create-an-index#embedding-models) to [convert data into dense and sparse vectors](/reference/api/latest/inference/generate-embeddings).

    ```python Python [expandable] theme={null}
    # Define the records
    data = [
        { "_id": "vec1", "chunk_text": "Apple Inc. issued a $10 billion corporate bond in 2023." },
        { "_id": "vec2", "chunk_text": "ETFs tracking the S&P 500 outperformed active funds last year." },
        { "_id": "vec3", "chunk_text": "Tesla's options volume surged after the latest earnings report." },
        { "_id": "vec4", "chunk_text": "Dividend aristocrats are known for consistently raising payouts." },
        { "_id": "vec5", "chunk_text": "The Federal Reserve raised interest rates by 0.25% to curb inflation." },
        { "_id": "vec6", "chunk_text": "Unemployment hit a record low of 3.7% in Q4 of 2024." },
        { "_id": "vec7", "chunk_text": "The CPI index rose by 6% in July 2024, raising concerns about purchasing power." },
        { "_id": "vec8", "chunk_text": "GDP growth in emerging markets outpaced developed economies." },
        { "_id": "vec9", "chunk_text": "Amazon's acquisition of MGM Studios was valued at $8.45 billion." },
        { "_id": "vec10", "chunk_text": "Alphabet reported a 20% increase in advertising revenue." },
        { "_id": "vec11", "chunk_text": "ExxonMobil announced a special dividend after record profits." },
        { "_id": "vec12", "chunk_text": "Tesla plans a 3-for-1 stock split to attract retail investors." },
        { "_id": "vec13", "chunk_text": "Credit card APRs reached an all-time high of 22.8% in 2024." },
        { "_id": "vec14", "chunk_text": "A 529 college savings plan offers tax advantages for education." },
        { "_id": "vec15", "chunk_text": "Emergency savings should ideally cover 6 months of expenses." },
        { "_id": "vec16", "chunk_text": "The average mortgage rate rose to 7.1% in December." },
        { "_id": "vec17", "chunk_text": "The SEC fined a hedge fund $50 million for insider trading." },
        { "_id": "vec18", "chunk_text": "New ESG regulations require companies to disclose climate risks." },
        { "_id": "vec19", "chunk_text": "The IRS introduced a new tax bracket for high earners." },
        { "_id": "vec20", "chunk_text": "Compliance with GDPR is mandatory for companies operating in Europe." },
        { "_id": "vec21", "chunk_text": "What are the best-performing green bonds in a rising rate environment?" },
        { "_id": "vec22", "chunk_text": "How does inflation impact the real yield of Treasury bonds?" },
        { "_id": "vec23", "chunk_text": "Top SPAC mergers in the technology sector for 2024." },
        { "_id": "vec24", "chunk_text": "Are stablecoins a viable hedge against currency devaluation?" },
        { "_id": "vec25", "chunk_text": "Comparison of Roth IRA vs 401(k) for high-income earners." },
        { "_id": "vec26", "chunk_text": "Stock splits and their effect on investor sentiment." },
        { "_id": "vec27", "chunk_text": "Tech IPOs that disappointed in their first year." },
        { "_id": "vec28", "chunk_text": "Impact of interest rate hikes on bank stocks." },
        { "_id": "vec29", "chunk_text": "Growth vs. value investing strategies in 2024." },
        { "_id": "vec30", "chunk_text": "The role of artificial intelligence in quantitative trading." },
        { "_id": "vec31", "chunk_text": "What are the implications of quantitative tightening on equities?" },
        { "_id": "vec32", "chunk_text": "How does compounding interest affect long-term investments?" },
        { "_id": "vec33", "chunk_text": "What are the best assets to hedge against inflation?" },
        { "_id": "vec34", "chunk_text": "Can ETFs provide better diversification than mutual funds?" },
        { "_id": "vec35", "chunk_text": "Unemployment hit at 2.4% in Q3 of 2024." },
        { "_id": "vec36", "chunk_text": "Unemployment is expected to hit 2.5% in Q3 of 2024." },
        { "_id": "vec37", "chunk_text": "In Q3 2025 unemployment for the prior year was revised to 2.2%"},
        { "_id": "vec38", "chunk_text": "Emerging markets witnessed increased foreign direct investment as global interest rates stabilized." },
        { "_id": "vec39", "chunk_text": "The rise in energy prices significantly impacted inflation trends during the first half of 2024." },
        { "_id": "vec40", "chunk_text": "Labor market trends show a declining participation rate despite record low unemployment in 2024." },
        { "_id": "vec41", "chunk_text": "Forecasts of global supply chain disruptions eased in late 2024, but consumer prices remained elevated due to persistent demand." },
        { "_id": "vec42", "chunk_text": "Tech sector layoffs in Q3 2024 have reshaped hiring trends across high-growth industries." },
        { "_id": "vec43", "chunk_text": "The U.S. dollar weakened against a basket of currencies as the global economy adjusted to shifting trade balances." },
        { "_id": "vec44", "chunk_text": "Central banks worldwide increased gold reserves to hedge against geopolitical and economic instability." },
        { "_id": "vec45", "chunk_text": "Corporate earnings in Q4 2024 were largely impacted by rising raw material costs and currency fluctuations." },
        { "_id": "vec46", "chunk_text": "Economic recovery in Q2 2024 relied heavily on government spending in infrastructure and green energy projects." },
        { "_id": "vec47", "chunk_text": "The housing market saw a rebound in late 2024, driven by falling mortgage rates and pent-up demand." },
        { "_id": "vec48", "chunk_text": "Wage growth outpaced inflation for the first time in years, signaling improved purchasing power in 2024." },
        { "_id": "vec49", "chunk_text": "China's economic growth in 2024 slowed to its lowest level in decades due to structural reforms and weak exports." },
        { "_id": "vec50", "chunk_text": "AI-driven automation in the manufacturing sector boosted productivity but raised concerns about job displacement." },
        { "_id": "vec51", "chunk_text": "The European Union introduced new fiscal policies in 2024 aimed at reducing public debt without stifling growth." },
        { "_id": "vec52", "chunk_text": "Record-breaking weather events in early 2024 have highlighted the growing economic impact of climate change." },
        { "_id": "vec53", "chunk_text": "Cryptocurrencies faced regulatory scrutiny in 2024, leading to volatility and reduced market capitalization." },
        { "_id": "vec54", "chunk_text": "The global tourism sector showed signs of recovery in late 2024 after years of pandemic-related setbacks." },
        { "_id": "vec55", "chunk_text": "Trade tensions between the U.S. and China escalated in 2024, impacting global supply chains and investment flows." },
        { "_id": "vec56", "chunk_text": "Consumer confidence indices remained resilient in Q2 2024 despite fears of an impending recession." },
        { "_id": "vec57", "chunk_text": "Startups in 2024 faced tighter funding conditions as venture capitalists focused on profitability over growth." },
        { "_id": "vec58", "chunk_text": "Oil production cuts in Q1 2024 by OPEC nations drove prices higher, influencing global energy policies." },
        { "_id": "vec59", "chunk_text": "The adoption of digital currencies by central banks increased in 2024, reshaping monetary policy frameworks." },
        { "_id": "vec60", "chunk_text": "Healthcare spending in 2024 surged as governments expanded access to preventive care and pandemic preparedness." },
        { "_id": "vec61", "chunk_text": "The World Bank reported declining poverty rates globally, but regional disparities persisted." },
        { "_id": "vec62", "chunk_text": "Private equity activity in 2024 focused on renewable energy and technology sectors amid shifting investor priorities." },
        { "_id": "vec63", "chunk_text": "Population aging emerged as a critical economic issue in 2024, especially in advanced economies." },
        { "_id": "vec64", "chunk_text": "Rising commodity prices in 2024 strained emerging markets dependent on imports of raw materials." },
        { "_id": "vec65", "chunk_text": "The global shipping industry experienced declining freight rates in 2024 due to overcapacity and reduced demand." },
        { "_id": "vec66", "chunk_text": "Bank lending to small and medium-sized enterprises surged in 2024 as governments incentivized entrepreneurship." },
        { "_id": "vec67", "chunk_text": "Renewable energy projects accounted for a record share of global infrastructure investment in 2024." },
        { "_id": "vec68", "chunk_text": "Cybersecurity spending reached new highs in 2024, reflecting the growing threat of digital attacks on infrastructure." },
        { "_id": "vec69", "chunk_text": "The agricultural sector faced challenges in 2024 due to extreme weather and rising input costs." },
        { "_id": "vec70", "chunk_text": "Consumer spending patterns shifted in 2024, with a greater focus on experiences over goods." },
        { "_id": "vec71", "chunk_text": "The economic impact of the 2008 financial crisis was mitigated by quantitative easing policies." },
        { "_id": "vec72", "chunk_text": "In early 2024, global GDP growth slowed, driven by weaker exports in Asia and Europe." },
        { "_id": "vec73", "chunk_text": "The historical relationship between inflation and unemployment is explained by the Phillips Curve." },
        { "_id": "vec74", "chunk_text": "The World Trade Organization's role in resolving disputes was tested in 2024." },
        { "_id": "vec75", "chunk_text": "The collapse of Silicon Valley Bank raised questions about regulatory oversight in 2024." },
        { "_id": "vec76", "chunk_text": "The cost of living crisis has been exacerbated by stagnant wage growth and rising inflation." },
        { "_id": "vec77", "chunk_text": "Supply chain resilience became a top priority for multinational corporations in 2024." },
        { "_id": "vec78", "chunk_text": "Consumer sentiment surveys in 2024 reflected optimism despite high interest rates." },
        { "_id": "vec79", "chunk_text": "The resurgence of industrial policy in Q1 2024 focused on decoupling critical supply chains." },
        { "_id": "vec80", "chunk_text": "Technological innovation in the fintech sector disrupted traditional banking in 2024." },
        { "_id": "vec81", "chunk_text": "The link between climate change and migration patterns is increasingly recognized." },
        { "_id": "vec82", "chunk_text": "Renewable energy subsidies in 2024 reduced the global reliance on fossil fuels." },
        { "_id": "vec83", "chunk_text": "The economic fallout of geopolitical tensions was evident in rising defense budgets worldwide." },
        { "_id": "vec84", "chunk_text": "The IMF's 2024 global outlook highlighted risks of stagflation in emerging markets." },
        { "_id": "vec85", "chunk_text": "Declining birth rates in advanced economies pose long-term challenges for labor markets." },
        { "_id": "vec86", "chunk_text": "Digital transformation initiatives in 2024 drove productivity gains in the services sector." },
        { "_id": "vec87", "chunk_text": "The U.S. labor market's resilience in 2024 defied predictions of a severe recession." },
        { "_id": "vec88", "chunk_text": "New fiscal measures in the European Union aimed to stabilize debt levels post-pandemic." },
        { "_id": "vec89", "chunk_text": "Venture capital investments in 2024 leaned heavily toward AI and automation startups." },
        { "_id": "vec90", "chunk_text": "The surge in e-commerce in 2024 was facilitated by advancements in logistics technology." },
        { "_id": "vec91", "chunk_text": "The impact of ESG investing on corporate strategies has been a major focus in 2024." },
        { "_id": "vec92", "chunk_text": "Income inequality widened in 2024 despite strong economic growth in developed nations." },
        { "_id": "vec93", "chunk_text": "The collapse of FTX highlighted the volatility and risks associated with cryptocurrencies." },
        { "_id": "vec94", "chunk_text": "Cyberattacks targeting financial institutions in 2024 led to record cybersecurity spending." },
        { "_id": "vec95", "chunk_text": "Automation in agriculture in 2024 increased yields but displaced rural workers." },
        { "_id": "vec96", "chunk_text": "New trade agreements signed 2022 will make an impact in 2024"},
    ]
    ```

    ```python Python theme={null}
    # Convert the chunk_text into dense vectors
    dense_embeddings = pc.inference.embed(
        model="llama-text-embed-v2",
        inputs=[d['chunk_text'] for d in data],
        parameters={"input_type": "passage", "truncate": "END"}
    )

    # Convert the chunk_text into sparse vectors
    sparse_embeddings = pc.inference.embed(
        model="pinecone-sparse-english-v0",
        inputs=[d['chunk_text'] for d in data],
        parameters={"input_type": "passage", "truncate": "END"}
    )
    ```
  </Step>

  <Step title="Upsert records with dense and sparse vectors">
    Use the [`upsert`](/reference/api/latest/data-plane/upsert) operation, specifying dense values in the `value` parameter and sparse values in the `sparse_values` parameter.

    <Note>
      Only dense indexes using the [dotproduct distance metric](/guides/index-data/indexing-overview#dotproduct) support dense and sparse vectors. Upserting records with dense and sparse vectors into dense indexes with a different distance metric will succeed, but querying will return an error.
    </Note>

    ```Python Python theme={null}
    # Target the hybrid index
    # To get the unique host for an index, 
    # see https://docs.pinecone.io/guides/manage-data/target-an-index
    index = pc.Index(host="INDEX_HOST")

    # Each record contains an ID, a dense vector, a sparse vector, and the original text as metadata
    records = []
    for d, de, se in zip(data, dense_embeddings, sparse_embeddings):
        records.append({
            "id": d['_id'],
            "values": de['values'],
            "sparse_values": {'indices': se['sparse_indices'], 'values': se['sparse_values']},
            "metadata": {'text': d['chunk_text']}
        })

    # Upsert the records into the hybrid index
    index.upsert(
        vectors=records,
        namespace="example-namespace"
    )
    ```
  </Step>

  <Step title="Search the hybrid index">
    Use the [`embed`](/reference/api/latest/inference/generate-embeddings) operation to convert your query into a dense vector and a sparse vector, and then use the [`query`](/reference/api/latest/data-plane/query) operation to search the hybrid index for the 40 most relevant records.

    ```Python Python theme={null}
    query = "Q3 2024 us economic data"

    # Convert the query into a dense vector
    dense_query_embedding = pc.inference.embed(
        model="llama-text-embed-v2",
        inputs=query,
        parameters={"input_type": "query", "truncate": "END"}
    )

    # Convert the query into a sparse vector
    sparse_query_embedding = pc.inference.embed(
        model="pinecone-sparse-english-v0",
        inputs=query,
        parameters={"input_type": "query", "truncate": "END"}
    )

    for d, s in zip(dense_query_embedding, sparse_query_embedding):
        query_response = index.query(
            namespace="example-namespace",
            top_k=40,
            vector=d['values'],
            sparse_vector={'indices': s['sparse_indices'], 'values': s['sparse_values']},
            include_values=False,
            include_metadata=True
        )
        print(query_response)
    ```

    ```python Response [expandable] theme={null}
    {'matches': [{'id': 'vec35',
                  'metadata': {'text': 'Unemployment hit at 2.4% in Q3 of 2024.'},
                  'score': 7.92519569,
                  'values': []},
                 {'id': 'vec46',
                  'metadata': {'text': 'Economic recovery in Q2 2024 relied '
                                       'heavily on government spending in '
                                       'infrastructure and green energy projects.'},
                  'score': 7.86733627,
                  'values': []},
                 {'id': 'vec36',
                  'metadata': {'text': 'Unemployment is expected to hit 2.5% in Q3 '
                                       'of 2024.'},
                  'score': 7.82636,
                  'values': []},
                 {'id': 'vec42',
                  'metadata': {'text': 'Tech sector layoffs in Q3 2024 have '
                                       'reshaped hiring trends across high-growth '
                                       'industries.'},
                  'score': 7.79465914,
                  'values': []},
                 {'id': 'vec49',
                  'metadata': {'text': "China's economic growth in 2024 slowed to "
                                       'its lowest level in decades due to '
                                       'structural reforms and weak exports.'},
                  'score': 7.46323156,
                  'values': []},
                 {'id': 'vec63',
                  'metadata': {'text': 'Population aging emerged as a critical '
                                       'economic issue in 2024, especially in '
                                       'advanced economies.'},
                  'score': 7.29055929,
                  'values': []},
                 {'id': 'vec92',
                  'metadata': {'text': 'Income inequality widened in 2024 despite '
                                       'strong economic growth in developed '
                                       'nations.'},
                  'score': 6.51210213,
                  'values': []},
                 {'id': 'vec52',
                  'metadata': {'text': 'Record-breaking weather events in early '
                                       '2024 have highlighted the growing economic '
                                       'impact of climate change.'},
                  'score': 6.4125514,
                  'values': []},
                 {'id': 'vec62',
                  'metadata': {'text': 'Private equity activity in 2024 focused on '
                                       'renewable energy and technology sectors '
                                       'amid shifting investor priorities.'},
                  'score': 4.8084693,
                  'values': []},
                 {'id': 'vec89',
                  'metadata': {'text': 'Venture capital investments in 2024 leaned '
                                       'heavily toward AI and automation '
                                       'startups.'},
                  'score': 4.7974205,
                  'values': []},
                 {'id': 'vec57',
                  'metadata': {'text': 'Startups in 2024 faced tighter funding '
                                       'conditions as venture capitalists focused '
                                       'on profitability over growth.'},
                  'score': 4.72518444,
                  'values': []},
                 {'id': 'vec37',
                  'metadata': {'text': 'In Q3 2025 unemployment for the prior year '
                                       'was revised to 2.2%'},
                  'score': 4.71824408,
                  'values': []},
                 {'id': 'vec69',
                  'metadata': {'text': 'The agricultural sector faced challenges '
                                       'in 2024 due to extreme weather and rising '
                                       'input costs.'},
                  'score': 4.66726208,
                  'values': []},
                 {'id': 'vec60',
                  'metadata': {'text': 'Healthcare spending in 2024 surged as '
                                       'governments expanded access to preventive '
                                       'care and pandemic preparedness.'},
                  'score': 4.62045908,
                  'values': []},
                 {'id': 'vec55',
                  'metadata': {'text': 'Trade tensions between the U.S. and China '
                                       'escalated in 2024, impacting global supply '
                                       'chains and investment flows.'},
                  'score': 4.59764862,
                  'values': []},
                 {'id': 'vec51',
                  'metadata': {'text': 'The European Union introduced new fiscal '
                                       'policies in 2024 aimed at reducing public '
                                       'debt without stifling growth.'},
                  'score': 4.57397079,
                  'values': []},
                 {'id': 'vec70',
                  'metadata': {'text': 'Consumer spending patterns shifted in '
                                       '2024, with a greater focus on experiences '
                                       'over goods.'},
                  'score': 4.55043507,
                  'values': []},
                 {'id': 'vec87',
                  'metadata': {'text': "The U.S. labor market's resilience in 2024 "
                                       'defied predictions of a severe recession.'},
                  'score': 4.51785707,
                  'values': []},
                 {'id': 'vec90',
                  'metadata': {'text': 'The surge in e-commerce in 2024 was '
                                       'facilitated by advancements in logistics '
                                       'technology.'},
                  'score': 4.47754288,
                  'values': []},
                 {'id': 'vec78',
                  'metadata': {'text': 'Consumer sentiment surveys in 2024 '
                                       'reflected optimism despite high interest '
                                       'rates.'},
                  'score': 4.46246624,
                  'values': []},
                 {'id': 'vec53',
                  'metadata': {'text': 'Cryptocurrencies faced regulatory scrutiny '
                                       'in 2024, leading to volatility and reduced '
                                       'market capitalization.'},
                  'score': 4.4435873,
                  'values': []},
                 {'id': 'vec45',
                  'metadata': {'text': 'Corporate earnings in Q4 2024 were largely '
                                       'impacted by rising raw material costs and '
                                       'currency fluctuations.'},
                  'score': 4.43836403,
                  'values': []},
                 {'id': 'vec82',
                  'metadata': {'text': 'Renewable energy subsidies in 2024 reduced '
                                       'the global reliance on fossil fuels.'},
                  'score': 4.43601322,
                  'values': []},
                 {'id': 'vec94',
                  'metadata': {'text': 'Cyberattacks targeting financial '
                                       'institutions in 2024 led to record '
                                       'cybersecurity spending.'},
                  'score': 4.41334057,
                  'values': []},
                 {'id': 'vec47',
                  'metadata': {'text': 'The housing market saw a rebound in late '
                                       '2024, driven by falling mortgage rates and '
                                       'pent-up demand.'},
                  'score': 4.39900732,
                  'values': []},
                 {'id': 'vec41',
                  'metadata': {'text': 'Forecasts of global supply chain '
                                       'disruptions eased in late 2024, but '
                                       'consumer prices remained elevated due to '
                                       'persistent demand.'},
                  'score': 4.37389421,
                  'values': []},
                 {'id': 'vec84',
                  'metadata': {'text': "The IMF's 2024 global outlook highlighted "
                                       'risks of stagflation in emerging markets.'},
                  'score': 4.37335157,
                  'values': []},
                 {'id': 'vec96',
                  'metadata': {'text': 'New trade agreements signed 2022 will make '
                                       'an impact in 2024'},
                  'score': 4.33860636,
                  'values': []},
                 {'id': 'vec79',
                  'metadata': {'text': 'The resurgence of industrial policy in Q1 '
                                       '2024 focused on decoupling critical supply '
                                       'chains.'},
                  'score': 4.33784199,
                  'values': []},
                 {'id': 'vec6',
                  'metadata': {'text': 'Unemployment hit a record low of 3.7% in '
                                       'Q4 of 2024.'},
                  'score': 4.33008051,
                  'values': []},
                 {'id': 'vec65',
                  'metadata': {'text': 'The global shipping industry experienced '
                                       'declining freight rates in 2024 due to '
                                       'overcapacity and reduced demand.'},
                  'score': 4.3228569,
                  'values': []},
                 {'id': 'vec64',
                  'metadata': {'text': 'Rising commodity prices in 2024 strained '
                                       'emerging markets dependent on imports of '
                                       'raw materials.'},
                  'score': 4.32269621,
                  'values': []},
                 {'id': 'vec95',
                  'metadata': {'text': 'Automation in agriculture in 2024 '
                                       'increased yields but displaced rural '
                                       'workers.'},
                  'score': 4.31127262,
                  'values': []},
                 {'id': 'vec86',
                  'metadata': {'text': 'Digital transformation initiatives in 2024 '
                                       'drove productivity gains in the services '
                                       'sector.'},
                  'score': 4.30181122,
                  'values': []},
                 {'id': 'vec66',
                  'metadata': {'text': 'Bank lending to small and medium-sized '
                                       'enterprises surged in 2024 as governments '
                                       'incentivized entrepreneurship.'},
                  'score': 4.27241945,
                  'values': []},
                 {'id': 'vec58',
                  'metadata': {'text': 'Oil production cuts in Q1 2024 by OPEC '
                                       'nations drove prices higher, influencing '
                                       'global energy policies.'},
                  'score': 4.21715498,
                  'values': []},
                 {'id': 'vec80',
                  'metadata': {'text': 'Technological innovation in the fintech '
                                       'sector disrupted traditional banking in '
                                       '2024.'},
                  'score': 4.17712116,
                  'values': []},
                 {'id': 'vec75',
                  'metadata': {'text': 'The collapse of Silicon Valley Bank raised '
                                       'questions about regulatory oversight in '
                                       '2024.'},
                  'score': 4.16192341,
                  'values': []},
                 {'id': 'vec56',
                  'metadata': {'text': 'Consumer confidence indices remained '
                                       'resilient in Q2 2024 despite fears of an '
                                       'impending recession.'},
                  'score': 4.15782213,
                  'values': []},
                 {'id': 'vec67',
                  'metadata': {'text': 'Renewable energy projects accounted for a '
                                       'record share of global infrastructure '
                                       'investment in 2024.'},
                  'score': 4.14623,
                  'values': []}],
     'namespace': 'example-namespace',
     'usage': {'read_units': 9}}
    ```
  </Step>

  <Step title="Search the hybrid index with explicit weighting">
    Because Pinecone views your sparse-dense vector as a single vector, it does not offer a built-in parameter to adjust the weight of a query's dense part against its sparse part; the index is agnostic to density or sparsity of coordinates in your vectors. You may, however, incorporate a linear weighting scheme by customizing your query vector, as demonstrated in the function below.

    The following example transforms vector values using an alpha parameter.

    ```Python Python theme={null}
    def hybrid_score_norm(dense, sparse, alpha: float):
        """Hybrid score using a convex combination

        alpha * dense + (1 - alpha) * sparse

        Args:
            dense: Array of floats representing
            sparse: a dict of `indices` and `values`
            alpha: scale between 0 and 1
        """
        if alpha < 0 or alpha > 1:
            raise ValueError("Alpha must be between 0 and 1")
        hs = {
            'indices': sparse['indices'],
            'values':  [v * (1 - alpha) for v in sparse['values']]
        }
        return [v * alpha for v in dense], hs
    ```

    The following example transforms a vector using the above function, then queries a Pinecone index.

    ```Python Python theme={null}
    sparse_vector = {
       'indices': [10, 45, 16],
       'values':  [0.5, 0.5, 0.2]
    }
    dense_vector = [0.1, 0.2, 0.3]

    hdense, hsparse = hybrid_score_norm(dense_vector, sparse_vector, alpha=0.75)

    query_response = index.query(
        namespace="example-namespace",
        top_k=10,
        vector=hdense,
        sparse_vector=hsparse
    )
    ```
  </Step>
</Steps>

## Use separate dense and sparse indexes

To perform hybrid search with separate indexes, follow these steps:

<Steps>
  <Step title="Create dense and sparse indexes">
    [Create a dense index](/guides/index-data/create-an-index#create-a-dense-index) and [create a sparse index](/guides/index-data/create-an-index#create-a-sparse-index), either with integrated embedding or for vectors created with external models.

    For example, the following code creates indexes with integrated embedding models.

    ```python Python theme={null}
    from pinecone import Pinecone

    pc = Pinecone(api_key="YOUR_API_KEY")

    dense_index_name = "dense-for-hybrid-py"
    sparse_index_name = "sparse-for-hybrid-py"

    if not pc.has_index(dense_index_name):
        pc.create_index_for_model(
            name=dense_index_name,
            cloud="aws",
            region="us-east-1",
            embed={
                "model":"llama-text-embed-v2",
                "field_map":{"text": "chunk_text"}
            }
        )

    if not pc.has_index(sparse_index_name):
        pc.create_index_for_model(
            name=sparse_index_name,
            cloud="aws",
            region="us-east-1",
            embed={
                "model":"pinecone-sparse-english-v0",
                "field_map":{"text": "chunk_text"}
            }
        )
    ```
  </Step>

  <Step title="Upsert dense and sparse vectors">
    [Upsert dense vectors](/guides/index-data/upsert-data#upsert-dense-vectors) into the dense index and [upsert sparse vectors](/guides/index-data/upsert-data#upsert-sparse-vectors) into the sparse index.

    Make sure to establish a linkage between the dense and sparse vectors so you can merge and deduplicate search results later. For example, the following uses `_id` as the linkage, but you can use any other custom field as well. Because the indexes are integrated with embedding models, you provide the source texts and Pinecone converts them to vectors automatically.

    ```python Python [expandable] theme={null}
    # Define the records
    records = [
        { "_id": "vec1", "chunk_text": "Apple Inc. issued a $10 billion corporate bond in 2023." },
        { "_id": "vec2", "chunk_text": "ETFs tracking the S&P 500 outperformed active funds last year." },
        { "_id": "vec3", "chunk_text": "Tesla's options volume surged after the latest earnings report." },
        { "_id": "vec4", "chunk_text": "Dividend aristocrats are known for consistently raising payouts." },
        { "_id": "vec5", "chunk_text": "The Federal Reserve raised interest rates by 0.25% to curb inflation." },
        { "_id": "vec6", "chunk_text": "Unemployment hit a record low of 3.7% in Q4 of 2024." },
        { "_id": "vec7", "chunk_text": "The CPI index rose by 6% in July 2024, raising concerns about purchasing power." },
        { "_id": "vec8", "chunk_text": "GDP growth in emerging markets outpaced developed economies." },
        { "_id": "vec9", "chunk_text": "Amazon's acquisition of MGM Studios was valued at $8.45 billion." },
        { "_id": "vec10", "chunk_text": "Alphabet reported a 20% increase in advertising revenue." },
        { "_id": "vec11", "chunk_text": "ExxonMobil announced a special dividend after record profits." },
        { "_id": "vec12", "chunk_text": "Tesla plans a 3-for-1 stock split to attract retail investors." },
        { "_id": "vec13", "chunk_text": "Credit card APRs reached an all-time high of 22.8% in 2024." },
        { "_id": "vec14", "chunk_text": "A 529 college savings plan offers tax advantages for education." },
        { "_id": "vec15", "chunk_text": "Emergency savings should ideally cover 6 months of expenses." },
        { "_id": "vec16", "chunk_text": "The average mortgage rate rose to 7.1% in December." },
        { "_id": "vec17", "chunk_text": "The SEC fined a hedge fund $50 million for insider trading." },
        { "_id": "vec18", "chunk_text": "New ESG regulations require companies to disclose climate risks." },
        { "_id": "vec19", "chunk_text": "The IRS introduced a new tax bracket for high earners." },
        { "_id": "vec20", "chunk_text": "Compliance with GDPR is mandatory for companies operating in Europe." },
        { "_id": "vec21", "chunk_text": "What are the best-performing green bonds in a rising rate environment?" },
        { "_id": "vec22", "chunk_text": "How does inflation impact the real yield of Treasury bonds?" },
        { "_id": "vec23", "chunk_text": "Top SPAC mergers in the technology sector for 2024." },
        { "_id": "vec24", "chunk_text": "Are stablecoins a viable hedge against currency devaluation?" },
        { "_id": "vec25", "chunk_text": "Comparison of Roth IRA vs 401(k) for high-income earners." },
        { "_id": "vec26", "chunk_text": "Stock splits and their effect on investor sentiment." },
        { "_id": "vec27", "chunk_text": "Tech IPOs that disappointed in their first year." },
        { "_id": "vec28", "chunk_text": "Impact of interest rate hikes on bank stocks." },
        { "_id": "vec29", "chunk_text": "Growth vs. value investing strategies in 2024." },
        { "_id": "vec30", "chunk_text": "The role of artificial intelligence in quantitative trading." },
        { "_id": "vec31", "chunk_text": "What are the implications of quantitative tightening on equities?" },
        { "_id": "vec32", "chunk_text": "How does compounding interest affect long-term investments?" },
        { "_id": "vec33", "chunk_text": "What are the best assets to hedge against inflation?" },
        { "_id": "vec34", "chunk_text": "Can ETFs provide better diversification than mutual funds?" },
        { "_id": "vec35", "chunk_text": "Unemployment hit at 2.4% in Q3 of 2024." },
        { "_id": "vec36", "chunk_text": "Unemployment is expected to hit 2.5% in Q3 of 2024." },
        { "_id": "vec37", "chunk_text": "In Q3 2025 unemployment for the prior year was revised to 2.2%"},
        { "_id": "vec38", "chunk_text": "Emerging markets witnessed increased foreign direct investment as global interest rates stabilized." },
        { "_id": "vec39", "chunk_text": "The rise in energy prices significantly impacted inflation trends during the first half of 2024." },
        { "_id": "vec40", "chunk_text": "Labor market trends show a declining participation rate despite record low unemployment in 2024." },
        { "_id": "vec41", "chunk_text": "Forecasts of global supply chain disruptions eased in late 2024, but consumer prices remained elevated due to persistent demand." },
        { "_id": "vec42", "chunk_text": "Tech sector layoffs in Q3 2024 have reshaped hiring trends across high-growth industries." },
        { "_id": "vec43", "chunk_text": "The U.S. dollar weakened against a basket of currencies as the global economy adjusted to shifting trade balances." },
        { "_id": "vec44", "chunk_text": "Central banks worldwide increased gold reserves to hedge against geopolitical and economic instability." },
        { "_id": "vec45", "chunk_text": "Corporate earnings in Q4 2024 were largely impacted by rising raw material costs and currency fluctuations." },
        { "_id": "vec46", "chunk_text": "Economic recovery in Q2 2024 relied heavily on government spending in infrastructure and green energy projects." },
        { "_id": "vec47", "chunk_text": "The housing market saw a rebound in late 2024, driven by falling mortgage rates and pent-up demand." },
        { "_id": "vec48", "chunk_text": "Wage growth outpaced inflation for the first time in years, signaling improved purchasing power in 2024." },
        { "_id": "vec49", "chunk_text": "China's economic growth in 2024 slowed to its lowest level in decades due to structural reforms and weak exports." },
        { "_id": "vec50", "chunk_text": "AI-driven automation in the manufacturing sector boosted productivity but raised concerns about job displacement." },
        { "_id": "vec51", "chunk_text": "The European Union introduced new fiscal policies in 2024 aimed at reducing public debt without stifling growth." },
        { "_id": "vec52", "chunk_text": "Record-breaking weather events in early 2024 have highlighted the growing economic impact of climate change." },
        { "_id": "vec53", "chunk_text": "Cryptocurrencies faced regulatory scrutiny in 2024, leading to volatility and reduced market capitalization." },
        { "_id": "vec54", "chunk_text": "The global tourism sector showed signs of recovery in late 2024 after years of pandemic-related setbacks." },
        { "_id": "vec55", "chunk_text": "Trade tensions between the U.S. and China escalated in 2024, impacting global supply chains and investment flows." },
        { "_id": "vec56", "chunk_text": "Consumer confidence indices remained resilient in Q2 2024 despite fears of an impending recession." },
        { "_id": "vec57", "chunk_text": "Startups in 2024 faced tighter funding conditions as venture capitalists focused on profitability over growth." },
        { "_id": "vec58", "chunk_text": "Oil production cuts in Q1 2024 by OPEC nations drove prices higher, influencing global energy policies." },
        { "_id": "vec59", "chunk_text": "The adoption of digital currencies by central banks increased in 2024, reshaping monetary policy frameworks." },
        { "_id": "vec60", "chunk_text": "Healthcare spending in 2024 surged as governments expanded access to preventive care and pandemic preparedness." },
        { "_id": "vec61", "chunk_text": "The World Bank reported declining poverty rates globally, but regional disparities persisted." },
        { "_id": "vec62", "chunk_text": "Private equity activity in 2024 focused on renewable energy and technology sectors amid shifting investor priorities." },
        { "_id": "vec63", "chunk_text": "Population aging emerged as a critical economic issue in 2024, especially in advanced economies." },
        { "_id": "vec64", "chunk_text": "Rising commodity prices in 2024 strained emerging markets dependent on imports of raw materials." },
        { "_id": "vec65", "chunk_text": "The global shipping industry experienced declining freight rates in 2024 due to overcapacity and reduced demand." },
        { "_id": "vec66", "chunk_text": "Bank lending to small and medium-sized enterprises surged in 2024 as governments incentivized entrepreneurship." },
        { "_id": "vec67", "chunk_text": "Renewable energy projects accounted for a record share of global infrastructure investment in 2024." },
        { "_id": "vec68", "chunk_text": "Cybersecurity spending reached new highs in 2024, reflecting the growing threat of digital attacks on infrastructure." },
        { "_id": "vec69", "chunk_text": "The agricultural sector faced challenges in 2024 due to extreme weather and rising input costs." },
        { "_id": "vec70", "chunk_text": "Consumer spending patterns shifted in 2024, with a greater focus on experiences over goods." },
        { "_id": "vec71", "chunk_text": "The economic impact of the 2008 financial crisis was mitigated by quantitative easing policies." },
        { "_id": "vec72", "chunk_text": "In early 2024, global GDP growth slowed, driven by weaker exports in Asia and Europe." },
        { "_id": "vec73", "chunk_text": "The historical relationship between inflation and unemployment is explained by the Phillips Curve." },
        { "_id": "vec74", "chunk_text": "The World Trade Organization's role in resolving disputes was tested in 2024." },
        { "_id": "vec75", "chunk_text": "The collapse of Silicon Valley Bank raised questions about regulatory oversight in 2024." },
        { "_id": "vec76", "chunk_text": "The cost of living crisis has been exacerbated by stagnant wage growth and rising inflation." },
        { "_id": "vec77", "chunk_text": "Supply chain resilience became a top priority for multinational corporations in 2024." },
        { "_id": "vec78", "chunk_text": "Consumer sentiment surveys in 2024 reflected optimism despite high interest rates." },
        { "_id": "vec79", "chunk_text": "The resurgence of industrial policy in Q1 2024 focused on decoupling critical supply chains." },
        { "_id": "vec80", "chunk_text": "Technological innovation in the fintech sector disrupted traditional banking in 2024." },
        { "_id": "vec81", "chunk_text": "The link between climate change and migration patterns is increasingly recognized." },
        { "_id": "vec82", "chunk_text": "Renewable energy subsidies in 2024 reduced the global reliance on fossil fuels." },
        { "_id": "vec83", "chunk_text": "The economic fallout of geopolitical tensions was evident in rising defense budgets worldwide." },
        { "_id": "vec84", "chunk_text": "The IMF's 2024 global outlook highlighted risks of stagflation in emerging markets." },
        { "_id": "vec85", "chunk_text": "Declining birth rates in advanced economies pose long-term challenges for labor markets." },
        { "_id": "vec86", "chunk_text": "Digital transformation initiatives in 2024 drove productivity gains in the services sector." },
        { "_id": "vec87", "chunk_text": "The U.S. labor market's resilience in 2024 defied predictions of a severe recession." },
        { "_id": "vec88", "chunk_text": "New fiscal measures in the European Union aimed to stabilize debt levels post-pandemic." },
        { "_id": "vec89", "chunk_text": "Venture capital investments in 2024 leaned heavily toward AI and automation startups." },
        { "_id": "vec90", "chunk_text": "The surge in e-commerce in 2024 was facilitated by advancements in logistics technology." },
        { "_id": "vec91", "chunk_text": "The impact of ESG investing on corporate strategies has been a major focus in 2024." },
        { "_id": "vec92", "chunk_text": "Income inequality widened in 2024 despite strong economic growth in developed nations." },
        { "_id": "vec93", "chunk_text": "The collapse of FTX highlighted the volatility and risks associated with cryptocurrencies." },
        { "_id": "vec94", "chunk_text": "Cyberattacks targeting financial institutions in 2024 led to record cybersecurity spending." },
        { "_id": "vec95", "chunk_text": "Automation in agriculture in 2024 increased yields but displaced rural workers." },
        { "_id": "vec96", "chunk_text": "New trade agreements signed 2022 will make an impact in 2024"},
    ]
    ```

    ```python Python theme={null}
    # Target the dense and sparse indexes
    # To get the unique host for an index, 
    # see https://docs.pinecone.io/guides/manage-data/target-an-index
    dense_index = pc.Index(host="INDEX_HOST")
    sparse_index = pc.Index(host="INDEX_HOST")

    # Upsert the records
    # The `chunk_text` fields are converted to dense and sparse vectors
    dense_index.upsert_records("example-namespace", records)
    sparse_index.upsert_records("example-namespace", records)
    ```
  </Step>

  <Step title="Search the dense index">
    Perform a [semantic search](/guides/search/semantic-search) on the dense index.

    For example, the following code searches the dense index for 40 records most semantically related to the query "Q3 2024 us economic data". Because the index is integrated with an embedding model, you provide the query as text and Pinecone converts the text to a dense vector automatically.

    ```python Python theme={null}
    query = "Q3 2024 us economic data"

    dense_results = dense_index.search(
        namespace="example-namespace",
        query={
            "top_k": 40,
            "inputs": {
                "text": query
            }
        }
    )

    print(dense_results)
    ```

    ```python Response [expandable] theme={null}
    {'result': {'hits': [{'_id': 'vec35',
                          '_score': 0.8629686832427979,
                          'fields': {'chunk_text': 'Unemployment hit at 2.4% in Q3 '
                                                   'of 2024.'}},
                         {'_id': 'vec36',
                          '_score': 0.8573639988899231,
                          'fields': {'chunk_text': 'Unemployment is expected to '
                                                   'hit 2.5% in Q3 of 2024.'}},
                         {'_id': 'vec6',
                          '_score': 0.8535352945327759,
                          'fields': {'chunk_text': 'Unemployment hit a record low '
                                                   'of 3.7% in Q4 of 2024.'}},
                         {'_id': 'vec42',
                          '_score': 0.8336166739463806,
                          'fields': {'chunk_text': 'Tech sector layoffs in Q3 2024 '
                                                   'have reshaped hiring trends '
                                                   'across high-growth '
                                                   'industries.'}},
                         {'_id': 'vec48',
                          '_score': 0.8328524827957153,
                          'fields': {'chunk_text': 'Wage growth outpaced inflation '
                                                   'for the first time in years, '
                                                   'signaling improved purchasing '
                                                   'power in 2024.'}},
                         {'_id': 'vec55',
                          '_score': 0.8322604298591614,
                          'fields': {'chunk_text': 'Trade tensions between the '
                                                   'U.S. and China escalated in '
                                                   '2024, impacting global supply '
                                                   'chains and investment flows.'}},
                         {'_id': 'vec45',
                          '_score': 0.8309446573257446,
                          'fields': {'chunk_text': 'Corporate earnings in Q4 2024 '
                                                   'were largely impacted by '
                                                   'rising raw material costs and '
                                                   'currency fluctuations.'}},
                         {'_id': 'vec72',
                          '_score': 0.8275909423828125,
                          'fields': {'chunk_text': 'In early 2024, global GDP '
                                                   'growth slowed, driven by '
                                                   'weaker exports in Asia and '
                                                   'Europe.'}},
                         {'_id': 'vec29',
                          '_score': 0.8270887136459351,
                          'fields': {'chunk_text': 'Growth vs. value investing '
                                                   'strategies in 2024.'}},
                         {'_id': 'vec46',
                          '_score': 0.8263787627220154,
                          'fields': {'chunk_text': 'Economic recovery in Q2 2024 '
                                                   'relied heavily on government '
                                                   'spending in infrastructure and '
                                                   'green energy projects.'}},
                         {'_id': 'vec79',
                          '_score': 0.8258304595947266,
                          'fields': {'chunk_text': 'The resurgence of industrial '
                                                   'policy in Q1 2024 focused on '
                                                   'decoupling critical supply '
                                                   'chains.'}},
                         {'_id': 'vec87',
                          '_score': 0.8257324695587158,
                          'fields': {'chunk_text': "The U.S. labor market's "
                                                   'resilience in 2024 defied '
                                                   'predictions of a severe '
                                                   'recession.'}},
                         {'_id': 'vec40',
                          '_score': 0.8253997564315796,
                          'fields': {'chunk_text': 'Labor market trends show a '
                                                   'declining participation rate '
                                                   'despite record low '
                                                   'unemployment in 2024.'}},
                         {'_id': 'vec37',
                          '_score': 0.8235862255096436,
                          'fields': {'chunk_text': 'In Q3 2025 unemployment for '
                                                   'the prior year was revised to '
                                                   '2.2%'}},
                         {'_id': 'vec58',
                          '_score': 0.8233317136764526,
                          'fields': {'chunk_text': 'Oil production cuts in Q1 2024 '
                                                   'by OPEC nations drove prices '
                                                   'higher, influencing global '
                                                   'energy policies.'}},
                         {'_id': 'vec47',
                          '_score': 0.8231339454650879,
                          'fields': {'chunk_text': 'The housing market saw a '
                                                   'rebound in late 2024, driven '
                                                   'by falling mortgage rates and '
                                                   'pent-up demand.'}},
                         {'_id': 'vec41',
                          '_score': 0.8187897801399231,
                          'fields': {'chunk_text': 'Forecasts of global supply '
                                                   'chain disruptions eased in '
                                                   'late 2024, but consumer prices '
                                                   'remained elevated due to '
                                                   'persistent demand.'}},
                         {'_id': 'vec56',
                          '_score': 0.8155254125595093,
                          'fields': {'chunk_text': 'Consumer confidence indices '
                                                   'remained resilient in Q2 2024 '
                                                   'despite fears of an impending '
                                                   'recession.'}},
                         {'_id': 'vec63',
                          '_score': 0.8136948347091675,
                          'fields': {'chunk_text': 'Population aging emerged as a '
                                                   'critical economic issue in '
                                                   '2024, especially in advanced '
                                                   'economies.'}},
                         {'_id': 'vec52',
                          '_score': 0.8129132390022278,
                          'fields': {'chunk_text': 'Record-breaking weather events '
                                                   'in early 2024 have highlighted '
                                                   'the growing economic impact of '
                                                   'climate change.'}},
                         {'_id': 'vec23',
                          '_score': 0.8126378655433655,
                          'fields': {'chunk_text': 'Top SPAC mergers in the '
                                                   'technology sector for 2024.'}},
                         {'_id': 'vec62',
                          '_score': 0.8116977214813232,
                          'fields': {'chunk_text': 'Private equity activity in '
                                                   '2024 focused on renewable '
                                                   'energy and technology sectors '
                                                   'amid shifting investor '
                                                   'priorities.'}},
                         {'_id': 'vec64',
                          '_score': 0.8109902739524841,
                          'fields': {'chunk_text': 'Rising commodity prices in '
                                                   '2024 strained emerging markets '
                                                   'dependent on imports of raw '
                                                   'materials.'}},
                         {'_id': 'vec54',
                          '_score': 0.8092231154441833,
                          'fields': {'chunk_text': 'The global tourism sector '
                                                   'showed signs of recovery in '
                                                   'late 2024 after years of '
                                                   'pandemic-related setbacks.'}},
                         {'_id': 'vec96',
                          '_score': 0.8075559735298157,
                          'fields': {'chunk_text': 'New trade agreements signed '
                                                   '2022 will make an impact in '
                                                   '2024'}},
                         {'_id': 'vec49',
                          '_score': 0.8062589764595032,
                          'fields': {'chunk_text': "China's economic growth in "
                                                   '2024 slowed to its lowest '
                                                   'level in decades due to '
                                                   'structural reforms and weak '
                                                   'exports.'}},
                         {'_id': 'vec7',
                          '_score': 0.8034461140632629,
                          'fields': {'chunk_text': 'The CPI index rose by 6% in '
                                                   'July 2024, raising concerns '
                                                   'about purchasing power.'}},
                         {'_id': 'vec84',
                          '_score': 0.8027160167694092,
                          'fields': {'chunk_text': "The IMF's 2024 global outlook "
                                                   'highlighted risks of '
                                                   'stagflation in emerging '
                                                   'markets.'}},
                         {'_id': 'vec13',
                          '_score': 0.8010239601135254,
                          'fields': {'chunk_text': 'Credit card APRs reached an '
                                                   'all-time high of 22.8% in '
                                                   '2024.'}},
                         {'_id': 'vec53',
                          '_score': 0.8007135391235352,
                          'fields': {'chunk_text': 'Cryptocurrencies faced '
                                                   'regulatory scrutiny in 2024, '
                                                   'leading to volatility and '
                                                   'reduced market '
                                                   'capitalization.'}},
                         {'_id': 'vec60',
                          '_score': 0.7980866432189941,
                          'fields': {'chunk_text': 'Healthcare spending in 2024 '
                                                   'surged as governments expanded '
                                                   'access to preventive care and '
                                                   'pandemic preparedness.'}},
                         {'_id': 'vec91',
                          '_score': 0.7980680465698242,
                          'fields': {'chunk_text': 'The impact of ESG investing on '
                                                   'corporate strategies has been '
                                                   'a major focus in 2024.'}},
                         {'_id': 'vec68',
                          '_score': 0.797269880771637,
                          'fields': {'chunk_text': 'Cybersecurity spending reached '
                                                   'new highs in 2024, reflecting '
                                                   'the growing threat of digital '
                                                   'attacks on infrastructure.'}},
                         {'_id': 'vec59',
                          '_score': 0.795337438583374,
                          'fields': {'chunk_text': 'The adoption of digital '
                                                   'currencies by central banks '
                                                   'increased in 2024, reshaping '
                                                   'monetary policy frameworks.'}},
                         {'_id': 'vec39',
                          '_score': 0.793889045715332,
                          'fields': {'chunk_text': 'The rise in energy prices '
                                                   'significantly impacted '
                                                   'inflation trends during the '
                                                   'first half of 2024.'}},
                         {'_id': 'vec66',
                          '_score': 0.7919396162033081,
                          'fields': {'chunk_text': 'Bank lending to small and '
                                                   'medium-sized enterprises '
                                                   'surged in 2024 as governments '
                                                   'incentivized '
                                                   'entrepreneurship.'}},
                         {'_id': 'vec57',
                          '_score': 0.7917722463607788,
                          'fields': {'chunk_text': 'Startups in 2024 faced tighter '
                                                   'funding conditions as venture '
                                                   'capitalists focused on '
                                                   'profitability over growth.'}},
                         {'_id': 'vec75',
                          '_score': 0.7907494306564331,
                          'fields': {'chunk_text': 'The collapse of Silicon Valley '
                                                   'Bank raised questions about '
                                                   'regulatory oversight in '
                                                   '2024.'}},
                         {'_id': 'vec51',
                          '_score': 0.790622889995575,
                          'fields': {'chunk_text': 'The European Union introduced '
                                                   'new fiscal policies in 2024 '
                                                   'aimed at reducing public debt '
                                                   'without stifling growth.'}},
                         {'_id': 'vec89',
                          '_score': 0.7899052500724792,
                          'fields': {'chunk_text': 'Venture capital investments in '
                                                   '2024 leaned heavily toward AI '
                                                   'and automation startups.'}}]},
     'usage': {'embed_total_tokens': 12, 'read_units': 1}}
    ```
  </Step>

  <Step title="Search the sparse index">
    Perform a [lexical search](/guides/search/lexical-search).

    For example, the following code searches the sparse index for 40 records that most exactly match the words in the query. Again, because the index is integrated with an embedding model, you provide the query as text and Pinecone converts the text to a sparse vector automatically.

    ```python Python theme={null}
    sparse_results = sparse_index.search(
        namespace="example-namespace",
        query={
            "top_k": 40,
            "inputs": {
                "text": query
            }
        }
    )

    print(sparse_results)
    ```

    ```python Response [expandable] theme={null}
    {'result': {'hits': [{'_id': 'vec35',
                          '_score': 7.0625,
                          'fields': {'chunk_text': 'Unemployment hit at 2.4% in Q3 '
                                                   'of 2024.'}},
                         {'_id': 'vec46',
                          '_score': 7.041015625,
                          'fields': {'chunk_text': 'Economic recovery in Q2 2024 '
                                                   'relied heavily on government '
                                                   'spending in infrastructure and '
                                                   'green energy projects.'}},
                         {'_id': 'vec36',
                          '_score': 6.96875,
                          'fields': {'chunk_text': 'Unemployment is expected to '
                                                   'hit 2.5% in Q3 of 2024.'}},
                         {'_id': 'vec42',
                          '_score': 6.9609375,
                          'fields': {'chunk_text': 'Tech sector layoffs in Q3 2024 '
                                                   'have reshaped hiring trends '
                                                   'across high-growth '
                                                   'industries.'}},
                         {'_id': 'vec49',
                          '_score': 6.65625,
                          'fields': {'chunk_text': "China's economic growth in "
                                                   '2024 slowed to its lowest '
                                                   'level in decades due to '
                                                   'structural reforms and weak '
                                                   'exports.'}},
                         {'_id': 'vec63',
                          '_score': 6.4765625,
                          'fields': {'chunk_text': 'Population aging emerged as a '
                                                   'critical economic issue in '
                                                   '2024, especially in advanced '
                                                   'economies.'}},
                         {'_id': 'vec92',
                          '_score': 5.72265625,
                          'fields': {'chunk_text': 'Income inequality widened in '
                                                   '2024 despite strong economic '
                                                   'growth in developed nations.'}},
                         {'_id': 'vec52',
                          '_score': 5.599609375,
                          'fields': {'chunk_text': 'Record-breaking weather events '
                                                   'in early 2024 have highlighted '
                                                   'the growing economic impact of '
                                                   'climate change.'}},
                         {'_id': 'vec89',
                          '_score': 4.0078125,
                          'fields': {'chunk_text': 'Venture capital investments in '
                                                   '2024 leaned heavily toward AI '
                                                   'and automation startups.'}},
                         {'_id': 'vec62',
                          '_score': 3.99609375,
                          'fields': {'chunk_text': 'Private equity activity in '
                                                   '2024 focused on renewable '
                                                   'energy and technology sectors '
                                                   'amid shifting investor '
                                                   'priorities.'}},
                         {'_id': 'vec57',
                          '_score': 3.93359375,
                          'fields': {'chunk_text': 'Startups in 2024 faced tighter '
                                                   'funding conditions as venture '
                                                   'capitalists focused on '
                                                   'profitability over growth.'}},
                         {'_id': 'vec69',
                          '_score': 3.8984375,
                          'fields': {'chunk_text': 'The agricultural sector faced '
                                                   'challenges in 2024 due to '
                                                   'extreme weather and rising '
                                                   'input costs.'}},
                         {'_id': 'vec37',
                          '_score': 3.89453125,
                          'fields': {'chunk_text': 'In Q3 2025 unemployment for '
                                                   'the prior year was revised to '
                                                   '2.2%'}},
                         {'_id': 'vec60',
                          '_score': 3.822265625,
                          'fields': {'chunk_text': 'Healthcare spending in 2024 '
                                                   'surged as governments expanded '
                                                   'access to preventive care and '
                                                   'pandemic preparedness.'}},
                         {'_id': 'vec51',
                          '_score': 3.783203125,
                          'fields': {'chunk_text': 'The European Union introduced '
                                                   'new fiscal policies in 2024 '
                                                   'aimed at reducing public debt '
                                                   'without stifling growth.'}},
                         {'_id': 'vec55',
                          '_score': 3.765625,
                          'fields': {'chunk_text': 'Trade tensions between the '
                                                   'U.S. and China escalated in '
                                                   '2024, impacting global supply '
                                                   'chains and investment flows.'}},
                         {'_id': 'vec70',
                          '_score': 3.76171875,
                          'fields': {'chunk_text': 'Consumer spending patterns '
                                                   'shifted in 2024, with a '
                                                   'greater focus on experiences '
                                                   'over goods.'}},
                         {'_id': 'vec90',
                          '_score': 3.70703125,
                          'fields': {'chunk_text': 'The surge in e-commerce in '
                                                   '2024 was facilitated by '
                                                   'advancements in logistics '
                                                   'technology.'}},
                         {'_id': 'vec87',
                          '_score': 3.69140625,
                          'fields': {'chunk_text': "The U.S. labor market's "
                                                   'resilience in 2024 defied '
                                                   'predictions of a severe '
                                                   'recession.'}},
                         {'_id': 'vec78',
                          '_score': 3.673828125,
                          'fields': {'chunk_text': 'Consumer sentiment surveys in '
                                                   '2024 reflected optimism '
                                                   'despite high interest rates.'}},
                         {'_id': 'vec82',
                          '_score': 3.66015625,
                          'fields': {'chunk_text': 'Renewable energy subsidies in '
                                                   '2024 reduced the global '
                                                   'reliance on fossil fuels.'}},
                         {'_id': 'vec53',
                          '_score': 3.642578125,
                          'fields': {'chunk_text': 'Cryptocurrencies faced '
                                                   'regulatory scrutiny in 2024, '
                                                   'leading to volatility and '
                                                   'reduced market '
                                                   'capitalization.'}},
                         {'_id': 'vec94',
                          '_score': 3.625,
                          'fields': {'chunk_text': 'Cyberattacks targeting '
                                                   'financial institutions in 2024 '
                                                   'led to record cybersecurity '
                                                   'spending.'}},
                         {'_id': 'vec45',
                          '_score': 3.607421875,
                          'fields': {'chunk_text': 'Corporate earnings in Q4 2024 '
                                                   'were largely impacted by '
                                                   'rising raw material costs and '
                                                   'currency fluctuations.'}},
                         {'_id': 'vec47',
                          '_score': 3.576171875,
                          'fields': {'chunk_text': 'The housing market saw a '
                                                   'rebound in late 2024, driven '
                                                   'by falling mortgage rates and '
                                                   'pent-up demand.'}},
                         {'_id': 'vec84',
                          '_score': 3.5703125,
                          'fields': {'chunk_text': "The IMF's 2024 global outlook "
                                                   'highlighted risks of '
                                                   'stagflation in emerging '
                                                   'markets.'}},
                         {'_id': 'vec41',
                          '_score': 3.5546875,
                          'fields': {'chunk_text': 'Forecasts of global supply '
                                                   'chain disruptions eased in '
                                                   'late 2024, but consumer prices '
                                                   'remained elevated due to '
                                                   'persistent demand.'}},
                         {'_id': 'vec65',
                          '_score': 3.537109375,
                          'fields': {'chunk_text': 'The global shipping industry '
                                                   'experienced declining freight '
                                                   'rates in 2024 due to '
                                                   'overcapacity and reduced '
                                                   'demand.'}},
                         {'_id': 'vec96',
                          '_score': 3.53125,
                          'fields': {'chunk_text': 'New trade agreements signed '
                                                   '2022 will make an impact in '
                                                   '2024'}},
                         {'_id': 'vec86',
                          '_score': 3.52734375,
                          'fields': {'chunk_text': 'Digital transformation '
                                                   'initiatives in 2024 drove '
                                                   'productivity gains in the '
                                                   'services sector.'}},
                         {'_id': 'vec95',
                          '_score': 3.5234375,
                          'fields': {'chunk_text': 'Automation in agriculture in '
                                                   '2024 increased yields but '
                                                   'displaced rural workers.'}},
                         {'_id': 'vec64',
                          '_score': 3.51171875,
                          'fields': {'chunk_text': 'Rising commodity prices in '
                                                   '2024 strained emerging markets '
                                                   'dependent on imports of raw '
                                                   'materials.'}},
                         {'_id': 'vec79',
                          '_score': 3.51171875,
                          'fields': {'chunk_text': 'The resurgence of industrial '
                                                   'policy in Q1 2024 focused on '
                                                   'decoupling critical supply '
                                                   'chains.'}},
                         {'_id': 'vec66',
                          '_score': 3.48046875,
                          'fields': {'chunk_text': 'Bank lending to small and '
                                                   'medium-sized enterprises '
                                                   'surged in 2024 as governments '
                                                   'incentivized '
                                                   'entrepreneurship.'}},
                         {'_id': 'vec6',
                          '_score': 3.4765625,
                          'fields': {'chunk_text': 'Unemployment hit a record low '
                                                   'of 3.7% in Q4 of 2024.'}},
                         {'_id': 'vec58',
                          '_score': 3.39453125,
                          'fields': {'chunk_text': 'Oil production cuts in Q1 2024 '
                                                   'by OPEC nations drove prices '
                                                   'higher, influencing global '
                                                   'energy policies.'}},
                         {'_id': 'vec80',
                          '_score': 3.390625,
                          'fields': {'chunk_text': 'Technological innovation in '
                                                   'the fintech sector disrupted '
                                                   'traditional banking in 2024.'}},
                         {'_id': 'vec75',
                          '_score': 3.37109375,
                          'fields': {'chunk_text': 'The collapse of Silicon Valley '
                                                   'Bank raised questions about '
                                                   'regulatory oversight in '
                                                   '2024.'}},
                         {'_id': 'vec67',
                          '_score': 3.357421875,
                          'fields': {'chunk_text': 'Renewable energy projects '
                                                   'accounted for a record share '
                                                   'of global infrastructure '
                                                   'investment in 2024.'}},
                         {'_id': 'vec56',
                          '_score': 3.341796875,
                          'fields': {'chunk_text': 'Consumer confidence indices '
                                                   'remained resilient in Q2 2024 '
                                                   'despite fears of an impending '
                                                   'recession.'}}]},
     'usage': {'embed_total_tokens': 9, 'read_units': 1}}
    ```
  </Step>

  <Step title="Merge and deduplicate the results">
    Merge the 40 dense and 40 sparse results and deduplicated them based on the field you used to link sparse and dense vectors.

    For example, the following code merges and deduplicates the results based on the `_id` field, resulting in 52 unique results.

    ```python Python theme={null}
    def merge_chunks(h1, h2):
        """Get the unique hits from two search results and return them as single array of {'_id', 'chunk_text'} dicts, printing each dict on a new line."""
        # Deduplicate by _id
        deduped_hits = {hit['_id']: hit for hit in h1['result']['hits'] + h2['result']['hits']}.values()
        # Sort by _score descending
        sorted_hits = sorted(deduped_hits, key=lambda x: x['_score'], reverse=True)
        # Transform to format for reranking
        result = [{'_id': hit['_id'], 'chunk_text': hit['fields']['chunk_text']} for hit in sorted_hits]
        return result

    merged_results = merge_chunks(sparse_results, dense_results)

    print('[\n   ' + ',\n   '.join(str(obj) for obj in merged_results) + '\n]')
    ```

    ```console Response [expandable] theme={null}
    [
       {'_id': 'vec92', 'chunk_text': 'Income inequality widened in 2024 despite strong economic growth in developed nations.'},
       {'_id': 'vec69', 'chunk_text': 'The agricultural sector faced challenges in 2024 due to extreme weather and rising input costs.'},
       {'_id': 'vec70', 'chunk_text': 'Consumer spending patterns shifted in 2024, with a greater focus on experiences over goods.'},
       {'_id': 'vec90', 'chunk_text': 'The surge in e-commerce in 2024 was facilitated by advancements in logistics technology.'},
       {'_id': 'vec78', 'chunk_text': 'Consumer sentiment surveys in 2024 reflected optimism despite high interest rates.'},
       {'_id': 'vec82', 'chunk_text': 'Renewable energy subsidies in 2024 reduced the global reliance on fossil fuels.'},
       {'_id': 'vec94', 'chunk_text': 'Cyberattacks targeting financial institutions in 2024 led to record cybersecurity spending.'},
       {'_id': 'vec65', 'chunk_text': 'The global shipping industry experienced declining freight rates in 2024 due to overcapacity and reduced demand.'},
       {'_id': 'vec86', 'chunk_text': 'Digital transformation initiatives in 2024 drove productivity gains in the services sector.'},
       {'_id': 'vec95', 'chunk_text': 'Automation in agriculture in 2024 increased yields but displaced rural workers.'},
       {'_id': 'vec80', 'chunk_text': 'Technological innovation in the fintech sector disrupted traditional banking in 2024.'},
       {'_id': 'vec67', 'chunk_text': 'Renewable energy projects accounted for a record share of global infrastructure investment in 2024.'},
       {'_id': 'vec35', 'chunk_text': 'Unemployment hit at 2.4% in Q3 of 2024.'},
       {'_id': 'vec36', 'chunk_text': 'Unemployment is expected to hit 2.5% in Q3 of 2024.'},
       {'_id': 'vec6', 'chunk_text': 'Unemployment hit a record low of 3.7% in Q4 of 2024.'},
       {'_id': 'vec42', 'chunk_text': 'Tech sector layoffs in Q3 2024 have reshaped hiring trends across high-growth industries.'},
       {'_id': 'vec48', 'chunk_text': 'Wage growth outpaced inflation for the first time in years, signaling improved purchasing power in 2024.'},
       {'_id': 'vec55', 'chunk_text': 'Trade tensions between the U.S. and China escalated in 2024, impacting global supply chains and investment flows.'},
       {'_id': 'vec45', 'chunk_text': 'Corporate earnings in Q4 2024 were largely impacted by rising raw material costs and currency fluctuations.'},
       {'_id': 'vec72', 'chunk_text': 'In early 2024, global GDP growth slowed, driven by weaker exports in Asia and Europe.'},
       {'_id': 'vec29', 'chunk_text': 'Growth vs. value investing strategies in 2024.'},
       {'_id': 'vec46', 'chunk_text': 'Economic recovery in Q2 2024 relied heavily on government spending in infrastructure and green energy projects.'},
       {'_id': 'vec79', 'chunk_text': 'The resurgence of industrial policy in Q1 2024 focused on decoupling critical supply chains.'},
       {'_id': 'vec87', 'chunk_text': "The U.S. labor market's resilience in 2024 defied predictions of a severe recession."},
       {'_id': 'vec40', 'chunk_text': 'Labor market trends show a declining participation rate despite record low unemployment in 2024.'},
       {'_id': 'vec37', 'chunk_text': 'In Q3 2025 unemployment for the prior year was revised to 2.2%'},
       {'_id': 'vec58', 'chunk_text': 'Oil production cuts in Q1 2024 by OPEC nations drove prices higher, influencing global energy policies.'},
       {'_id': 'vec47', 'chunk_text': 'The housing market saw a rebound in late 2024, driven by falling mortgage rates and pent-up demand.'},
       {'_id': 'vec41', 'chunk_text': 'Forecasts of global supply chain disruptions eased in late 2024, but consumer prices remained elevated due to persistent demand.'},
       {'_id': 'vec56', 'chunk_text': 'Consumer confidence indices remained resilient in Q2 2024 despite fears of an impending recession.'},
       {'_id': 'vec63', 'chunk_text': 'Population aging emerged as a critical economic issue in 2024, especially in advanced economies.'},
       {'_id': 'vec52', 'chunk_text': 'Record-breaking weather events in early 2024 have highlighted the growing economic impact of climate change.'},
       {'_id': 'vec23', 'chunk_text': 'Top SPAC mergers in the technology sector for 2024.'},
       {'_id': 'vec62', 'chunk_text': 'Private equity activity in 2024 focused on renewable energy and technology sectors amid shifting investor priorities.'},
       {'_id': 'vec64', 'chunk_text': 'Rising commodity prices in 2024 strained emerging markets dependent on imports of raw materials.'},
       {'_id': 'vec54', 'chunk_text': 'The global tourism sector showed signs of recovery in late 2024 after years of pandemic-related setbacks.'},
       {'_id': 'vec96', 'chunk_text': 'New trade agreements signed 2022 will make an impact in 2024'},
       {'_id': 'vec49', 'chunk_text': "China's economic growth in 2024 slowed to its lowest level in decades due to structural reforms and weak exports."},
       {'_id': 'vec7', 'chunk_text': 'The CPI index rose by 6% in July 2024, raising concerns about purchasing power.'},
       {'_id': 'vec84', 'chunk_text': "The IMF's 2024 global outlook highlighted risks of stagflation in emerging markets."},
       {'_id': 'vec13', 'chunk_text': 'Credit card APRs reached an all-time high of 22.8% in 2024.'},
       {'_id': 'vec53', 'chunk_text': 'Cryptocurrencies faced regulatory scrutiny in 2024, leading to volatility and reduced market capitalization.'},
       {'_id': 'vec60', 'chunk_text': 'Healthcare spending in 2024 surged as governments expanded access to preventive care and pandemic preparedness.'},
       {'_id': 'vec91', 'chunk_text': 'The impact of ESG investing on corporate strategies has been a major focus in 2024.'},
       {'_id': 'vec68', 'chunk_text': 'Cybersecurity spending reached new highs in 2024, reflecting the growing threat of digital attacks on infrastructure.'},
       {'_id': 'vec59', 'chunk_text': 'The adoption of digital currencies by central banks increased in 2024, reshaping monetary policy frameworks.'},
       {'_id': 'vec39', 'chunk_text': 'The rise in energy prices significantly impacted inflation trends during the first half of 2024.'},
       {'_id': 'vec66', 'chunk_text': 'Bank lending to small and medium-sized enterprises surged in 2024 as governments incentivized entrepreneurship.'},
       {'_id': 'vec57', 'chunk_text': 'Startups in 2024 faced tighter funding conditions as venture capitalists focused on profitability over growth.'},
       {'_id': 'vec75', 'chunk_text': 'The collapse of Silicon Valley Bank raised questions about regulatory oversight in 2024.'},
       {'_id': 'vec51', 'chunk_text': 'The European Union introduced new fiscal policies in 2024 aimed at reducing public debt without stifling growth.'},
       {'_id': 'vec89', 'chunk_text': 'Venture capital investments in 2024 leaned heavily toward AI and automation startups.'}
    ]
    ```
  </Step>

  <Step title="Rerank the results">
    Use one of Pinecone's [hosted reranking models](/guides/search/rerank-results#reranking-models) to rerank the merged and deduplicated results based on a unified relevance score and then return a smaller set of the most highly relevant results.

    For example, the following code sends the 52 unique results from the last step to the `bge-reranker-v2-m3` reranking model and returns the top 10 most relevant results.

    ```python Python theme={null}
    result = pc.inference.rerank(
        model="bge-reranker-v2-m3",
        query=query,
        documents=merged_results,
        rank_fields=["chunk_text"],
        top_n=10,
        return_documents=True,
        parameters={
            "truncate": "END"
        }
    )

    print("Query", query)
    print('-----')
    for row in result.data:
        print(f"{row['document']['_id']} - {round(row['score'], 2)} - {row['document']['chunk_text']}")
    ```

    ```console Response [expandable] theme={null}
    Query Q3 2024 us economic data
    -----
    vec36 - 0.84 - Unemployment is expected to hit 2.5% in Q3 of 2024.
    vec35 - 0.76 - Unemployment hit at 2.4% in Q3 of 2024.
    vec48 - 0.33 - Wage growth outpaced inflation for the first time in years, signaling improved purchasing power in 2024.
    vec37 - 0.25 - In Q3 2025 unemployment for the prior year was revised to 2.2%
    vec42 - 0.21 - Tech sector layoffs in Q3 2024 have reshaped hiring trends across high-growth industries.
    vec87 - 0.2 - The U.S. labor market's resilience in 2024 defied predictions of a severe recession.
    vec63 - 0.08 - Population aging emerged as a critical economic issue in 2024, especially in advanced economies.
    vec92 - 0.08 - Income inequality widened in 2024 despite strong economic growth in developed nations.
    vec72 - 0.07 - In early 2024, global GDP growth slowed, driven by weaker exports in Asia and Europe.
    vec46 - 0.06 - Economic recovery in Q2 2024 relied heavily on government spending in infrastructure and green energy projects.
    ```
  </Step>
</Steps>

---

# Source: https://docs.pinecone.io/guides/index-data/implement-multitenancy.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Implement multitenancy

> Use namespaces to isolate tenant data securely.

[Multitenancy](https://en.wikipedia.org/wiki/Multitenancy) is a software architecture where a single instance of a system serves multiple customers, or tenants, while ensuring data isolation between them for privacy and security.

This page shows you how to implement multitenancy in Pinecone using a **serverless index with one namespace per tenant**.

<Note>
  For design guidance on choosing between namespaces, metadata filtering, and other approaches, see [Design for multi-tenancy](/guides/index-data/data-modeling#design-for-multi-tenancy).
</Note>

<Note>
  While the Standard and Enterprise plans support up to [100,000 namespaces per index](/reference/api/database-limits#namespaces-per-serverless-index), Pinecone can accommodate million-scale namespaces and beyond for specific use cases. If your application requires more than 100,000 namespaces, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).
</Note>

## How it works

<img className="block max-w-full dark:hidden" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namepsaces.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=8a85b3a90f4c64964ac2d19b72c9e537" data-og-width="1400" width="1400" data-og-height="880" height="880" data-path="images/multitenant-saas-namepsaces.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namepsaces.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=ecf6f1c273f4802044152a03c2e9990f 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namepsaces.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=2636f6f7b3b635bea0c4e7b7c82d6555 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namepsaces.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=04e9f5f77a2c242055638347b2949138 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namepsaces.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=719792f97af9dbcba9cbc4030bcf8ae3 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namepsaces.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=6577382ddaebdc1d1dd2b0ed5849b042 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namepsaces.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=ff6a56b6d70a48b5f8646f5fb0c75598 2500w" />

<img className="hidden max-w-full dark:block" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namespaces-dark.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=46f954e9010597b5303aee2984be1e29" data-og-width="1400" width="1400" data-og-height="880" height="880" data-path="images/multitenant-saas-namespaces-dark.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namespaces-dark.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=44cc681389721a2730e2cb0ece7c8bfb 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namespaces-dark.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=3ee30fd16eb0d1c99cf4cf26864ed5b3 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namespaces-dark.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=b3390cd3c7c51b37934bf94c0927d0ee 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namespaces-dark.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=1621cef8d384c50572c16c206b92e821 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namespaces-dark.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=5e9e90a83acc40ceab104caa1493f61b 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-namespaces-dark.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=908341daa70cac819162457eb2615763 2500w" />

In Pinecone, an [index](/guides/index-data/indexing-overview) is the highest-level organizational unit of data, where you define the dimension of vectors to be stored in the index and the measure of similarity to be used when querying the index.

Within an index, records are stored in [namespaces](/guides/index-data/indexing-overview#namespaces), and all [upserts](/guides/index-data/upsert-data), [queries](/guides/search/search-overview), and other [data plane operations](/reference/api/latest/data-plane) always target one namespace.

This structure makes it easy to implement multitenancy. For example, for an AI-powered SaaS application where you need to isolate the data of each customer, you would assign each customer to a namespace and target their writes and queries to that namespace (diagram above).

In cases where you have different workload patterns (e.g., RAG and semantic search), you would use a different index for each workload, with one namespace per customer in each index:

<img className="block max-w-full dark:hidden" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=d30f68fcbf1b6e259850350461f10e3c" data-og-width="1560" width="1560" data-og-height="1200" height="1200" data-path="images/multitenant-saas-indexes.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=ff79239d5622f8c8430a71d5a64abee8 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=9e76456ea4ff54d9333b1e330f6622e2 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=a99629c4e37af33b8c7fe93201cc5075 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=5674100409641614c0a2a23d99fe4ce2 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=898ceabfffd1c5d18b3759151aac9276 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=1b5f238711bd3e42f9a19c29e6f096ce 2500w" />

<img className="hidden max-w-full dark:block" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes-dark.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=1c64b2977ebdf57a2bbc8274b91e399e" data-og-width="1560" width="1560" data-og-height="1200" height="1200" data-path="images/multitenant-saas-indexes-dark.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes-dark.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=3e38602271765a6cb8e98108b1f0decb 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes-dark.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=6d6b3925edd6dd5ea1ab0bb3246699d6 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes-dark.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=93166d83515098bde4f88aa0748feb3f 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes-dark.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=7b5a0d541d8b132bd4f7290886dd2372 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes-dark.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=fe477a1edbd28071b43363656d02745f 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/multitenant-saas-indexes-dark.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=3c49967484298ee8e0675ef7a620eb5b 2500w" />

<Accordion title="Understand the benefits">
  * **Tenant isolation:** In the [serverless architecture](/guides/get-started/database-architecture), each namespace is stored separately, so using namespaces provides physical isolation of data between tenants/customers. This reduces the risk of application bugs that could query the wrong tenant's data.

  * **No noisy neighbors:** Reads and writes always target a single namespace, so the behavior of one tenant/customer does not affect other tenants/customers.

  * **No maintenance effort:** Serverless indexes scale automatically based on usage; you don't configure or manage any compute or storage resources.

  * **Cost efficiency:** Query cost is based on namespace size (1 RU per 1 GB). With 100 tenants of 1 GB each, querying one tenant's namespace costs 1 RU. Using metadata filtering in a single 100 GB namespace would cost 100 RUs for the same query, because it scans all data regardless of filters.

  * **Simple tenant offboarding:** To offboard a tenant/customer, you just [delete the relevant namespace](/guides/manage-data/delete-data#delete-all-records-from-a-namespace). This is a lightweight and almost instant operation.
</Accordion>

## 1. Create a serverless index

Based on a [breakthrough architecture](/guides/get-started/database-architecture), serverless indexes scale automatically based on usage, and you pay only for the amount of data stored and operations performed. Combined with the isolation of tenant data using namespaces (next step), serverless indexes are ideal for multitenant use cases.

To [create a serverless index](/guides/index-data/create-an-index#create-a-serverless-index), use the `spec` parameter to define the cloud and region where the index should be deployed. For Python, you also need to import the `ServerlessSpec` class.

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone
  from pinecone import ServerlessSpec

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.create_index(
    name="multitenant-app",
    dimension=8,
    metric="cosine",
    spec=ServerlessSpec(
      cloud="aws",
      region="us-east-1"
    )
  )
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({
    apiKey: 'YOUR_API_KEY'
  });

  await pc.createIndex({
    name: 'multitenant-app',
    dimension: 8,
    metric: 'cosine',
    spec: {
      serverless: {
        cloud: 'aws',
        region: 'us-east-1'
      }
    }
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;

  public class CreateServerlessIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          pc.createServerlessIndex("multitenant-app", "cosine", 8, "aws", "us-east-1");
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
    	bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // Serverless index
      indexName := "multi-tenant-app"
      vectorType := "dense"
      dimension := int32(8)
      metric := pinecone.Cosine
      deletionProtection := pinecone.DeletionProtectionDisabled

      idx, err := pc.CreateServerlessIndex(ctx, &pinecone.CreateServerlessIndexRequest{
          Name:               indexName,
          VectorType:         &vectorType,
          Dimension:          &dimension,
          Metric:             &metric,
          Cloud:              pinecone.Aws,
          Region:             "us-east-1",
          DeletionProtection: &deletionProtection,
      })
      if err != nil {
          log.Fatalf("Failed to create serverless index: %v", err)
      } else {
          fmt.Printf("Successfully created serverless index: %v", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var createIndexRequest = await pinecone.CreateIndexAsync(new CreateIndexRequest
  {
      Name = "multitenant-app",
      Dimension = 8,
      Metric = MetricType.Cosine,
      Spec = new ServerlessIndexSpec
      {
          Serverless = new ServerlessSpec
          {
              Cloud = ServerlessSpecCloud.Aws,
              Region = "us-east-1",
          }
      },
      DeletionProtection = DeletionProtection.Disabled
  });
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s "https://api.pinecone.io/indexes" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
           "name": "multitenant-app",
           "dimension": 8,
           "metric": "cosine",
           "spec": {
              "serverless": {
                 "cloud": "aws",
                 "region": "us-east-1"
              }
           }
        }'
  ```
</CodeGroup>

## 2. Isolate tenant data

In a multitenant solution, you need to isolate data between tenants. To achieve this in Pinecone, use one namespace per tenant. In the [serverless architecture](/guides/get-started/database-architecture), each namespace is stored separately, so this approach ensures physical isolation of each tenant's data.

To [create a namespace for a tenant](/guides/index-data/indexing-overview#namespaces#creating-a-namespace), specify the `namespace` parameter when first [upserting](/guides/index-data/upsert-data) the tenant's records. For example, the following code upserts records for `tenant1` and `tenant2` into the `multitenant-app` index:

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")
  index = pc.Index("multitenant-app")

  index.upsert(
    vectors=[
      {"id": "A", "values": [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1]},
      {"id": "B", "values": [0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2]},
      {"id": "C", "values": [0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3]},
      {"id": "D", "values": [0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4]}
    ],
    namespace="tenant1"
  )

  index.upsert(
    vectors=[
      {"id": "E", "values": [0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5]},
      {"id": "F", "values": [0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6]},
      {"id": "G", "values": [0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7]},
      {"id": "H", "values": [0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8]}
    ],
    namespace="tenant2"
  )
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" });
  const index = pc.index("multitenant-app");

  await index.namespace("tenant1").upsert([
    {
      "id": "A", 
      "values": [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1]
    },
    {
      "id": "B", 
      "values": [0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2]
    },
    {
      "id": "C", 
      "values": [0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3]
    },
    {
      "id": "D", 
      "values": [0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4]
    }
  ]);

  await index.namespace("tenant2").upsert([
    {
      "id": "E", 
      "values": [0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5]
    },
    {
      "id": "F", 
      "values": [0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6]
    },
    {
      "id": "G", 
      "values": [0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7]
    },
    {
      "id": "H", 
      "values": [0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8]
    }
  ]);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.clients.Pinecone;

  import java.util.Arrays;
  import java.util.List;

  public class UpsertExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          String indexName = "multitenant-app";
          Index index = pc.getIndexConnection(indexName);
          List<Float> values1 = Arrays.asList(0.1f, 0.1f, 0.1f, 0.1f, 0.1f, 0.1f, 0.1f, 0.1f);
          List<Float> values2 = Arrays.asList(0.2f, 0.2f, 0.2f, 0.2f, 0.2f, 0.2f, 0.2f, 0.2f);
          List<Float> values3 = Arrays.asList(0.3f, 0.3f, 0.3f, 0.3f, 0.3f, 0.3f, 0.3f, 0.3f);
          List<Float> values4 = Arrays.asList(0.4f, 0.4f, 0.4f, 0.4f, 0.4f, 0.4f, 0.4f, 0.4f);
          List<Float> values5 = Arrays.asList(0.5f, 0.5f, 0.5f, 0.5f, 0.5f, 0.5f, 0.5f, 0.5f);
          List<Float> values6 = Arrays.asList(0.6f, 0.6f, 0.6f, 0.6f, 0.6f, 0.6f, 0.6f, 0.6f);
          List<Float> values7 = Arrays.asList(0.7f, 0.7f, 0.7f, 0.7f, 0.7f, 0.7f, 0.7f, 0.7f);
          List<Float> values8 = Arrays.asList(0.8f, 0.8f, 0.8f, 0.8f, 0.8f, 0.8f, 0.8f, 0.8f);

          index.upsert("A", values1, "tenant1");
          index.upsert("B", values2, "tenant1");
          index.upsert("C", values3, "tenant1");
          index.upsert("D", values4, "tenant1");
          index.upsert("E", values5, "tenant2");
          index.upsert("F", values6, "tenant2");
          index.upsert("G", values7, "tenant2");
          index.upsert("H", values8, "tenant2");
      }
  }
  ```

  ```go Go theme={null}
  // Add to the main function:
  idx, err := pc.DescribeIndex(ctx, indexName)
  if err != nil {
      log.Fatalf("Failed to describe index \"%v\": %v", idx.Name, err)
  }

  idxConnection1, err := pc.Index(pinecone.NewIndexConnParams{Host: idx.Host, Namespace: "tenant1"})
  if err != nil {
      log.Fatalf("Failed to create IndexConnection1 for Host %v: %v", idx.Host, err)
  }

  // This reuses the gRPC connection of idxConnection1 while targeting a different namespace
  idxConnection2 := idxConnection1.WithNamespace("tenant2")

  vectors1 := []*pinecone.Vector{
      {
          Id:     "A",
          Values: []float32{0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1},
      },
      {
          Id:     "B",
          Values: []float32{0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2},
      },
      {
          Id:     "C",
          Values: []float32{0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3},
      },   
      {
          Id:     "D",
          Values: []float32{0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4},
      },   
  }

  vectors2 := []*pinecone.Vector{
      {
          Id:     "E",
          Values: []float32{0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5},
      },
      {
          Id:     "F",
          Values: []float32{0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6},
      },
      {
          Id:     "G",
          Values: []float32{0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7},
      },   
      {
          Id:     "H",
          Values: []float32{0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8},
      },   
  }

  count1, err := idxConnection1.UpsertVectors(ctx, vectors1)
  if err != nil {
      log.Fatalf("Failed to upsert vectors: %v", err)
  } else {
      fmt.Printf("Successfully upserted %d vector(s)!\n", count1)
  }

  count2, err := idxConnection2.UpsertVectors(ctx, vectors2)
  if err != nil {
      log.Fatalf("Failed to upsert vectors: %v", err)
  } else {
      fmt.Printf("Successfully upserted %d vector(s)!\n", count2)
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var index = pinecone.Index("multitenant-app");

  var upsertResponse1 = await index.UpsertAsync(new UpsertRequest {
      Vectors = new[]
      {
          new Vector
          {
              Id = "A",
              Values = new[] { 0.1f, 0.1f, 0.1f, 0.1f, 0.1f, 0.1f, 0.1f, 0.1f },
          },
          new Vector
          {
              Id = "B",
              Values = new[] { 0.2f, 0.2f, 0.2f, 0.2f, 0.2f, 0.2f, 0.2f, 0.2f },
          },
          new Vector
          {
              Id = "C",
              Values = new[] { 0.3f, 0.3f, 0.3f, 0.3f, 0.3f, 0.3f, 0.3f, 0.3f },
          },
          new Vector
          {
              Id = "D",
              Values = new[] { 0.4f, 0.4f, 0.4f, 0.4f, 0.4f, 0.4f, 0.4f, 0.4f },
          }
      },
      Namespace = "tenant1",
  });

  var upsertResponse2 = await index.UpsertAsync(new UpsertRequest {
      Vectors = new[]
      {
          new Vector
          {
              Id = "E",
              Values = new[] { 0.5f, 0.5f, 0.5f, 0.5f, 0.5f, 0.5f, 0.5f, 0.5f },
          },
          new Vector
          {
              Id = "F",
              Values = new[] { 0.6f, 0.6f, 0.6f, 0.6f, 0.6f, 0.6f, 0.6f, 0.6f },
          },
          new Vector
          {
              Id = "G",
              Values = new[] { 0.7f, 0.7f, 0.7f, 0.7f, 0.7f, 0.7f, 0.7f, 0.7f },
          },
          new Vector
          {
              Id = "H",
              Values = new[] { 0.8f, 0.8f, 0.8f, 0.8f, 0.8f, 0.8f, 0.8f, 0.8f },
          }
      },
      Namespace = "tenant2",
  });
  ```

  ```bash curl theme={null}
  # The `POST` requests below uses the unique endpoint for an index.
  # See https://docs.pinecone.io/guides/manage-data/target-an-index for details.
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/vectors/upsert" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "vectors": [
        {
          "id": "A", 
          "values": [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1]
        },
        {
          "id": "B", 
          "values": [0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2]
        },
        {
          "id": "C", 
          "values": [0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3]
        },
        {
          "id": "D", 
          "values": [0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4]
        }
      ],
      "namespace": "tenant1"
    }'

  curl "https://$INDEX_HOST/vectors/upsert" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "vectors": [
        {
          "id": "E", 
          "values": [0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5]
        },
        {
          "id": "F", 
          "values": [0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6]
        },
        {
          "id": "G", 
          "values": [0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7]
        },
        {
          "id": "H", 
          "values": [0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8]
        }
      ],
      "namespace": "tenant2"
    }'
  ```
</CodeGroup>

When upserting additional records for a tenant, or when [updating](/guides/manage-data/update-data) or [deleting](/guides/manage-data/delete-data) records for a tenant, specify the tenant's `namespace`. For example, the following code updates the dense vector value of record `A` in `tenant1`:

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")
  index = pc.Index("multitenant-app")

  index.update(id="A", values=[0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8], namespace="tenant1")
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" });
  const index = pc.index("multitenant-app");

  await index.namespace('tenant1').update({
   	id: 'A',
   	values: [0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8]
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.clients.Pinecone;
  import io.pinecone.proto.UpdateResponse;

  import java.util.Arrays;
  import java.util.List;

  public class UpdateExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          Index index = pc.getIndexConnection("multitenant-app");
          List<Float> values = Arrays.asList(0.1f, 0.2f, 0.3f, 0.4f, 0.5f, 0.6f, 0.7f, 0.8f);
          UpdateResponse updateResponse = index.update("A", values, null, "tenant1", null, null);
          System.out.println(updateResponse);
      }
  }
  ```

  ```go Go theme={null}
  // Add to the main function:
  idxConn1.UpdateVector(ctx, &pinecone.UpdateVectorRequest{
      Id:     "A",
      Values: []float32{0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8},
  })
  if err != nil {
      log.Fatalf("Failed to update vector with ID %v: %v", id, err)
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var index = pinecone.Index("multitenant-app");

  var upsertResponse = await index.UpsertAsync(new UpsertRequest {
      Vectors = new[]
      {
          new Vector
          {
              Id = "A",
              Values = new[] { 0.1f, 0.2f, 0.3f, 0.4f, 0.5f, 0.6f, 0.7f, 0.8f },
          }
      },
      Namespace = "tenant1",
  });
  ```

  ```bash curl theme={null}
  # The `POST` request below uses the unique endpoint for an index.
  # See https://docs.pinecone.io/guides/manage-data/target-an-index for details.
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/vectors/update" \
  	-H "Api-Key: $PINECONE_API_KEY" \
  	-H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
  	-d '{
  			"id": "A",
  			"values": [01., 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8],
  			"namespace": "tenant1"
  		}'
  ```
</CodeGroup>

## 3. Query tenant data

In a multitenant solution, you need to ensure that the queries of one tenant do not affect the experience of other tenants/customers. To achieve this in Pinecone, target each tenant's [queries](/guides/search/search-overview) at the namespace for the tenant.

For example, the following code queries only `tenant2` for the 3 vectors that are most similar to an example query vector:

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")
  index = pc.Index("multitenant-app")

  query_results = index.query(
      namespace="tenant2",
      vector=[0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7],
      top_k=3,
      include_values=True
  )

  print(query_results)

  # Returns:
  # {'matches': [{'id': 'F',
  #               'score': 1.00000012,
  #               'values': [0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6]},
  #              {'id': 'G',
  #               'score': 1.0,
  #               'values': [0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7]},
  #              {'id': 'E',
  #               'score': 1.0,
  #               'values': [0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5]}],
  #  'namespace': 'tenant2',
  #  'usage': {'read_units': 6}}
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" });
  const index = pc.index("multitenant-app");

  const queryResponse = await index.namespace("tenant2").query({
    topK: 3,
    vector: [0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7],
    includeValues: true
  });

  console.log(queryResponse);

  // Returns:
  {
    "matches": [
      {
        "id": "F",
        "score": 1.00000012,
        "values": [
          0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6
        ]
      },
      {
        "id": "E",
        "score": 1,
        "values": [ 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5
        ]
      },
      {
        "id": "G",
        "score": 1,
        "values": [
          0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7
        ]
      }
    ],
    "namespace": "tenant2",
    "usage": {
      "readUnits": 6
    }
  }
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.clients.Pinecone;
  import io.pinecone.unsigned_indices_model.QueryResponseWithUnsignedIndices;

  import java.util.Arrays;
  import java.util.List;

  public class QueryExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          String indexName = "multitenant-app";
          Index index = pc.getIndexConnection(indexName);
          List<Float> queryVector = Arrays.asList(0.7f, 0.7f, 0.7f, 0.7f, 0.7f, 0.7f, 0.7f, 0.7f);
          QueryResponseWithUnsignedIndices queryResponse = index.query(3, queryVector2, null, null, null, "tenant2", null, true, false);
          System.out.println(queryResponse);
      }
  }

  // Results:
  // class QueryResponseWithUnsignedIndices {
  //     matches: [ScoredVectorWithUnsignedIndices {
  //         score: 1.00000012
  //         id: F
  //         values: [0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6]
  //         metadata: 
  //         sparseValuesWithUnsignedIndices: SparseValuesWithUnsignedIndices {
  //             indicesWithUnsigned32Int: []
  //             values: []
  //         }
  //     }, ScoredVectorWithUnsignedIndices {
  //         score: 1
  //         id: E
  //         values: [0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5]
  //         metadata: 
  //         sparseValuesWithUnsignedIndices: SparseValuesWithUnsignedIndices {
  //             indicesWithUnsigned32Int: []
  //             values: []
  //         }
  //     }, ScoredVectorWithUnsignedIndices {
  //         score: 0.07999992
  //         id: G
  //         values: [0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7]
  //         metadata: 
  //         sparseValuesWithUnsignedIndices: SparseValuesWithUnsignedIndices {
  //             indicesWithUnsigned32Int: []
  //             values: []
  //         }
  //     }]
  //     namespace: tenant2
  //     usage: read_units: 6
  // }
  ```

  ```go Go theme={null}
  // Add to the main function:
  queryVector := []float32{0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7}

  res, err := idxConnection2.QueryByVectorValues(ctx, &pinecone.QueryByVectorValuesRequest{
      Vector:        queryVector,
      TopK:          3,
      IncludeValues: true,
  })
  if err != nil {
      log.Fatalf("Error encountered when querying by vector: %v", err)
  } else {
      fmt.Printf(prettifyStruct(res))
  }

  // Returns:
  // {
  //   "matches": [
  //     {
  //       "vector": {
  //         "id": "F",
  //         "values": [
  //           0.6,
  //           0.6,
  //           0.6,
  //           0.6,
  //           0.6,
  //           0.6,
  //           0.6,
  //           0.6
  //         ]
  //       },
  //       "score": 1.0000001
  //     },
  //     {
  //       "vector": {
  //         "id": "G",
  //         "values": [
  //           0.7,
  //           0.7,
  //           0.7,
  //           0.7,
  //           0.7,
  //           0.7,
  //           0.7,
  //           0.7
  //         ]
  //       },
  //       "score": 1
  //     },
  //     {
  //       "vector": {
  //         "id": "H",
  //         "values": [
  //           0.8,
  //           0.8,
  //           0.8,
  //           0.8,
  //           0.8,
  //           0.8,
  //           0.8,
  //           0.8
  //         ]
  //       },
  //       "score": 1
  //     }
  //   ],
  //   "usage": {
  //     "read_units": 6
  //   }
  // }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var index = pinecone.Index("multitenant-app");

  var queryResponse = await index.QueryAsync(new QueryRequest {
      Vector = new[] { 0.7f, 0.7f, 0.7f, 0.7f, 0.7f, 0.7f, 0.7f, 0.7f },
      Namespace = "tenant2",
      TopK = 3,
  });

  Console.WriteLine(queryRespnose);
  ```

  ```shell curl theme={null}
  # The `POST` requests below uses the unique endpoint for an index.
  # See https://docs.pinecone.io/guides/manage-data/target-an-index for details.
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/query" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "namespace": "tenant2",
      "vector": [0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7],
      "topK": 3,
      "includeValues": true
    }'
  #
  # Output:
  # {
  #   "matches": [
  #     {
  #       "id": "F",
  #       "score": 1.00000012,
  #       "values": [0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6]
  #     },
  #     {
  #       "id": "E",
  #       "score": 1,
  #       "values": [0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5]
  #     },
  #     {
  #       "id": "G",
  #       "score": 1,
  #       "values": [0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.7]
  #     }
  #   ],
  #   "namespace": "tenant2",
  #   "usage": {"read_units": 6}
  # }
  ```
</CodeGroup>

## 4. Offboard a tenant

In a multitenant solution, you also need it to be quick and easy to offboard a tenant and delete all of its records. To achieve this in Pinecone, you just [delete the namespace](/guides/manage-data/delete-data#delete-an-entire-namespace) for the specific tenant.

For example, the following code deletes the namespace and all records for `tenant1`:

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")
  index = pc.Index("multitenant-app")

  index.delete(delete_all=True, namespace='tenant1')
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" });
  const index = pc.index("multitenant-app");

  await index.namespace('tenant1').deleteAll();
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.clients.Pinecone;

  public class DeleteVectorsExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          Index index = pc.getIndexConnection("multitenant-app");
          index.deleteAll("tenant1");
      }
  }
  ```

  ```go Go theme={null}
  // Add to the main function:
  idxConnection1.DeleteAllVectorsInNamespace(ctx)
  if err != nil {
      log.Fatalf("Failed to delete vectors in namespace \"%v\": %v", idxConnection2.Namespace, err)
  }
  ```

  ```csharp C# theme={null}
  var index = pinecone.Index("multitenant-app");

  var deleteResponse = await index.DeleteAsync(new DeleteRequest {
      DeleteAll = true,
      Namespace = "tenant1",
  });
  ```

  ```bash curl theme={null}
  # The `POST` request below uses the unique endpoint for an index.
  # See https://docs.pinecone.io/guides/manage-data/target-an-index for details.
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"
  curl "https://$INDEX_HOST/vectors/delete" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "deleteAll": true,
      "namespace": "tenant1"
    }
  '
  ```
</CodeGroup>

## Alternative: Metadata filtering

When tenant isolation is not a strict requirement, or when you need to query across multiple tenants simultaneously, you can store all records in a single namespace and use metadata fields to assign records to tenants/customers. At query time, you can then [filter by metadata](/guides/index-data/indexing-overview#metadata).

<Warning>
  This approach has significant performance and cost tradeoffs compared to using namespaces:

  * Higher query costs: Queries scan the entire namespace regardless of filters, so you pay for scanning all tenants' data even though results are filtered to one tenant.
  * Slower performance: Large namespaces increase query latency, and large filters add network overhead on the request side.
  * Filter size limits: Each `$in` or `$nin` operator is limited to 10,000 values. Exceeding this limit will cause requests to fail. See [Metadata filter limits](/reference/api/database-limits#metadata-filter-limits).

  Anti-pattern: Avoid filtering by large lists of individual user IDs. Instead, use access control groups (organization, project, role), namespaces, or post-filter client-side (for semantic search).

  For detailed guidance on choosing between namespaces and metadata filtering, see [Design for multi-tenancy](/guides/index-data/data-modeling#design-for-multi-tenancy).
</Warning>

For more background on this approach, see [Multitenancy in Vector Databases](https://www.pinecone.io/learn/series/vector-databases-in-production-for-busy-engineers/vector-database-multi-tenancy/).

---

# Source: https://docs.pinecone.io/guides/index-data/import-data.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Import records

> Import large datasets efficiently from S3, GCS, or Azure into Pinecone indexes.

Importing from object storage is the most efficient and cost-effective way to load large numbers of records into an index.

To run through this guide in your browser, see the [Bulk import colab notebook](https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/pinecone-import.ipynb).

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Standard and Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

## Before you import

Before you can import records, ensure you have a serverless index, a storage integration, and data formatted in a Parquet file and uploaded to an Amazon S3 bucket, Google Cloud Storage bucket, or Azure Blob Storage container.

### Create an index

[Create a serverless index](/guides/index-data/create-an-index) for your data.

Be sure to create your index on a cloud that supports importing from the object storage you want to use:

|                                       | …to an **AWS** index | …to a **GCP** index | …to an **Azure** index |
| ------------------------------------- | :------------------: | :-----------------: | :--------------------: |
| Import from **AWS S3**…               |           ✅          |          ❌          |            ❌           |
| Import from **Google Cloud Storage**… |           ✅          |          ✅          |            ✅           |
| Import from **Azure Blob Storage**…   |           ✅          |          ✅          |            ✅           |

### Add a storage integration

To import records from a public data source, a storage integration is not required. However, to import records from a secure data source, you must create an integration to allow Pinecone access to data in your object storage. See the following guides:

* [Integrate with Amazon S3](/guides/operations/integrations/integrate-with-amazon-s3)
* [Integrate with Google Cloud Storage](/guides/operations/integrations/integrate-with-google-cloud-storage)
* [Integrate with Azure Blob Storage](/guides/operations/integrations/integrate-with-azure-blob-storage)

### Prepare your data

1. In your Amazon S3 bucket, Google Cloud Storage bucket, or Azure Blob Storage container, create an import directory containing a subdirectory for each namespace you want to import into. The namespaces must not yet exist in your index.

   For example, to import data into the namespaces `example_namespace1` and `example_namespace2`, your directory structure would look like this:

   ```
   <BUCKET_OR_CONTAINER_NAME>/
   --/<IMPORT_DIR>/
   ----/example_namespace1/
   ----/example_namespace2/
   ```

   <Tip>
     To import into the default namespace, use a subdirectory called `__default__`. The default namespace must be empty.
   </Tip>

2. For each namespace, create one or more Parquet files defining the records to import.

   Parquet files must contain specific columns, depending on the index type:

   <Tabs>
     <Tab title="Dense index">
       To import into a namespace in a [dense index](/guides/index-data/indexing-overview#dense-indexes), the Parquet file must contain the following columns:

       | Column name | Parquet type  | Description                                                                                                                     |
       | ----------- | ------------- | ------------------------------------------------------------------------------------------------------------------------------- |
       | `id`        | `STRING`      | Required. The unique [identifier for each record](/guides/get-started/concepts#record-id).                                      |
       | `values`    | `LIST<FLOAT>` | Required. A list of floating-point values that make up the [dense vector embedding](/guides/get-started/concepts#dense-vector). |
       | `metadata`  | `STRING`      | Optional. Additional [metadata](/guides/get-started/concepts#metadata) for each record. To omit from specific rows, use `NULL`. |

       <Warning>
         The Parquet file cannot contain additional columns.
       </Warning>

       For example:

       ```parquet  theme={null}
       id | values                   | metadata
       --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
       1  | [ 3.82  2.48 -4.15 ... ] | {"year": 1984, "month": 6, "source": "source1", "title": "Example1", "text": "When ..."}
       2  | [ 1.82  3.48 -2.15 ... ] | {"year": 1990, "month": 4, "source": "source2", "title": "Example2", "text": "Who ..."}
       ```
     </Tab>

     <Tab title="Sparse index">
       To import into a namespace in a [sparse index](/guides/index-data/indexing-overview#sparse-indexes), the Parquet file must contain the following columns:

       | Column name     | Parquet type                                          | Description                                                                                                                                                                                     |
       | --------------- | ----------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
       | `id`            | `STRING`                                              | Required. The unique [identifier for each record](/guides/get-started/concepts#record-id).                                                                                                      |
       | `sparse_values` | `STRUCT<indices: LIST<UINT_32>, values: LIST<FLOAT>>` | Required. A list of floating-point values (sparse values) and a list of integer values (sparse indices) that make up the [sparse vector embedding](/guides/get-started/concepts#sparse-vector). |
       | `metadata`      | `STRING`                                              | Optional. Additional [metadata](/guides/get-started/concepts#metadata) for each record. To omit from specific rows, use `NULL`.                                                                 |

       <Warning>
         The Parquet file cannot contain additional columns.
       </Warning>

       For example:

       ```parquet  theme={null}
       id | sparse_values                                                                                       | metadata
       --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
       1  | {"indices": [ 822745112 1009084850 1221765879 ... ], "values": [1.7958984 0.41577148 2.828125 ...]} | {"year": 1984, "month": 6, "source": "source1", "title": "Example1", "text": "When ..."}
       2  | {"indices": [ 504939989 1293001993 3201939490 ... ], "values": [1.4383747 0.72849722 1.384775 ...]} | {"year": 1990, "month": 4, "source": "source2", "title": "Example2", "text": "Who ..."}
       ```
     </Tab>

     <Tab title="Hybrid index">
       To import into a namespace in a [hybrid index](/guides/search/hybrid-search#use-a-single-hybrid-index), the Parquet file must contain the following columns:

       | Column name     | Parquet type                                          | Description                                                                                                                                                               |
       | --------------- | ----------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
       | `id`            | `STRING`                                              | Required. The unique [identifier for each record](/guides/get-started/concepts#record-id).                                                                                |
       | `values`        | `LIST<FLOAT>`                                         | Required. A list of floating-point values that make up the [dense vector embedding](/guides/get-started/concepts#dense-vector).                                           |
       | `sparse_values` | `STRUCT<indices: LIST<UINT_32>, values: LIST<FLOAT>>` | Optional. A list of floating-point values that make up the [sparse vector embedding](/guides/get-started/concepts#sparse-vector). To omit from specific rows, use `NULL`. |
       | `metadata`      | `STRING`                                              | Optional. Additional [metadata](/guides/get-started/concepts#metadata) for each record. To omit from specific rows, use `NULL`.                                           |

       <Warning>
         The Parquet file cannot contain additional columns.
       </Warning>

       For example:

       ```parquet  theme={null}
       id | values                   | sparse_values                                                                          | metadata
       --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
       1  | [ 3.82  2.48 -4.15 ... ] | {"indices": [1082468256, 1009084850, 1221765879, ...], "values": [2.0, 3.0, 4.0, ...]} | {"year": 1984, "month": 6, "source": "source1", "title": "Example1", "text": "When ..."}
       2  | [ 1.82  3.48 -2.15 ... ] | {"indices": [2225824123, 1293001993, 3201939490, ...], "values": [5.0, 2.0, 3.0, ...]} | {"year": 1990, "month": 4, "source": "source2", "title": "Example2", "text": "Who ..."}
       ```
     </Tab>
   </Tabs>

3. Upload the Parquet files into the relevant subdirectory.

   For example, if you have subdirectories for the namespaces `example_namespace1` and `example_namespace2` and upload 4 Parquet files into each, your directory structure would look as follows after the upload:

   ```
   <BUCKET_OR_CONTAINER_NAME>/
   --/<IMPORT_DIR>/
   ----/example_namespace1/
   ------0.parquet
   ------1.parquet
   ------2.parquet
   ------3.parquet
   ----/example_namespace2/
   ------4.parquet
   ------5.parquet
   ------6.parquet
   ------7.parquet
   ```

## Import records into an index

<Warning>
  Review [import limits](#import-limits) before starting an import.
</Warning>

Use the [`start_import`](/reference/api/latest/data-plane/start_import) operation to start an asynchronous import of vectors from object storage into an index.

* For `uri`, specify the URI of the bucket and import directory containing the namespaces and Parquet files you want to import. For example:

  * Amazon S3: `s3://BUCKET_NAME/IMPORT_DIR`
  * Google Cloud Storage: `gs://BUCKET_NAME/IMPORT_DIR`
  * Azure Blob Storage: `https://STORAGE_ACCOUNT.blob.core.windows.net/CONTAINER_NAME/IMPORT_DIR`

* For `integration_id`, specify the Integration ID of the Amazon S3, Google Cloud Storage, or Azure Blob Storage integration you created. The ID is found on the [Storage integrations](https://app.pinecone.io/organizations/-/projects/-/storage) page of the Pinecone console.

  <Note>
    An Integration ID is not needed to import from a public bucket.
  </Note>

* For `error_mode`, use `CONTINUE` or `ABORT`.

  * With `ABORT`, the operation stops if any records fail to import.
  * With `CONTINUE`, the operation continues on error, but there is not any notification about which records, if any, failed to import. To see how many records were successfully imported, use the [describe an import](#describe-an-import) operation.

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone, ImportErrorMode

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")
  root = "s3://example_bucket/import"

  index.start_import(
      uri=root,
      integration_id="a12b3d4c-47d2-492c-a97a-dd98c8dbefde", # Optional for public buckets
      error_mode=ImportErrorMode.CONTINUE # or ImportErrorMode.ABORT
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  const storageURI = 's3://example_bucket/import';
  const errorMode = 'continue'; // or 'abort'
  const integrationID = 'a12b3d4c-47d2-492c-a97a-dd98c8dbefde'; // Optional for public buckets

  await index.startImport(storageURI, errorMode, integrationID); 
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import io.pinecone.clients.AsyncIndex;
  import org.openapitools.db_data.client.ApiException;
  import org.openapitools.db_data.client.model.ImportErrorMode;
  import org.openapitools.db_data.client.model.StartImportResponse;

  public class StartImport {
      public static void main(String[] args) throws ApiException {
          // Initialize a Pinecone client with your API key
          Pinecone pinecone = new Pinecone.Builder("YOUR_API_KEY").build();

          // Get async imports connection object
          AsyncIndex asyncIndex = pinecone.getAsyncIndexConnection("docs-example");

          // s3 uri
          String uri = "s3://example_bucket/import";

          // Integration ID (optional for public buckets)
          String integrationId = "a12b3d4c-47d2-492c-a97a-dd98c8dbefde";

          // Start an import
          StartImportResponse response = asyncIndex.startImport(uri, integrationId, ImportErrorMode.OnErrorEnum.CONTINUE);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
  	}

      uri := "s3://example_bucket/import"
      errorMode := "continue" // or "abort"
      importRes, err := idxConnection.StartImport(ctx, uri, nil, (*pinecone.ImportErrorMode)(&errorMode))
      if err != nil {
          log.Fatalf("Failed to start import: %v", err)
      }
      fmt.Printf("Import started with ID: %s", importRes.Id)
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var uri = "s3://example_bucket/import";

  var response = await index.StartBulkImportAsync(new StartImportRequest
  {
      Uri = uri,
      IntegrationId = "a12b3d4c-47d2-492c-a97a-dd98c8dbefde",
      ErrorMode = new ImportErrorMode { OnError = ImportErrorModeOnError.Continue }
  });
  ```

  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/bulk/imports" \
    -H 'Api-Key: $YOUR_API_KEY' \
    -H 'Content-Type: application/json' \
    -H 'X-Pinecone-Api-Version: 2025-10' \
    -d '{
          "integrationId": "a12b3d4c-47d2-492c-a97a-dd98c8dbefde",
          "uri": "s3://example_bucket/import",
          "errorMode": {
              "onError": "continue"
              }
          }'
  ```
</CodeGroup>

The response contains an `id` that you can use to [check the status of the import](#list-imports):

```json Response theme={null}
{
   "id": "101"
}
```

Once all the data is loaded, the [index builder](/guides/get-started/database-architecture#index-builder) indexes the records, which usually takes at least 10 minutes. During this indexing process, the expected job status is `InProgress`, but `100.0` percent complete. Once all the imported records are indexed and fully available for querying, the import operation is set to `Completed`.

<Tip>
  You can start a new import using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/indexes). Find the index you want to import into, and click the **ellipsis (...) menu > Import data**.
</Tip>

## Track import progress

The amount of time required for an import depends on various factors, including:

* The number of records to import
* The number of namespaces to import, and the the number of records in each
* The total size (in bytes) of the import

To track an import's progress, check its status bar in the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/import) or use the [`describe_import`](/reference/api/latest/data-plane/describe_import) operation with the import ID:

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  index.describe_import(id="101")
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });


  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  const results = await index.describeImport(id='101');
  console.log(results);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import io.pinecone.clients.AsyncIndex;
  import org.openapitools.db_data.client.ApiException;
  import org.openapitools.db_data.client.model.ImportModel;

  public class DescribeImport {
      public static void main(String[] args) throws ApiException {
          // Initialize a Pinecone client with your API key
          Pinecone pinecone = new Pinecone.Builder("YOUR_API_KEY").build();

          // Get async imports connection object
          AsyncIndex asyncIndex = pinecone.getAsyncIndexConnection("docs-example");

          // Describe import
          ImportModel importDetails = asyncIndex.describeImport("101");

          System.out.println(importDetails);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      importID := "101"

      importDesc, err := idxConnection.DescribeImport(ctx, importID)
      if err != nil {
          log.Fatalf("Failed to describe import: %s - %v", importID, err)
      }
      fmt.Printf("Import ID: %s, Status: %s", importDesc.Id, importDesc.Status)
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var importDetails = await index.DescribeBulkImportAsync("101");
  ```

  ```bash curl theme={null}
  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X GET "https://{INDEX_HOST}/bulk/imports/101" \
    -H 'Api-Key: $YOUR_API_KEY' \
    -H 'X-Pinecone-Api-Version: 2025-10'
  ```
</CodeGroup>

The response contains the import details, including the import `status`, `percent_complete`, and `records_imported`:

```json Response theme={null}
{
  "id": "101",
  "uri": "s3://example_bucket/import",
  "status": "InProgress",
  "created_at": "2024-08-19T20:49:00.754Z",
  "finished_at": "2024-08-19T20:49:00.754Z",
  "percent_complete": 42.2,
  "records_imported": 1000000
}
```

If the import fails, the response contains an `error` field with the reason for the failure. See the [Troubleshooting](#troubleshooting) section for more information.

```json Response theme={null}
{
  "id": "102",
  "uri": "s3://example_bucket/import",
  "status": "Failed",
  "percent_complete": 0.0,
  "records_imported": 0,
  "created_at": "2025-08-21T11:29:47.886797+00:00",
  "error": "User error: The namespace \"namespace1\" already exists. Imports are only allowed into nonexistent namespaces.",
  "finished_at": "2025-08-21T11:30:05.506423+00:00"
}
```

## Manage imports

### List imports

Use the [`list_imports`](/reference/api/latest/data-plane/list_imports) operation to list all of the recent and ongoing imports. By default, the operation returns up to 100 imports per page. If the `limit` parameter is passed, the operation returns up to that number of imports per page instead. For example, if `limit=3`, up to 3 imports are returned per page. Whenever there are additional imports to return, the response includes a `pagination_token` for fetching the next page of imports.

<Tabs>
  <Tab title="Python SDK">
    When using the Python SDK, `list_import` paginates automatically.

    ```python Python theme={null}
    from pinecone import Pinecone, ImportErrorMode

    pc = Pinecone(api_key="YOUR_API_KEY")

    # To get the unique host for an index, 
    # see https://docs.pinecone.io/guides/manage-data/target-an-index
    index = pc.Index(host="INDEX_HOST")

    # List using a generator that handles pagination
    for i in index.list_imports():
        print(f"id: {i.id} status: {i.status}")

    # List using a generator that fetches all results at once
    operations = list(index.list_imports())
    print(operations)
    ```

    ```json Response theme={null}
    {
      "data": [
        {
          "id": "1",
          "uri": "s3://BUCKET_NAME/PATH/TO/DIR",
          "status": "Pending",
          "started_at": "2024-08-19T20:49:00.754Z",
          "finished_at": "2024-08-19T20:49:00.754Z",
          "percent_complete": 42.2,
          "records_imported": 1000000
        }
      ],
      "pagination": {
        "next": "Tm90aGluZyB0byBzZWUgaGVyZQo="
      }
    }
    ```

    <Tip>
      You can view the list of imports for an index in the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/indexes/). Select the index and navigate to the **Imports** tab.
    </Tip>
  </Tab>

  <Tab title="Other SDKs">
    When using the Node.js SDK, Java SDK, Go SDK, .NET SDK, or REST API to list recent and ongoing imports, you must manually fetch each page of results. To view the next page of results, include the `paginationToken` provided in the response.

    <CodeGroup>
      ```javascript JavaScript theme={null}
      import { Pinecone } from '@pinecone-database/pinecone';

      const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      const index = pc.index("INDEX_NAME", "INDEX_HOST")

      const results = await index.listImports({ limit: 10, paginationToken: 'Tm90aGluZyB0byBzZWUgaGVyZQo' });
      console.log(results);
      ```

      ```java Java theme={null}
      import io.pinecone.clients.Pinecone;
      import io.pinecone.clients.AsyncIndex;
      import org.openapitools.db_data.client.ApiException;
      import org.openapitools.db_data.client.model.ListImportsResponse;

      public class ListImports {
          public static void main(String[] args) throws ApiException {
              // Initialize a Pinecone client with your API key
              Pinecone pinecone = new Pinecone.Builder("YOUR_API_KEY").build();

              // Get async imports connection object
              AsyncIndex asyncIndex = pinecone.getAsyncIndexConnection("docs-example");

              // List imports
              ListImportsResponse response = asyncIndex.listImports(10, "Tm90aGluZyB0byBzZWUgaGVyZQo");

              System.out.println(response);
          }
      }
      ```

      ```go Go theme={null}
      package main

      import (
          "context"
          "fmt"
          "log"

          "github.com/pinecone-io/go-pinecone/v4/pinecone"
      )

      func main() {
          ctx := context.Background()

          pc, err := pinecone.NewClient(pinecone.NewClientParams{
              ApiKey: "YOUR_API_KEY",
          })
          if err != nil {
              log.Fatalf("Failed to create Client: %v", err)
          }

          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST"})
          if err != nil {
              log.Fatalf("Failed to create IndexConnection for Host: %v", err)
        	}

          limit := int32(10)
          firstImportPage, err := idxConnection.ListImports(ctx, &limit, nil)
          if err != nil {
              log.Fatalf("Failed to list imports: %v", err)
          }
          fmt.Printf("First page of imports: %+v", firstImportPage.Imports)

          paginationToken := firstImportPage.NextPaginationToken
          nextImportPage, err := idxConnection.ListImports(ctx, &limit, paginationToken)
          if err != nil {
              log.Fatalf("Failed to list imports: %v", err)
          }
          fmt.Printf("Second page of imports: %+v", nextImportPage.Imports)
      }
      ```

      ```csharp C# theme={null}
      using Pinecone;

      var pinecone = new PineconeClient("YOUR_API_KEY");

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      var index = pinecone.Index(host: "INDEX_HOST");

      var imports = await index.ListBulkImportsAsync(new ListBulkImportsRequest
      {
          Limit = 10,
          PaginationToken = "Tm90aGluZyB0byBzZWUgaGVyZQo"
      });
      ```

      ```bash curl theme={null}
      # To get the unique host for an index,
      # see https://docs.pinecone.io/guides/manage-data/target-an-index
      PINECONE_API_KEY="YOUR_API_KEY"
      INDEX_HOST="INDEX_HOST"

      curl -X GET "https://$INDEX_HOST/bulk/imports?paginationToken==Tm90aGluZyB0byBzZWUgaGVyZQo" \
        -H 'Api-Key: $YOUR_API_KEY' \
        -H 'X-Pinecone-Api-Version: 2025-10'
      ```
    </CodeGroup>
  </Tab>
</Tabs>

### Cancel an import

The [`cancel_import`](/reference/api/latest/data-plane/cancel_import) operation cancels an import if it is not yet finished. It has no effect if the import is already complete.

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  index.cancel_import(id="101")
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  await index.cancelImport(id='101');
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import io.pinecone.clients.AsyncIndex;
  import org.openapitools.db_data.client.ApiException;

  public class CancelImport {
      public static void main(String[] args) throws ApiException {
          // Initialize a Pinecone client with your API key
          Pinecone pinecone = new Pinecone.Builder("YOUR_API_KEY").build();

          // Get async imports connection object
          AsyncIndex asyncIndex = pinecone.getAsyncIndexConnection("docs-example");

          // Cancel import
          asyncIndex.cancelImport("2");
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      importID := "101"

      err = idxConnection.CancelImport(ctx, importID)
      if err != nil {
          log.Fatalf("Failed to cancel import: %s", importID)
      }

      importDesc, err := idxConnection.DescribeImport(ctx, importID)
      if err != nil {
          log.Fatalf("Failed to describe import: %s - %v", importID, err)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var cancelResponse = await index.CancelBulkImportAsync("101");
  ```

  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X DELETE "https://{INDEX_HOST}/bulk/imports/101" \
    -H 'Api-Key: $YOUR_API_KEY' \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

```json Response theme={null}
{}
```

<Tip>
  You can cancel your import using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/import). To cancel an ongoing import, select the index you are importing into and navigate to the **Imports** tab. Then, click the **ellipsis (...) menu > Cancel**.
</Tip>

## Import limits

<Note>
  If your import exceeds these limits, you'll get an `Exceeds system limit` error. Pinecone can help unblock these imports quickly. [Contact Pinecone support](https://app.pinecone.io/organizations/-/settings/support/ticket) for assistance.
</Note>

| Metric                    | Limit   |
| :------------------------ | :------ |
| Max namespaces per import | 10,000  |
| Max size per namespace    | 500 GB  |
| Max files per import      | 100,000 |
| Max size per file         | 10 GB   |

Also:

* You cannot import data from an AWS S3 bucket into a Pinecone index hosted on GCP or Azure.
* You cannot import data from S3 Express One Zone storage.
* You cannot import data into an existing namespace.
* When importing data into the `__default__` namespace of an index, the default namespace must be empty.
* Each import takes at least 10 minutes to complete.
* When importing into an [index with integrated embedding](/guides/index-data/indexing-overview#vector-embedding), records must contain vectors, not text. To add records with text, you must use [upsert](/guides/index-data/upsert-data).

## Troubleshooting

When an import fails, you'll see an error message with the reason for the failure in the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/import) or in the response to the [describe an import](/reference/api/latest/data-plane/describe_import) operation.

<AccordionGroup>
  <Accordion title="Namespace already exists">
    You cannot import data into an existing namespace. If your import directory structure contains a folder with the name of an existing namespace in your index, the import will fail with the following error:

    ```
    User error: The namespace "example-namespace" already exists. Imports are only allowed into nonexistent namespaces.
    ```

    To fix this, rename the folder to use a namespace name that does not yet exist.
  </Accordion>

  <Accordion title="No namespace found">
    In object storage, your directory structure must be as follows:

    ```
    example_bucket/
    --/imports/
    ----/example_namespace1/
    ------0.parquet
    ------1.parquet
    ------2.parquet
    ------3.parquet
    ----/example_namespace2/
    ------4.parquet
    ------5.parquet
    ------6.parquet
    ------7.parquet
    ```

    If a Parquet file is not nested under a namespace subdirectory, the import will fail with the following error:

    ```
    User error: \"test-import/0.parquet\": No namespace detected. Each file should be nested under a subdirectory of the URI prefix. This indicates which namespace it should be imported into.
    ```

    To fix this, move the Parquet file to a namespace subdirectory.
  </Accordion>

  <Accordion title="Parquet files not found">
    Each namespace subdirectory must contain Parquet files with data to import. If a namespace subdirectory does not include Parquet files, the import will fail with the following error:

    ```
    User error: No Parquet files found under \"gs://example_bucket/imports\". Files must be stored with the specified bucket prefix.
    ```

    To fix this, add Parquet files to the namespace subdirectory.
  </Accordion>

  <Accordion title="Invalid import URI">
    In your [start import](/reference/api/latest/data-plane/start_import) request, the import `uri` must specify only the bucket and import directory containing the namespaces and Parquet files you want to import. If the `uri` also contains a namespaces directory or a Parquet filename, the import will fail with the following error:

    ```
    User error: \"test-import/0.parquet\": It looks like you specified a complete path to a parquet file as the URI prefix to import from. Note that the URI prefix should give an ancestor directory with subdirectories to specify each namespace to import into. See https://docs.pinecone.io/guides/data/understanding-imports#directory-structure.
    ```

    To fix this, remove the namespaces directory or Parquet filename from the `uri`.
  </Accordion>

  <Accordion title="Invalid Parquet files">
    When a Parquet file is not formatted correctly, the import will fail with a message like one of the following:

    ```shell File schema errors theme={null}
    Missing required column \"{0}\"
    Unsupported column \"{0}\"
    ```

    ```shell File corruption errors theme={null}
    Parquet footer could not be parsed. Are you sure this is valid parquet?
    ```

    ```shell Type errors theme={null}
    The expected data type for column \"{column}\" is \"{expected}\", but got \"{given}\"
    The expected data type for metadata is a JSON encoded string in UTF-8 format, but got \"{given}\"
    ```

    These errors are returned for both `CONTINUE` and `ABORT` error modes.

    To fix these errors, check the specific error message and follow the instructions in the [Prepare your data](#prepare-your-data) section.
  </Accordion>

  <Accordion title="Invalid records">
    When the `error_mode` is `ABORT` and a file contains invalid records, the import will stop processing on the first invalid record and return an error message identifying the file name and row:

    ```
    User error: error reading record (file \"/0.parquet\", row 0):
    ```

    This will be followed by an error message identifying the specific issue. For example:

    ```shell Missing values theme={null}
    missing required values in column \"{column}\"
    ```

    ```shell Invalid metadata  theme={null}
    Failed to parse metadata: {msg}
    ```

    ```shell Invalid vectors theme={null}
    Upserting dense vectors is not supported for sparse indexes
    ```

    When the `error_mode` is `CONTINUE`, the import will skip individual invalid records. However, if all records are invalid and skipped (for example, the vector type in the file does not match the vector type of the index), the import will fail with a general message:

    ```
    User error: No vectors added, all rows were skipped for namespace: example-namespace
    ```

    To fix these errors, check the specific error message and follow the instructions in the [Prepare your data](#prepare-your-data) section.
  </Accordion>

  <Accordion title="Duplicate records">
    When your import contains duplicate vectors (records with identical vector values), the duplicates are marked as skipped and not imported. Only one occurrence of each unique vector is added to the index.

    This applies to both `CONTINUE` and `ABORT` error modes:

    * With `ABORT`: The import fails when it encounters a duplicate vector within the import.
    * With `CONTINUE`: The import proceeds, skipping duplicate records silently.

    **Example scenario:**
    If your Parquet file contains:

    ```parquet  theme={null}
    id | values
    ---|---------
    1  | [0.1, 0.2, 0.3]
    2  | [0.1, 0.2, 0.3]  ← Duplicate of record 1, will be skipped
    3  | [0.4, 0.5, 0.6]
    ```

    Only records 1 and 3 will be imported.

    To prevent this from happening, deduplicate your source data before creating Parquet files by removing records with identical vector values.
  </Accordion>
</AccordionGroup>

## See also

* [Integrate with Amazon S3](/guides/operations/integrations/integrate-with-amazon-s3)
* [Integrate with Google Cloud Storage](/guides/operations/integrations/integrate-with-google-cloud-storage)
* [Integrate with Azure Blob Storage](/guides/operations/integrations/integrate-with-azure-blob-storage)
* [Pinecone's pricing](https://www.pinecone.io/pricing/)

---

# Source: https://docs.pinecone.io/guides/optimize/increase-relevance.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Increase search relevance

> Learn techniques to improve search result quality.

This page describes helpful techniques for improving search accuracy and relevance.

## Rerank results

[Reranking](/guides/search/rerank-results) is used as part of a two-stage vector retrieval process to improve the quality of results. You first query an index for a given number of relevant results, and then you send the query and results to a reranking model. The reranking model scores the results based on their semantic relevance to the query and returns a new, more accurate ranking. This approach is one of the simplest methods for improving quality in retrieval augmented generation (RAG) pipelines.

Pinecone provides [hosted reranking models](/guides/search/rerank-results#reranking-models) so it's easy to manage two-stage vector retrieval on a single platform. You can use a hosted model to rerank results as an integrated part of a query, or you can use a hosted model to rerank results as a standalone operation.

## Filter by metadata

Every [record](/guides/get-started/concepts#record) in an index must contain an ID and a dense or sparse vector, depending on the [type of index](/guides/index-data/indexing-overview#indexes). In addition, you can include metadata key-value pairs to store related information or context. When you search the index, you can then include a metadata filter to limit the search to records matching a filter expression.

For example, if an index contains records about books, you could use a metadata field to associate each record with a genre, like `"genre": "fiction"` or `"genre": "poetry"`. When you query the index, you could then use a metadata filter to limit your search to records related to a specific genre.

For more details, see [Filter by metadata](/guides/search/filter-by-metadata).

## Use hybrid search

[Semantic search](/guides/search/semantic-search) and [lexical search](/guides/search/lexical-search) are powerful information retrieval techniques, but each has notable limitations. For example, semantic search can miss results based on exact keyword matches, especially in scenarios involving domain-specific terminology, while lexical search can miss results based on relationships, such as synonyms and paraphrases.

To work around these limitations, you can perform hybrid search, which combines semantic and lexical search. There are two ways to do this:

* [Use a single hybrid index](/guides/search/hybrid-search#use-a-single-hybrid-index). This is the **recommended** approach for most use cases because you make requests to a single index, the linkage between dense and sparse vectors is implicit, and you can perform hybrid queries with a single request.
* [Use separate dense and sparse indexes](/guides/search/hybrid-search#use-separate-dense-and-sparse-indexes). This approach provides more flexibility but requires managing two indexes, maintaining linkages between vectors, and querying each index separately before merging results.

For more details, including guidance on choosing the right approach, see [Hybrid search](/guides/search/hybrid-search).

## Explore chunking strategies

You can chunk your content in different ways to get better results. Consider factors like the length of the content, the complexity of queries, and how results will be used in your application.

For more details, see [Chunking strategies](https://www.pinecone.io/learn/chunking-strategies/).

---

# Source: https://docs.pinecone.io/guides/optimize/increase-throughput.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Increase throughput

> Learn techniques to improve data operation performance and query throughput.

## Import from object storage

[Importing from object storage](/guides/index-data/import-data) is the most efficient and cost-effective method to load large numbers of records into an index. You store your data as Parquet files in object storage, integrate your object storage with Pinecone, and then start an asynchronous, long-running operation that imports and indexes your records.

## Upsert in batches

[Upserting in batches](/guides/index-data/upsert-data#upsert-in-batches) is another efficient way to ingest large numbers of records (up to 1000 per batch). Batch upserting is also a good option if you cannot work around bulk import's current [limitations](/guides/index-data/import-data#import-limits).

## Upsert/search in parallel

Pinecone is thread-safe, so you can send multiple [upsert](/guides/index-data/upsert-data#upsert-in-parallel) requests and multiple [query](/guides/search/semantic-search#parallel-queries) requests in parallel to help increase throughput.

## Python SDK options

### Use gRPC

Use the [Python SDK with gRPC extras](/reference/sdks/python/overview) to run data operations such as upserts and queries over [gRPC](https://grpc.io/) rather than HTTP for a modest performance improvement.

### Upsert from a dataframe

To quickly ingest data when using the Python SDK, use the [`upsert_from_dataframe` method](/reference/sdks/python/overview#upsert-from-a-dataframe). The method includes retry logic and `batch_size`, and is performant especially with Parquet file data sets.

## See also

Read more about [high-throughput optimizations](https://www.pinecone.io/blog/working-at-scale/) on our blog.

---

# Source: https://docs.pinecone.io/troubleshooting/index-creation-error-max-serverless.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Serverless index creation error - max serverless indexes

## Problem

Each project is limited to 20 serverless indexes. Trying to create more than 20 serverless indexes in a project raises the following `403 (FORBIDDEN)` error:

```console console theme={null}
This project already contains 20 serverless indexes, the maximum per project. 
Delete any unused indexes and try again, or create a new project for more serverless indexes. 
For additional help, please contact support@pinecone.io.
```

## Solution

[Delete any unused serverless indexes](/guides/manage-data/manage-indexes#delete-an-index) in the project and try again, or create a new project to hold additional serverless indexes.

Also consider using [namespaces](/guides/index-data/indexing-overview#namespaces) to partition vectors of the same dimensionality within a single index. Namespaces can help speed up queries as well as comply with [multitenancy](/guides/index-data/implement-multitenancy) requirements.

---

# Source: https://docs.pinecone.io/troubleshooting/index-creation-error-missing-spec.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Index creation error - missing spec parameter

## Problem

Using the [new API](/reference/api), creating an index requires passing appropriate values into the `spec` parameter. Without this `spec` parameter, the `create_index` method raises the following error:

```console console theme={null}
TypeError: Pinecone.create_index() missing 1 required positional argument: 'spec'
```

## Solution

Set the `spec` parameter. For guidance on how to set this parameter, see [Create an index](/guides/index-data/create-an-index#create-a-serverless-index).

---

# Source: https://docs.pinecone.io/guides/index-data/indexing-overview.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Indexing overview

> Understand key concepts related to indexing data in Pinecone.

export const word_0 = "records"

## Indexes

In Pinecone, you store vector data in indexes. There are two types of indexes: dense and sparse.

### Dense indexes

Dense indexes store dense vectors, which are a series of numbers that represent the meaning and relationships of text, images, or other types of data. Each number in a dense vector corresponds to a point in a multidimensional space. Vectors that are closer together in that space are semantically similar.

When you query a dense index, Pinecone retrieves the dense vectors that are the most semantically similar to the query. This is often called **semantic search**, nearest neighbor search, similarity search, or just vector search.

Learn more:

* [Create a dense index](/guides/index-data/create-an-index#create-a-dense-index)
* [Upsert dense vectors](/guides/index-data/upsert-data#upsert-dense-vectors)
* [Semantic search](/guides/search/semantic-search)

### Sparse indexes

Sparse indexes store sparse vectors, which are a series of numbers that represent the words or phrases in a document. Sparse vectors have a very large number of dimensions, where only a small proportion of values are non-zero. The dimensions represent words from a dictionary, and the values represent the importance of these words in the document.

When you search a sparse index, Pinecone retrieves the sparse vectors that most exactly match the words or phrases in the query. Query terms are scored independently and then summed, with the most similar records scored highest. This is often called **lexical search** or **keyword search**.

Learn more:

* [Create a sparse index](/guides/index-data/create-an-index#create-a-sparse-index)
* [Upsert sparse vectors](/guides/index-data/upsert-data#upsert-sparse-vectors)
* [Lexical search](/guides/search/lexical-search)

<Tip>
  Semantic search can miss results based on exact keyword matches, while lexical search can miss results based on relationships. To work around these limitations, you can perform [hybrid search](/guides/search/hybrid-search).
</Tip>

#### Limitations

<Note>
  These limitations are subject to change during the public preview period.
</Note>

Sparse indexes have the following limitations:

* Max non-zero values per sparse vector: 1000
* Max upserts per second per sparse index: 10
* Max queries per second per sparse index: 100
* Max `top_k` value per query: 1000

  <Note>
    You may get fewer than `top_k` results if `top_k` is larger than the number of sparse vectors in your index that match your query. That is, any vectors where the dotproduct score is `0` will be discarded.
  </Note>
* Max query results size: 4MB

## Namespaces

Within an index, records are partitioned into namespaces, and all [upserts](/guides/index-data/upsert-data), [queries](/guides/search/search-overview), and other [data operations](/guides/index-data/upsert-data) always target one namespace. This has two main benefits:

* **Multitenancy:** When you need to isolate data between customers, you can use one namespace per customer and target each customer's writes and queries to their dedicated namespace. See [Implement multitenancy](/guides/index-data/implement-multitenancy) for end-to-end guidance.

* **Faster queries:** When you divide records into namespaces in a logical way, you speed up queries by ensuring only relevant records are scanned. The same applies to fetching records, listing record IDs, and other data operations.

Namespaces are created automatically during [upsert](/guides/index-data/upsert-data). If a namespace doesn't exist, it is created implicitly.

<Note>
  While the Standard and Enterprise plans support up to [100,000 namespaces per index](/reference/api/database-limits#namespaces-per-serverless-index), Pinecone can accommodate million-scale namespaces and beyond for specific use cases. If your application requires more than 100,000 namespaces, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).
</Note>

<img className="block max-w-full dark:hidden" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=641c2aa9a3238bf70698c583097c1f29" data-og-width="1400" width="1400" data-og-height="880" height="880" data-path="images/quickstart-upsert.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=5b394ed73cf8248f9a9ae8f9d3cdbd2d 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=3bd0b45458ebbcab40605f149b5847d5 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=9c1e1b064228344b3caf4c0a1aa8ab82 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=d40dd481d2e7cc8882d766d6df59fcba 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=bf6be475e7bed3453299f6bbecd6aa54 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=46c9efa0b08b8ac3c907a121289c19f2 2500w" />

<img className="hidden max-w-full dark:block" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=14a3e6c2847455db0821ebbf9bd51df9" data-og-width="1400" width="1400" data-og-height="880" height="880" data-path="images/quickstart-upsert-dark.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=229b97b3ea4581730afab68709201084 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=02fcdee17c689d31769c145fb86f259b 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=dd13a98f1154ffd8c4cc7cd4d37c0d33 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=b13689e423b344828452efcf91c737b4 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=319293ecf19a483936595aaebfb5cb31 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/quickstart-upsert-dark.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=78ad3d1499ae4d57ff9e9cd0a8e56093 2500w" />

## Vector embedding

[Dense vectors](/guides/get-started/concepts#dense-vector) and [sparse vectors](/guides/get-started/concepts#sparse-vector) are the basic units of data in Pinecone and what Pinecone was specially designed to store and work with. Dense vectors represents the semantics of data such as text, images, and audio recordings, while sparse vectors represent documents or queries in a way that captures keyword information.

To transform data into vector format, you use an embedding model. You can either use Pinecone's integrated embedding models to convert your source data to vectors automatically, or you can use an external embedding model and bring your own vectors to Pinecone.

### Integrated embedding

1. [Create an index](/guides/index-data/create-an-index) that is integrated with one of Pinecone's [hosted embedding models](/guides/index-data/create-an-index#embedding-models).
2. [Upsert](/guides/index-data/upsert-data) your source text. Pinecone uses the integrated model to convert the text to vectors automatically.
3. [Search](/guides/search/search-overview) with a query text. Again, Pinecone uses the integrated model to convert the text to a vector automatically.

<Note>
  Indexes with integrated embedding do not support [updating](/guides/manage-data/update-data) or [importing](/guides/index-data/import-data) with text.
</Note>

### Bring your own vectors

1. Use an embedding model to convert your text to vectors. The model can be [hosted by Pinecone](/reference/api/latest/inference/generate-embeddings) or an external provider.
2. [Create an index](/guides/index-data/create-an-index) that matches the characteristics of the model.
3. [Upsert](/guides/index-data/upsert-data) your vectors directly.
4. Use the same external embedding model to convert a query to a vector.
5. [Search](/guides/search/search-overview) with your query vector directly.

## Data ingestion

<Tip>
  To control costs when ingesting large datasets (10,000,000+ records), use [import](/guides/index-data/import-data) instead of upsert.
</Tip>

There are two ways to ingest data into an index:

* [Importing from object storage](/guides/index-data/import-data) is the most efficient and cost-effective way to load large numbers of records into an index. You store your data as Parquet files in object storage, integrate your object storage with Pinecone, and then start an asynchronous, long-running operation that imports and indexes your records.

* [Upserting](/guides/index-data/upsert-data) is intended for ongoing writes to an index. [Batch upserting](/guides/index-data/upsert-data#upsert-in-batches) can improve throughput performance and is a good option for larger numbers of records (up to 1000 per batch) if you cannot work around import's current limitations.

## Metadata

Every [record](/guides/get-started/concepts#record) in an index must contain an ID and a vector. In addition, you can include metadata key-value pairs to store additional information or context. When you query the index, you can then include a [metadata filter](/guides/search/filter-by-metadata) to limit the search to records matching a filter expression. Searches without metadata filters do not consider metadata and search the entire namespace.

### Metadata format

* Metadata fields must be key-value pairs in a flat JSON object. Nested JSON objects are not supported.
* Keys must be strings and must not start with a `$`.
* Values must be one of the following data types:
  * String
  * Integer (converted to a 64-bit floating point by Pinecone)
  * Floating point
  * Boolean (`true`, `false`)
  * List of strings
* Null metadata values aren't supported. Instead of setting a key to `null`, remove the key from the metadata payload.

**Examples**

<CodeGroup>
  ```json Valid metadata theme={null}
  {
    "document_id": "document1",
    "document_title": "Introduction to Vector Databases",
    "chunk_number": 1,
    "chunk_text": "First chunk of the document content...",
    "is_public": true,
    "tags": ["beginner", "database", "vector-db"],
    "scores": ["85", "92"]
  }
  ```

  ```json Invalid metadata theme={null}
  {
    "document": {       // Nested JSON objects are not supported
      "document_id": "document1",
      "document_title": "Introduction to Vector Databases",
    },
    "$chunk_number": 1, // Keys must not start with a `$`
    "chunk_text": null, // Null values are not supported
    "is_public": true,
    "tags": ["beginner", "database", "vector-db"],
    "scores": [85, 92]  // Lists of non-strings are not supported
  }
  ```
</CodeGroup>

### Metadata size

Pinecone supports 40KB of metadata per record.

### Metadata filter expressions

Pinecone's filtering language supports the following operators:

| Operator  | Function                                                                                                                           | Supported types         |
| :-------- | :--------------------------------------------------------------------------------------------------------------------------------- | :---------------------- |
| `$eq`     | Matches {word_0} with metadata values that are equal to a specified value. Example: `{"genre": {"$eq": "documentary"}}`            | Number, string, boolean |
| `$ne`     | Matches {word_0} with metadata values that are not equal to a specified value. Example: `{"genre": {"$ne": "drama"}}`              | Number, string, boolean |
| `$gt`     | Matches {word_0} with metadata values that are greater than a specified value. Example: `{"year": {"$gt": 2019}}`                  | Number                  |
| `$gte`    | Matches {word_0} with metadata values that are greater than or equal to a specified value. Example:`{"year": {"$gte": 2020}}`      | Number                  |
| `$lt`     | Matches {word_0} with metadata values that are less than a specified value. Example: `{"year": {"$lt": 2020}}`                     | Number                  |
| `$lte`    | Matches {word_0} with metadata values that are less than or equal to a specified value. Example: `{"year": {"$lte": 2020}}`        | Number                  |
| `$in`     | Matches {word_0} with metadata values that are in a specified array. Example: `{"genre": {"$in": ["comedy", "documentary"]}}`      | String, number          |
| `$nin`    | Matches {word_0} with metadata values that are not in a specified array. Example: `{"genre": {"$nin": ["comedy", "documentary"]}}` | String, number          |
| `$exists` | Matches {word_0} with the specified metadata field. Example: `{"genre": {"$exists": true}}`                                        | Number, string, boolean |
| `$and`    | Joins query clauses with a logical `AND`. Example: `{"$and": [{"genre": {"$eq": "drama"}}, {"year": {"$gte": 2020}}]}`             | -                       |
| `$or`     | Joins query clauses with a logical `OR`. Example: `{"$or": [{"genre": {"$eq": "drama"}}, {"year": {"$gte": 2020}}]}`               | -                       |

<Note>
  Only `$and` and `$or` are allowed at the top level of the query expression.
</Note>

<Note>
  Each `$in` or `$nin` operator accepts a maximum of 10,000 values. Exceeding this limit will cause the request to fail. For more information, see [Metadata filter limits](/reference/api/database-limits#metadata-filter-limits).
</Note>

For example, the following has a `"genre"` metadata field with a list of strings:

```JSON JSON theme={null}
{ "genre": ["comedy", "documentary"] }
```

This means `"genre"` takes on both values, and requests with the following filters will match:

```JSON JSON theme={null}
{"genre":"comedy"}

{"genre": {"$in":["documentary","action"]}}

{"$and": [{"genre": "comedy"}, {"genre":"documentary"}]}
```

However, requests with the following filter will **not** match:

```JSON JSON theme={null}
{ "$and": [{ "genre": "comedy" }, { "genre": "drama" }] }
```

Additionally, requests with the following filters will **not** match because they are invalid. They will result in a compilation error:

```json JSON theme={null}
# INVALID QUERY:
{"genre": ["comedy", "documentary"]}
```

```json JSON theme={null}
# INVALID QUERY:
{"genre": {"$eq": ["comedy", "documentary"]}}
```

---

# Source: https://docs.pinecone.io/reference/cli/installation.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# CLI installation

<Note>
  This feature is in [public preview](/release-notes/feature-availability).
</Note>

This document describes how to install the Pinecone CLI on your local machine.

## Platforms

Pinecone's CLI is available for the following platforms:

* **macOS**: Intel (x86\_64) and Apple Silicon (ARM64)
* **Linux**: x86\_64, ARM64, and i386 architectures
* **Windows**: x86\_64 and i386 architectures

## Installation

### Homebrew (macOS)

The most convenient way to install the CLI on Mac:

```bash  theme={null}
brew tap pinecone-io/tap
brew install pinecone-io/tap/pinecone
```

To upgrade to the latest version:

```bash  theme={null}
brew update
brew upgrade pinecone
```

### Direct download (all platforms)

Pre-built binaries for all supported platforms are available on the [GitHub Releases page](https://github.com/pinecone-io/cli/releases).

### Build from source

To build from source, see the [CONTRIBUTING.md](https://github.com/pinecone-io/cli/blob/main/CONTRIBUTING.md) file for detailed instructions. To build from source, you need Go v1.23+.

---

# Source: https://docs.pinecone.io/integrations/instill.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Instill AI

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Instill AI specializes in developing cutting-edge solutions for data, models, and pipeline orchestration. Their flagship source-available product, Instill Core, is a no-code/low-code platform designed to facilitate the development, deployment, and management of AI workflows. By simplifying the integration of various AI models and data sources, they enable businesses to harness the power of AI without requiring extensive technical expertise. Their solutions cater to a wide range of applications, from predictive analytics and autonomous AI agents to enterprise private knowledge bases, AI assistants, and beyond.

The Pinecone integration with Instill allows developers to incorporate its API for vector upsert and query tasks into AI pipelines. Developers configure their Pinecone component within Instill Core by providing the necessary API key and base URL. They can then perform tasks such as querying vector similarities to retrieve the most relevant results, complete with metadata and similarity scores, or upserting new vector data to keep their datasets up-to-date. This integration enables the addition of knowledge to LLMs via Retrieval Augmented Generation (RAG), significantly enhancing the capabilities of autonomous agents, chatbots, question-answering systems, and multi-agent systems.

<PrimarySecondaryCTA primaryHref={"https://www.instill.tech/docs/component/data/pinecone"} primaryLabel={"Get started"} />

---

# Source: https://docs.pinecone.io/guides/operations/integrations/integrate-with-amazon-s3.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Integrate with Amazon S3

> Set up Amazon S3 integrationfor data import and audit logs.

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Standard and Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

This page shows you how to integrate Pinecone with an Amazon S3 bucket. Once your integration is set up, you can use it to [import data](/guides/index-data/import-data) from your Amazon S3 bucket into a Pinecone index hosted on AWS, or to [export audit logs](/guides/production/configure-audit-logs) to your Amazon S3 bucket.

## Before you begin

Ensure you have the following:

* A [Pinecone account](https://app.pinecone.io/).
* An [Amazon S3 bucket](https://docs.aws.amazon.com/AmazonS3/latest/userguide/creating-buckets.html).

## 1. Create an IAM policy

In the [AWS IAM console](https://console.aws.amazon.com/iam/home):

1. In the navigation pane, click **Policies**.
2. Click **Create policy**.
3. In **Select a service** section, select **S3**.
4. Select the following actions to allow:
   * `ListBucket`: Permission to list some or all of the objects in an S3 bucket. Required for [importing data](/guides/index-data/import-data) and [exporting audit logs](/guides/production/configure-audit-logs).
   * `GetObject`: Permission to retrieve objects from an S3 bucket. Required for [importing data](/guides/index-data/import-data).
   * `PutObject`: Permission to add an object to an S3 bucket. Required for [exporting audit logs](/guides/production/configure-audit-logs).
5. In the **Resources** section, select **Specific**.
6. For the **bucket**, specify the ARN of the bucket you created. For example: `arn:aws:s3:::example-bucket-name`
7. For the **object**, specify an object ARN as the target resource. For example: `arn:aws:s3:::example-bucket-name/*`
8. Click **Next**.
9. Specify the name of your policy. For example:  "Pinecone-S3-Access".
10. Click **Create policy**.

### Targeting a subdirectory (optional)

To write [audit logs](/guides/production/configure-audit-logs) to a specific subdirectory within your S3 bucket (e.g., `my-bucket/pinecone-logs/`), you need to configure your IAM policy differently for `ListBucket` vs. object-level actions:

1. For `ListBucket`, use a **Condition** block with `StringLike` to specify the prefix. Include both the directory path with and without the trailing wildcard:

   ```json  theme={null}
   {
       "Sid": "ListBucketWithPrefix",
       "Effect": "Allow",
       "Action": "s3:ListBucket",
       "Resource": "arn:aws:s3:::example-bucket-name",
       "Condition": {
           "StringLike": {
               "s3:prefix": [
                   "pinecone-logs/",
                   "pinecone-logs/*"
               ]
           }
       }
   }
   ```

2. For `PutObject` and `GetObject`, use the **Resource** specifier with the subdirectory path:

   ```json  theme={null}
   {
       "Sid": "ObjectActionsInSubdirectory",
       "Effect": "Allow",
       "Action": [
           "s3:PutObject",
           "s3:GetObject"
       ],
       "Resource": "arn:aws:s3:::example-bucket-name/pinecone-logs/*"
   }
   ```

**Complete example policy for subdirectory access:**

```json  theme={null}
{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Sid": "ListBucketWithPrefix",
            "Effect": "Allow",
            "Action": "s3:ListBucket",
            "Resource": "arn:aws:s3:::example-bucket-name",
            "Condition": {
                "StringLike": {
                    "s3:prefix": [
                        "pinecone-logs/",
                        "pinecone-logs/*"
                    ]
                }
            }
        },
        {
            "Sid": "ObjectActionsInSubdirectory",
            "Effect": "Allow",
            "Action": [
                "s3:PutObject",
                "s3:GetObject"
            ],
            "Resource": "arn:aws:s3:::example-bucket-name/pinecone-logs/*"
        }
    ]
}
```

<Note>
  The key difference is that `ListBucket` operates on the bucket resource and uses conditions to filter by prefix, while object-level actions (`PutObject`, `GetObject`) operate directly on object resources specified in the ARN.
</Note>

## 2. Set up access using an IAM role

In the [AWS IAM console](https://console.aws.amazon.com/iam/home):

1. In the navigation pane, click **Roles**.

2. Click **Create role**.

3. In the **Trusted entity type** section, select **AWS account**.

4. Select **Another AWS account**.

5. Enter the Pinecone AWS VPC account ID: `713131977538`

6. Click **Next**.

7. Select the [policy you created](#1-create-an-iam-policy).

8. Click **Next**.

9. Specify the role name. For example: "Pinecone".

10. Click **Create role**.

11. Click the role you created.

12. On the **Summary** page for the role, find the **ARN**.

    For example: `arn:aws:iam::123456789012:role/PineconeAccess`

13. Copy the **ARN**.

    You will need to enter the ARN into Pinecone later.

## 3. Add a storage integration

<Note>
  This step is required for [importing data](/guides/index-data/import-data). It is not required for [storing audit logs](/guides/production/configure-audit-logs).
</Note>

In the [Pinecone console](https://app.pinecone.io/organizations/-/projects), add an integration with Amazon S3..

1. Select your project.
2. Go to [**Manage > Storage integrations**](https://app.pinecone.io/organizations/-/projects/-/storage).
3. Click **Add integration**.
4. Enter a unique integration name.
5. Select **Amazon S3**.
6. Enter the **ARN** of the [IAM role you created](/guides/operations/integrations/integrate-with-amazon-s3#2-set-up-access-using-an-iam-role).
7. Click **Add integration**.

## Next steps

* [Import data](/guides/index-data/import-data) from your Amazon S3 bucket into a Pinecone index.
* [Configure audit logs](/guides/production/configure-audit-logs) to export logs to your Amazon S3 bucket.

---

# Source: https://docs.pinecone.io/guides/operations/integrations/integrate-with-azure-blob-storage.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Integrate with Azure Blob Storage

> Set up Azure Blob Storage integration for data import.

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Standard and Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

This page describes how to integrate Pinecone with Azure Blob Storage. After setting up an integration, you can [import data](/guides/index-data/import-data) from an Azure Blob Storage container into a Pinecone index hosted on AWS, GCP, or Azure.

## Before you begin

Ensure you have the following:

* A [Pinecone account](https://app.pinecone.io/)
* An [Azure Blob Storage container](https://learn.microsoft.com/azure/storage/blobs/storage-blobs-introduction)

## 1. Create an app registration and service principal

Pinecone uses a service principal to access your Azure Blob Storage container.

1. [Create an app registration](https://learn.microsoft.com/entra/identity-platform/quickstart-register-app) for your Pinecone integration. This automatically creates a service principal.

   When creating your app registration:

   * Do not specify a **Redirect URI**.
   * Copy the **Application (client) ID** and the **Directory (tenant) ID**. You'll use these values when adding a storage integration in Pinecone.

2. [Create a client secret](https://learn.microsoft.com/entra/identity-platform/how-to-add-credentials?tabs=client-secret) for the service principal.

   Copy the secret's **Value** (not its **ID**). You'll use this when creating a storage integration in Pinecone.

## 2. Grant access to the storage account

[Assign the service principal to your storage account](https://learn.microsoft.com/azure/storage/common/storage-auth-aad-rbac-portal#assign-azure-rbac-roles-using-the-azure-portal):

1. In the Azure portal, navigate to the subscription associated with your storage account.
2. Select **Access control (IAM)**.
3. Click **Add** > **Add role assignment**.
4. Select **Storage Blob Data Reader** or another role that has permission to list and read blobs in a container.
5. Click **Next**.
6. Select **User, group, or service principal** and click **Select members**.
7. Select the app you created in the previous step.
8. Click **Review + assign** (you may need to click this twice).

## 3. In Pinecone, add a storage integration

In the [Pinecone console](https://app.pinecone.io/organizations/-/projects), add an integration with Azure Blob Storage:

1. Select your project.
2. Go to [**Manage > Storage integrations**](https://app.pinecone.io/organizations/-/projects/-/storage).
3. Click **Add integration**.
4. Enter a unique integration name.
5. Select **Azure Blob Storage**.
6. For **Tenant ID**, **Client ID**, and **Client secret**, enter the values you copied from Azure.
7. Click **Add integration**.

## Next steps

[Import data](/guides/index-data/import-data) from your Azure Blob Storage container into your Pinecone index.

---

# Source: https://docs.pinecone.io/guides/operations/integrations/integrate-with-google-cloud-storage.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Integrate with Google Cloud Storage

> Integrate Google Cloud Storage for bulk data import

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Standard and Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

This page shows you how to integrate Pinecone with a Google Cloud Storage (GCS) bucket. Once your integration is set up, you can use it to [import data](/guides/index-data/import-data) from your bucket into a Pinecone index hosted on AWS, GCP, or Azure.

## Before you begin

Ensure you have the following:

* A [Pinecone account](https://app.pinecone.io/)
* A [Google Cloud Storage bucket](https://cloud.google.com/storage/docs/creating-buckets)

## 1. Create a service account and key

Pinecone will use a service account to access your GCS bucket.

1. [Create a service account](https://cloud.google.com/iam/docs/service-accounts-create) for your Pinecone integration.
2. [Create a service account key](https://cloud.google.com/iam/docs/keys-create-delete#creating). Select **JSON** as the key type.

   The key will be downloaded to your computer. You'll use this key when adding a storage integration in Pinecone.

## 2. Grant access to the bucket

[Add your service account as a principal to the bucket](https://cloud.google.com/storage/docs/access-control/using-iam-permissions#bucket-add).

* For the principal, use your service account email address.
* For the role, select **Storage Object Viewer** or another role that has permission to list and read objects in a bucket.

## 3. Add a storage integration

In the [Pinecone console](https://app.pinecone.io/organizations/-/projects), add an integration with Google Cloud Storage:

1. Select your project.
2. Go to [**Manage > Storage integrations**](https://app.pinecone.io/organizations/-/projects/-/storage).
3. Click **Add integration**.
4. Enter a unique integration name.
5. Select **Google Cloud Storage**.
6. Open the JSON key file for your service account.
7. Copy the contents of the key file and paste them into the **Index account key JSON** field.
8. Click **Add integration**.

## Next steps

[Import data](/guides/index-data/import-data) from your GCS bucket into your Pinecone index.

---

# Source: https://docs.pinecone.io/integrations/build-integration/integration-ecosystem.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Integration ecosystem

Anyone can use the [Pinecone SDKs](/reference/pinecone-sdks) or the [Pinecone API](/reference/api/introduction) to build a native Pinecone integration. We encourage you to build and launch freely to support your users.

Listing on our official [Integrations](/integrations/overview) page is a curated process. We prioritize applications from integrations that are actively used by mutual customers. If you believe your integration qualifies for official partner support, please fill out our [application form](https://dash.partnerstack.com/application?company=pinecone\&group=technologypartners).

## Additional information

* [Attribute usage to your integration](/integrations/build-integration/attribute-usage-to-your-integration)
* [Connect your users to Pinecone](/integrations/build-integration/connect-your-users-to-pinecone)

---

# Source: https://docs.pinecone.io/reference/api/introduction.md

# Source: https://docs.pinecone.io/reference/api/assistant/introduction.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Assistant API reference

Use the [Assistant API](/guides/assistant/quickstart/sdk-quickstart) to upload documents, ask questions, and receive responses that reference your documents. This is known as [retrieval-augmented generation (RAG)](https://www.pinecone.io/learn/retrieval-augmented-generation/).

## SDK support

The following Pinecone SDKs support the Assistant API:

<CardGroup cols={3}>
  <Card title="Python SDK" icon="python" href="/reference/sdks/python/overview" />

  <Card title="Node.js SDK" icon="node-js" href="/reference/sdks/node/overview" />
</CardGroup>

## Versioning

The Assistant API is versioned to ensure that your applications continue to work as expected as the platform evolves. For more details, see [API versioning](/reference/api/versioning) in the Pinecone Database documentation.

---

# Source: https://docs.pinecone.io/reference/java-sdk.md

# Java SDK

<Tip>
  See the [Pinecone Java SDK
  documentation](https://github.com/pinecone-io/pinecone-java-client/blob/main/README.md) for full installation
  instructions and usage examples.

  To make a feature request or report an issue, please [file an issue](https://github.com/pinecone-io/pinecone-java-client/issues).
</Tip>

## Requirements

The Pinecone Java SDK Java 1.8 or later.

## SDK versions

SDK versions are pinned to specific [API versions](/reference/api/versioning). When a new API version is released, a new version of the SDK is also released.

The mappings between API versions and Java SDK versions are as follows:

| API version        | SDK version |
| :----------------- | :---------- |
| `2025-04` (latest) | v5.x        |
| `2025-01`          | v4.x        |
| `2024-10`          | v3.x        |
| `2024-07`          | v2.x        |
| `2024-04`          | v1.x        |

When a new stable API version is released, you should upgrade your SDK to the latest version to ensure compatibility with the latest API changes.

## Install

To install the latest version of the [Java SDK](https://github.com/pinecone-io/pinecone-java-client), add a dependency to the current module:

```shell Java theme={null}
# Maven
<dependency>
  <groupId>io.pinecone</groupId>
  <artifactId>pinecone-client</artifactId>
  <version>5.0.0</version>
</dependency>

# Gradle
implementation "io.pinecone:pinecone-client:5.0.0"
```

Alternatively, you can download the standalone uberjar [pinecone-client-4.0.0-all.jar](https://repo1.maven.org/maven2/io/pinecone/pinecone-client/4.0.0/pinecone-client-4.0.0-all.jar), which bundles the Pinecone SDK and all dependencies together. You can include this in your classpath like you do with any third-party JAR without having to obtain the `pinecone-client` dependencies separately.

## Upgrade

<Warning>
  Before upgrading to `v4.0.0`, update all relevant code to account for the breaking changes explained [here](/release-notes/2025#2025-02-07-3).
</Warning>

If you are already using the Java SDK, upgrade the dependency in the current module to the latest version:

```shell Java theme={null}
# Maven
<dependency>
  <groupId>io.pinecone</groupId>
  <artifactId>pinecone-client</artifactId>
  <version>5.0.0</version>
</dependency>

# Gradle
implementation "io.pinecone:pinecone-client:5.0.0"
```

## Initialize

Once installed, you can import the SDK and then use an [API key](/guides/production/security-overview#api-keys) to initialize a client instance:

```Java  theme={null}
import io.pinecone.clients.Pinecone;
import org.openapitools.db_control.client.model.*;

public class InitializeClientExample {
    public static void main(String[] args) {
        Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
    }
}
```

---

# Source: https://docs.pinecone.io/integrations/jina.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Jina AI

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Jina Embeddings leverage powerful models to generate high-quality text embeddings that can process inputs up to 8,000 tokens. Jina Embeddings are designed to be highly versatile, catering to both domain-specific use cases, such as e-commerce, and language-specific needs, including Chinese and German. By providing robust models and the expertise to fine-tune them for specific requirements, Jina AI empowers developers to enhance their search functionalities, improve natural language understanding, and drive more insightful data analysis.

By integrating Pinecone with Jina, you can add knowledge to LLMs via retrieval augmented generation (RAG), greatly enhancing LLM ability for autonomous agents, chatbots, question-answering, and multi-agent systems.

<PrimarySecondaryCTA primaryHref={"https://jina.ai/embeddings/#apiform"} primaryLabel={"Get started"} primaryTarget={"_blank"} secondaryHref={"https://www.pinecone.io/models/jina-embeddings-v2-base-en/"} secondaryLabel={"Try the model"} />

---

# Source: https://docs.pinecone.io/troubleshooting/keep-customer-data-separate.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Keep customer data separate in Pinecone

Some use cases require vectors to be segmented by their customers, either physically or logically. The table below describes three techniques to accomplish this and the pros and cons of considering each:

| **Techniques**                                                                                                    | **Pros**                                                                                                                                                                         | **Cons**                                                                                       |
| ----------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------- |
| **Separate Indexes**<br />Each customer would have a separate index                                               | • Customer data is truly separated physically by indexes                                                                                                                         | • You cannot query across customers if you wish<br />• Cost and maintenance of several indexes |
| **Namespaces**<br />You can isolate data within a single index using namespaces                                   | • You can only query one namespace at a time, which would keep customer data separate logically<br />• Cheaper than #1 potentially by making more efficient use of index storage | • You cannot query across namespaces<br />• Customer data is only separated logically          |
| **Metadata Filtering**<br />All data is kept in a single index and logically separated by filtering at query time | • Most versatile if you wish to query across customers<br />• As with namespaces cheaper than #1 potentially                                                                     | • Customer data is separated only by filtering at query time                                   |

---

# Source: https://docs.pinecone.io/reference/api/known-limitations.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Known limitations

This page describes known limitations and feature restrictions in Pinecone.

## General

* [Upserts](/guides/index-data/upsert-data)
  * Pinecone is eventually consistent, so there can be a slight delay before upserted records are available to query.

    After upserting records, use the [`describe_index_stats`](/reference/api/2024-10/data-plane/describeindexstats) operation to check if the current vector count matches the number of records you expect, although this method may not work for pod-based indexes with multiple replicas.
  * Only indexes using the [dotproduct distance metric](/guides/index-data/indexing-overview#dotproduct) support querying sparse-dense vectors.

    Upserting, updating, and fetching sparse-dense vectors in indexes with a different distance metric will succeed, but querying will return an error.
  * Indexes created before February 22, 2023 do not support sparse vectors.
* [Metadata](/guides/index-data/upsert-data#upsert-with-metadata-filters)
  * Null metadata values aren't supported. Instead of setting a key to `null`, remove the key from the metadata payload.
  * Nested JSON objects are not supported.

## Serverless indexes

Serverless indexes do not support the following features:

* [Filtering index statistics by metadata](/reference/api/2024-10/data-plane/describeindexstats)
* [Private endpoints](/guides/production/connect-to-aws-privatelink)

  * This feature is available on AWS only.

---

# Source: https://docs.pinecone.io/integrations/langchain.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# LangChain

> Using LangChain and Pinecone to add knowledge to LLMs

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

LangChain provides modules for managing and optimizing the use of large language models (LLMs) in applications. Its core philosophy is to facilitate data-aware applications where the language model interacts with other data sources and its environment. This framework consists of several parts that simplify the entire application lifecycle:

* Write your applications in LangChain/LangChain.js. Get started quickly by using Templates for reference.
* Use LangSmith to inspect, test, and monitor your chains to constantly improve and deploy with confidence.
* Turn any chain into an API with LangServe.

By integrating Pinecone with LangChain, you can add knowledge to LLMs via retrieval augmented generation (RAG), greatly enhancing LLM ability for autonomous agents, chatbots, question-answering, and multi-agent systems.

<PrimarySecondaryCTA primaryHref={"https://github.com/langchain-ai/pinecone-serverless"} primaryLabel={"Get started"} primaryTarget={"_blank"} secondaryHref={"#setup-guide"} secondaryLabel={"View setup guide"} />

## Setup guide

This guide shows you how to integrate Pinecone, a high-performance vector database, with [LangChain](https://www.langchain.com/), a framework for building applications powered by large language models (LLMs).

Pinecone enables developers to build scalable, real-time recommendation and search systems based on vector similarity search. LangChain, on the other hand, provides modules for managing and optimizing the use of language models in applications. Its core philosophy is to facilitate data-aware applications where the language model interacts with other data sources and its environment.

By integrating Pinecone with LangChain, you can add knowledge to LLMs via [Retrieval Augmented Generation (RAG)](https://www.pinecone.io/learn/series/rag/), greatly enhancing LLM ability for autonomous agents, chatbots, question-answering, and multi-agent systems.

<Note>
  This guide demonstrates only one way out of many that you can use LangChain and Pinecone together. For additional examples, see:

  * [LangChain AI Handbook](https://www.pinecone.io/learn/series/langchain/)
  * [Retrieval Augmentation for LLMs](https://github.com/pinecone-io/examples/blob/master/learn/generation/langchain/handbook/05-langchain-retrieval-augmentation.ipynb)
  * [Retrieval Augmented Conversational Agent](https://github.com/pinecone-io/examples/blob/master/learn/generation/langchain/handbook/08-langchain-retrieval-agent.ipynb)
</Note>

## Key concepts

The `PineconeVectorStore` class provided by LangChain can be used to interact with Pinecone indexes. It's important to remember that you must have an existing Pinecone index before you can create a `PineconeVectorStore` object.

### Initializing a vector store

To initialize a `PineconeVectorStore` object, you must provide the name of the Pinecone index and an `Embeddings` object initialized through LangChain. There are two general approaches to initializing a `PineconeVectorStore` object:

1. Initialize without adding records:

```Python Python theme={null}
    import os
    from langchain_pinecone import PineconeVectorStore
    from langchain_openai import OpenAIEmbeddings

    os.environ['OPENAI_API_KEY'] = '<YOUR_OPENAI_API_KEY>'
    os.environ['PINECONE_API_KEY'] = '<YOUR_PINECONE_API_KEY>'

    index_name = "<YOUR_PINECONE_INDEX_NAME>"
    embeddings = OpenAIEmbeddings()

    vectorstore = PineconeVectorStore(index_name=index_name, embedding=embeddings)
```

You can also use the `from_existing_index` method of LangChain's `PineconeVectorStore` class to initialize a vector store.

2. Initialize while adding records:

The `from_documents` and `from_texts` methods of LangChain's `PineconeVectorStore` class add records to a Pinecone index and return a `PineconeVectorStore` object.

The `from_documents` method accepts a list of LangChain's `Document` class objects, which can be created using LangChain's `CharacterTextSplitter` class. The `from_texts` method accepts a list of strings. Similarly to above, you must provide the name of an existing Pinecone index and an `Embeddings` object.

Both of these methods handle the embedding of the provided text data and the creation of records in your Pinecone index.

```Python Python theme={null}
    import os
    from langchain_pinecone import PineconeVectorStore
    from langchain_openai import OpenAIEmbeddings
    from langchain_community.document_loaders import TextLoader
    from langchain_text_splitters import CharacterTextSplitter

    os.environ['OPENAI_API_KEY'] = '<YOUR_OPENAI_API_KEY>'
    os.environ['PINECONE_API_KEY'] = '<YOUR_PINECONE_API_KEY>'

    index_name = "<YOUR_PINECONE_INDEX_NAME>"
    embeddings = OpenAIEmbeddings()

    # path to an example text file
    loader = TextLoader("../../modules/state_of_the_union.txt")
    documents = loader.load()
    text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
    docs = text_splitter.split_documents(documents)

    vectorstore_from_docs = PineconeVectorStore.from_documents(
        docs,
        index_name=index_name,
        embedding=embeddings
    )

    texts = ["Tonight, I call on the Senate to: Pass the Freedom to Vote Act.", "ne of the most serious constitutional responsibilities a President has is nominating someone to serve on the United States Supreme Court.", "One of our nation’s top legal minds, who will continue Justice Breyer’s legacy of excellence."]

    vectorstore_from_texts = PineconeVectorStore.from_texts(
        texts,
        index_name=index_name,
        embedding=embeddings
    )
```

### Add more records

Once you have initialized a `PineconeVectorStore` object, you can add more records to the underlying Pinecone index (and thus also the linked LangChain object) using either the `add_documents` or `add_texts` methods.

Like their counterparts that also initialize a `PineconeVectorStore` object, both of these methods also handle the embedding of the provided text data and the creation of records in your Pinecone index.

```Python Python theme={null}
    # path to an example text file
    loader = TextLoader("../../modules/inaugural_address.txt")
    documents = loader.load()
    text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
    docs = text_splitter.split_documents(documents)

    vectorstore = PineconeVectorStore(index_name=index_name, embedding=embeddings)

    vectorstore.add_documents(docs)
```

```Python Python theme={null}
    vectorstore = PineconeVectorStore(index_name=index_name, embedding=embeddings)

    vectorstore.add_texts(["More text to embed and add to the index!"])
```

### Perform a similarity search

A `similarity_search` on a `PineconeVectorStore` object returns a list of LangChain `Document` objects most similar to the query provided. While the `similarity_search` uses a Pinecone query to find the most similar results, this method includes additional steps and returns results of a different type.

The `similarity_search` method accepts raw text and automatically embeds it using the `Embedding` object provided when you initialized the `PineconeVectorStore`. You can also provide a `k` value to determine the number of LangChain `Document` objects to return. The default value is `k=4`.

```Python Python theme={null}
    query = "Who is Ketanji Brown Jackson?"
    vectorstore.similarity_search(query)
    
    # Response:
    # [
    #    Document(page_content='Ketanji Onyika Brown Jackson is an American lawyer and jurist who is an associate justice of the Supreme Court of the United...', metadata={'chunk': 0.0, 'source': 'https://en.wikipedia.org/wiki/Ketanji_Brown_Jackson', 'title': 'Ketanji Brown Jackson', 'wiki-id': '6573'}),  
    #    Document(page_content='Jackson was nominated to the Supreme Court by President Joe Biden on February 25, 2022, and confirmed by the U.S. Senate...', metadata={'chunk': 1.0, 'source': 'https://en.wikipedia.org/wiki/Ketanji_Brown_Jackson', 'title': 'Ketanji Brown Jackson', 'wiki-id': '6573'}),  
    #    Document(page_content='Jackson grew up in Miami and attended Miami Palmetto Senior High School. She distinguished herself as a champion debater...', metadata={'chunk': 3.0, 'source': 'https://en.wikipedia.org/wiki/Ketanji_Brown_Jackson', 'title': 'Ketanji Brown Jackson', 'wiki-id': '6573'}),
    #    Document(page_content='After high school, Jackson matriculated at Harvard University to study government, having applied despite her guidance...', metadata={'chunk': 5.0, 'source': 'https://en.wikipedia.org/wiki/Ketanji_Brown_Jackson', 'title': 'Ketanji Brown Jackson', 'wiki-id': '6573'})
    # ]   
```

You can also optionally apply a metadata filter to your similarity search. The filtering query language is the same as for Pinecone queries, as detailed in [Filtering with metadata](https://docs.pinecone.io/guides/index-data/indexing-overview#metadata).

```Python Python theme={null}
    query = "Tell me more about Ketanji Brown Jackson."
    vectorstore.similarity_search(query, filter={'source': 'https://en.wikipedia.org/wiki/Ketanji_Brown_Jackson'})
```

### Namespaces

Several methods of the `PineconeVectorStore` class support using [namespaces](/guides/index-data/indexing-overview#namespaces). You can also initialize your `PineconeVectorStore` object with a namespace to restrict all further operations to that space.

```Python Python theme={null}
    index_name = "<YOUR_PINECONE_INDEX_NAME>"
    embeddings = OpenAIEmbeddings()

    vectorstore = PineconeVectorStore(index_name=index_name, embedding=embeddings, namespace="example-namespace")
```

If you initialize your `PineconeVectorStore` object without a namespace, you can specify the target namespace within the operation.

```Python Python theme={null}
    # path to an example text file
    loader = TextLoader("../../modules/congressional_address.txt")
    documents = loader.load()
    text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
    docs = text_splitter.split_documents(documents)

    vectorstore_from_docs = PineconeVectorStore.from_documents(
        docs,
        index_name=index_name,
        embedding=embeddings,
        namespace="example-namespace"
    )

    vectorstore_from_texts = PineconeVectorStore.from_texts(
        texts,
        index_name=index_name,
        embedding=embeddings,
        namespace="example-namespace"
    )

    vectorstore_from_docs.add_documents(docs, namespace="example-namespace")

    vectorstore_from_texts.add_texts(["More text!"], namespace="example-namespace")
```

```Python Python theme={null}
    query = "Who is Ketanji Brown Jackson?"
    vectorstore.similarity_search(query, namespace="example-namespace")
```

## Tutorial

### 1. Set up your environment

Before you begin, install some necessary libraries and set environment variables for your Pinecone and OpenAI API keys:

```Shell  theme={null}
pip install -qU \
  "pinecone[grpc]"==5.1.0 \
  pinecone-datasets==0.7.0 \
  langchain-pinecone==0.1.2 \
  langchain-openai==0.1.23 \
  langchain==0.2.15
```

```Shell  theme={null}
# Set environment variables for API keys
export PINECONE_API_KEY=<your Pinecone API key available at app.pinecone.io>
export OPENAI_API_KEY=<your OpenAI API key, available at platform.openai.com/api-keys>
```

```Python Python theme={null}
import os
pinecone_api_key = os.environ.get('PINECONE_API_KEY')
openai_api_key = os.environ.get('OPENAI_API_KEY')
```

### 2. Build the knowledge base

1. Load a [sample Pinecone dataset](/guides/data/use-public-pinecone-datasets) into memory:

   ```Python Python theme={null}
   import pinecone_datasets  
   dataset = pinecone_datasets.load_dataset('wikipedia-simple-text-embedding-ada-002-100K')  
   len(dataset)  

   # Response:
   # 100000
   ```

2. Reduce the dataset and format it for upserting into Pinecone:

   ```Python Python theme={null}
   # we will use rows of the dataset up to index 30_000
   dataset.documents.drop(dataset.documents.index[30_000:], inplace=True)
   # we drop sparse_values as they are not needed for this example  
   dataset.documents.drop(['metadata'], axis=1, inplace=True)  
   dataset.documents.rename(columns={'blob': 'metadata'}, inplace=True)  
   ```

### 3. Index the data in Pinecone

1. Initialize your client connection to Pinecone and create an index. This step uses the Pinecone API key you set as an environment variable [earlier](#1-set-up-your-environment).

   ```Python Python theme={null}
   from pinecone.grpc import PineconeGRPC as Pinecone
   from pinecone import ServerlessSpec, PodSpec  
   import time  
   # configure client  
   pc = Pinecone(api_key=pinecone_api_key)  
   spec = ServerlessSpec(cloud='aws', region='us-east-1')  
   # check for and delete index if already exists  
   index_name = 'langchain-retrieval-augmentation-fast'  
   if pc.has_index(index_name):  
       pc.delete_index(name=index_name)  
   # create a new index  
   pc.create_index(  
       index_name,  
       dimension=1536,  # dimensionality of text-embedding-ada-002  
       metric='dotproduct',  
       spec=spec  
   )  
   ```

2. Target the index and check its current stats:

   ```Python Python theme={null}
   index = pc.Index(index_name)  
   index.describe_index_stats()  

   # Response:
   # {'dimension': 1536,  
   # 'index_fullness': 0.0,  
   # 'namespaces': {},  
   # 'total_vector_count': 0}  
   ```

   You'll see that the index has a `total_vector_count` of `0`, as you haven't added any vectors yet.

3. Now upsert the data to Pinecone:

   ```Python Python theme={null}
   for batch in dataset.iter_documents(batch_size=100):  
       index.upsert(batch)  
   ```

4. Once the data is indexed, check the index stats once again:

   ```Python Python theme={null}
   index.describe_index_stats()  

   # Response:
   # {'dimension': 1536,  
   # 'index_fullness': 0.0,  
   # 'namespaces': {},  
   # 'total_vector_count': 70000} 
   ```

### 4. Initialize a LangChain vector store

Now that you've built your Pinecone index, you need to initialize a LangChain vector store using the index. This step uses the OpenAI API key you set as an environment variable [earlier](#1-set-up-your-environment). Note that OpenAI is a paid service and so running the remainder of this tutorial may incur some small cost.

1. Initialize a LangChain embedding object:

   ```Python Python theme={null}
   from langchain_openai import OpenAIEmbeddings  
   # get openai api key from platform.openai.com  
   model_name = 'text-embedding-ada-002'  
   embeddings = OpenAIEmbeddings(  
       model=model_name,  
       openai_api_key=openai_api_key  
   )  
   ```

2. Initialize the LangChain vector store:

   The `text_field` parameter sets the name of the metadata field that stores the raw text when you upsert records using a LangChain operation such as `vectorstore.from_documents` or `vectorstore.add_texts`.
   This metadata field is used as the `page_content` in the `Document` objects retrieved from query-like LangChain operations such as `vectorstore.similarity_search`.
   If you do not specify a value for `text_field`, it will default to `"text"`.

   ```Python Python theme={null}
   from langchain_pinecone import PineconeVectorStore  
   text_field = "text"  
   vectorstore = PineconeVectorStore(  
       index, embeddings, text_field  
   )  
   ```

3. Now you can query the vector store directly using `vectorstore.similarity_search`:

   ```Python Python theme={null}
   query = "who was Benito Mussolini?"  
   vectorstore.similarity_search(  
       query,  # our search query  
       k=3  # return 3 most relevant docs  
   )  

   # Response:
   # [Document(page_content='Benito Amilcare Andrea Mussolini KSMOM GCTE (29 July 1883 – 28 April 1945) was an Italian politician and journalist...', metadata={'chunk': 0.0, 'source': 'https://simple.wikipedia.org/wiki/Benito%20Mussolini', 'title': 'Benito Mussolini', 'wiki-id': '6754'}),  
   # Document(page_content='Fascism as practiced by Mussolini\nMussolini\'s form of Fascism, "Italian Fascism"- unlike Nazism, the racist ideology...', metadata={'chunk': 1.0, 'source': 'https://simple.wikipedia.org/wiki/Benito%20Mussolini', 'title': 'Benito Mussolini', 'wiki-id': '6754'}),  
   # Document(page_content='Veneto was made part of Italy in 1866 after a war with Austria. Italian soldiers won Latium in 1870. That was when...', metadata={'chunk': 5.0, 'source': 'https://simple.wikipedia.org/wiki/Italy', 'title': 'Italy', 'wiki-id': '363'})]
   ```

All of these sample results are good and relevant. But what else can you do with this? There are many tasks, one of the most interesting (and well supported by LangChain) is called "Generative Question-Answering" or GQA.

### 5. Use Pinecone and LangChain for RAG

In RAG, you take the query as a question that is to be answered by a LLM, but the LLM must answer the question based on the information it is seeing from the vectorstore.

1. To do this, initialize a `RetrievalQA` object like so:

   ```Python Python theme={null}
   from langchain_openai import ChatOpenAI  
   from langchain.chains import RetrievalQA  
   # completion llm  
   llm = ChatOpenAI(  
       openai_api_key=OPENAI_API_KEY,  
       model_name='gpt-3.5-turbo',  
       temperature=0.0  
   )  
   qa = RetrievalQA.from_chain_type(  
       llm=llm,  
       chain_type="stuff",  
       retriever=vectorstore.as_retriever()  
   )  
   qa.invoke(query)  

   # Response:
   # Benito Mussolini was an Italian politician and journalist who served as the Prime Minister of Italy from 1922 until 1943. He was the leader of the National Fascist Party and played a significant role in the rise of fascism in Italy...
   ```

2. You can also include the sources of information that the LLM is using to answer your question using a slightly different version of `RetrievalQA` called `RetrievalQAWithSourcesChain`:

   ```Python Python theme={null}
   from langchain.chains import RetrievalQAWithSourcesChain  
   qa_with_sources = RetrievalQAWithSourcesChain.from_chain_type(  
       llm=llm,  
       chain_type="stuff",  
       retriever=vectorstore.as_retriever()  
   )  
   qa_with_sources.invoke(query)

   # Response:
   # {'question': 'who was Benito Mussolini?',  
   # 'answer': "Benito Mussolini was an Italian politician and journalist who served as the Prime Minister of Italy from 1922 until 1943. He was the leader of the National Fascist Party and played a significant role in the rise of fascism in Italy...",  
   # 'sources': 'https://simple.wikipedia.org/wiki/Benito%20Mussolini'}  
   ```

### 6. Clean up

When you no longer need the index, use the `delete_index` operation to delete it:

```Python Python theme={null}
pc.delete_index(name=index_name)
```

## Related articles

* [LangChain AI Handbook](https://www.pinecone.io/learn/series/langchain/)

---

# Source: https://docs.pinecone.io/integrations/langtrace.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Langtrace

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Scale3 Labs recently launched Langtrace AI, an open-source monitoring and evaluation platform for LLM-powered applications. Langtrace is built based on Open Telemetry(OTEL) standards and supports native tracing for the most popular LLM vendors, VectorDBs, and frameworks(like Langchain and LlamaIndex).

Langtrace AI supports tracing Pinecone natively, which means the Langtrace SDK can generate OTEL standard traces with automatic instrumentation in just 2 lines of code. These traces can be ingested by an observability tool that supports OTEL, such as Datadog, Grafana/Prometheus, SigNoz, Sentry, etc. Langtrace also has a visualization client that is optimized for visualizing the traces generated in an LLM stack.

By having a Pinecone integration, Pinecone users can get access to rich and high cardinal tracing for the Pinecone API calls using Langtrace, which they can ingest into their observability tool of choice. This helps customers gain insights into the DB calls and help with debugging and troubleshooting applications in case of incidents.

<PrimarySecondaryCTA primaryHref={"https://langtrace.ai/blog/tracing-pinecone-using-langtrace"} primaryLabel={"Get started"} secondaryHref={"https://www.youtube.com/watch?v=9BM0M8lwTBg"} secondaryLabel={"View video tutorial"} />

---

# Source: https://docs.pinecone.io/guides/search/lexical-search.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Lexical search

> Perform keyword-based search on sparse indexes

This page shows you how to search a [sparse index](/guides/index-data/indexing-overview#sparse-indexes) for records that most exactly match the words or phrases in a query. This is often called lexical search or keyword search.

Lexical search uses [sparse vectors](https://www.pinecone.io/learn/sparse-retrieval/), which have a very large number of dimensions, where only a small proportion of values are non-zero. The dimensions represent words from a dictionary, and the values represent the importance of these words in the document. Words are scored independently and then summed, with the most similar records scored highest.

## Search with text

<Note>
  Searching with text is supported only for [indexes with integrated embedding](/guides/index-data/indexing-overview#integrated-embedding).
</Note>

To search a sparse index with a query text, use the [`search_records`](/reference/api/latest/data-plane/search_records) operation with the following parameters:

* `namespace`: The [namespace](/guides/index-data/indexing-overview#namespaces) to query. To use the default namespace, set to `"__default__"`.
* `query.inputs.text`:  The query text. Pinecone uses the [embedding model](/guides/index-data/create-an-index#embedding-models) integrated with the index to convert the text to a sparse vector automatically.
* `query.top_k`:  The number of records to return.
* `query.match_terms`: (Optional) A list of terms that must be present in each search result. For more details, see [Filter by required terms](#filter-by-required-terms).
* `fields`: (Optional) The fields to return in the response. If not specified, the response includes all fields.

For example, the following code converts the query “What is AAPL's outlook, considering both product launches and market conditions?” to a sparse vector and then searches for the 3 most similar vectors in the `example-namespace` namespace:

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  results = index.search(
      namespace="example-namespace", 
      query={
          "inputs": {"text": "What is AAPL's outlook, considering both product launches and market conditions?"}, 
          "top_k": 3
      },
      fields=["chunk_text", "quarter"]
  )

  print(results)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const namespace = pc.index("INDEX_NAME", "INDEX_HOST").namespace("example-namespace");

  const response = await namespace.searchRecords({
    query: {
      topK: 3,
      inputs: { text: "What is AAPL's outlook, considering both product launches and market conditions?" },
    },
    fields: ['chunk_text', 'quarter']
  });

  console.log(response);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import org.openapitools.db_data.client.ApiException;
  import org.openapitools.db_data.client.model.SearchRecordsResponse;

  import java.util.*;

  public class SearchText {
      public static void main(String[] args) throws ApiException {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);

          Index index = new Index(config, connection, "integrated-sparse-java");

          String query = "What is AAPL's outlook, considering both product launches and market conditions?";
          List<String> fields = new ArrayList<>();
          fields.add("category");
          fields.add("chunk_text");

          // Search the sparse index
          SearchRecordsResponse recordsResponse = index.searchRecordsByText(query,  "example-namespace", fields, 3, null, null);

          // Print the results
          System.out.println(recordsResponse);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
    	bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
      } 

      res, err := idxConnection.SearchRecords(ctx, &pinecone.SearchRecordsRequest{
          Query: pinecone.SearchRecordsQuery{
              TopK: 3,
              Inputs: &map[string]interface{}{
                  "text": "What is AAPL's outlook, considering both product launches and market conditions?",
              },
          },
          Fields: &[]string{"chunk_text", "category"},
      })
      if err != nil {
          log.Fatalf("Failed to search records: %v", err)
      }
      fmt.Printf(prettifyStruct(res))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var response = await index.SearchRecordsAsync(
      "example-namespace",
      new SearchRecordsRequest
      {
          Query = new SearchRecordsRequestQuery
          {
              TopK = 3,
              Inputs = new Dictionary<string, object?> { { "text", "What is AAPL's outlook, considering both product launches and market conditions?" } },
          },
          Fields = ["category", "chunk_text"],
      }
  );

  Console.WriteLine(response);
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/records/namespaces/example-namespace/search" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "query": {
            "inputs": { "text": "What is AAPL'\''s outlook, considering both product launches and market conditions?" },
            "top_k": 3
          },
          "fields": ["chunk_text", "quarter"]
      }'
  ```
</CodeGroup>

The results will look as follows. The most similar records are scored highest.

<CodeGroup>
  ```python Python theme={null}
  {'result': {'hits': [{'_id': 'vec2',
                        '_score': 10.77734375,
                        'fields': {'chunk_text': "Analysts suggest that AAPL'''s "
                                                 'upcoming Q4 product launch '
                                                 'event might solidify its '
                                                 'position in the premium '
                                                 'smartphone market.',
                                   'quarter': 'Q4'}},
                       {'_id': 'vec3',
                        '_score': 6.49066162109375,
                        'fields': {'chunk_text': "AAPL'''s strategic Q3 "
                                                 'partnerships with '
                                                 'semiconductor suppliers could '
                                                 'mitigate component risks and '
                                                 'stabilize iPhone production.',
                                   'quarter': 'Q3'}},
                       {'_id': 'vec1',
                        '_score': 5.3671875,
                        'fields': {'chunk_text': 'AAPL reported a year-over-year '
                                                 'revenue increase, expecting '
                                                 'stronger Q3 demand for its '
                                                 'flagship phones.',
                                   'quarter': 'Q3'}}]},
   'usage': {'embed_total_tokens': 18, 'read_units': 1}}
  ```

  ```javascript JavaScript theme={null}
  {
    result: { 
      hits: [ 
        {
          _id: "vec2",
          _score: 10.82421875,
          fields: {
            chunk_text: "Analysts suggest that AAPL'''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
            quarter: "Q4"
          }
        },
        {
          _id: "vec3",
          _score: 6.49066162109375,
          fields: {
            chunk_text: "AAPL'''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production.",
            quarter: "Q3"
          }
        },
        {
          _id: "vec1",
          _score: 5.3671875,
          fields: {
            chunk_text: "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
            quarter: "Q3"
          }
        }
      ]
    },
    usage: { 
      readUnits: 1, 
      embedTotalTokens: 18 
    }
  }
  ```

  ```java Java theme={null}
  class SearchRecordsResponse {
      result: class SearchRecordsResponseResult {
          hits: [class Hit {
              id: vec2
              score: 10.82421875
              fields: {chunk_text=Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market., quarter=Q4}
              additionalProperties: null
          }, class Hit {
              id: vec3
              score: 6.49066162109375
              fields: {chunk_text=AAAPL'''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production., quarter=Q3}
              additionalProperties: null
          }, class Hit {
              id: vec1
              score: 5.3671875
              fields: {chunk_text=AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones., quarter=Q3}
              additionalProperties: null
          }]
          additionalProperties: null
      }
      usage: class SearchUsage {
          readUnits: 1
          embedTotalTokens: 18
      }
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  {
    "result": {
      "hits": [
        {
          "_id": "vec2",
          "_score": 10.833984,
          "fields": {
            "chunk_text": "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
            "quarter": "Q4"
          }
        },
        {
          "_id": "vec3",
          "_score": 6.473572,
          "fields": {
            "chunk_text": "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production.",
            "quarter": "Q3"
          }
        },
        {
          "_id": "vec1",
          "_score": 5.3710938,
          "fields": {
            "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
            "quarter": "Q3"
          }
        }
      ]
    },
    "usage": {
      "read_units": 6,
      "embed_total_tokens": 18
    }
  }
  ```

  ```csharp C# theme={null}
  {
      "result": {
          "hits": [
              {
                  "_id": "vec2",
                  "_score": 10.833984,
                  "fields": {
                      "chunk_text": "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
                      "quarter": "Q4"
                  }
              },
              {
                  "_id": "vec3",
                  "_score": 6.473572,
                  "fields": {
                      "chunk_text": "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production.",
                      "quarter": "Q3"
                  }
              },
              {
                  "_id": "vec1",
                  "_score": 5.3710938,
                  "fields": {
                      "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
                      "quarter": "Q3"
                  }
              }
          ]
      },
      "usage": {
          "read_units": 6,
          "embed_total_tokens": 18
      }
  }
  ```

  ```json curl theme={null}
  {
    "result": {
      "hits": [
        {
          "_id": "vec2",
          "_score": 10.82421875,
          "fields": {
            "chunk_text": "Analysts suggest that AAPL'''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
            "quarter": "Q4"
          }
        },
        {
          "_id": "vec3",
          "_score": 6.49066162109375,
          "fields": {
            "chunk_text": "AAPL'''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production.",
            "quarter": "Q3"
          }
        },
        {
          "_id": "vec1",
          "_score": 5.3671875,
          "fields": {
            "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
            "quarter": "Q3"
          }
        }
      ]
    },
    "usage": {
      "embed_total_tokens": 18,
      "read_units": 1
    }
  }
  ```
</CodeGroup>

## Search with a sparse vector

To search a sparse index with a sparse vector representation of a query, use the [`query`](/reference/api/latest/data-plane/query) operation with the following parameters:

* `namespace`: The [namespace](/guides/index-data/indexing-overview#namespaces) to query. To use the default namespace, set to `"__default__"`.
* `sparse_vector`: The sparse vector values and indices.
* `top_k`: The number of results to return.
* `include_values`: Whether to include the vector values of the matching records in the response. Defaults to `false`.
* `include_metadata`: Whether to include the metadata of the matching records in the response. Defaults to `false`.

<Note>
  When querying with `top_k` over 1000, for optimal performance set `include_values=false` and `include_metadata=false` to return only IDs and scores.

  Since vectors values are retrieved from object storage, include them in your query results only when you need them (especially with higher `top_k` values). For details, see [Decrease latency](/guides/optimize/decrease-latency#avoid-including-vector-values-when-not-needed).
</Note>

For example, the following code uses a sparse vector representation of the query "What is AAPL's outlook, considering both product launches and market conditions?" to search for the 3 most similar vectors in the `example-namespace` namespace:

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  results = index.query(
      namespace="example-namespace",
      sparse_vector={
        "values": [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],
        "indices": [767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697]
      }, 
      top_k=3,
      include_metadata=True,
      include_values=False
  )

  print(results)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  const queryResponse = await index.namespace('example-namespace').query({
      sparseVector: {
          indices: [767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697],
          values: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
      },
      topK: 3,
      includeValues: false,
      includeMetadata: true
  });

  console.log(queryResponse);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import io.pinecone.unsigned_indices_model.QueryResponseWithUnsignedIndices;
  import io.pinecone.clients.Index;

  import java.util.*;

  public class SearchSparseIndex {
      public static void main(String[] args) throws InterruptedException {
          // Instantiate Pinecone class
          Pinecone pinecone = new Pinecone.Builder("YOUR_API_KEY").build();

          String indexName = "docs-example";

          Index index = pinecone.getIndexConnection(indexName);

          List<Long> sparseIndices = Arrays.asList(
                  767227209L, 1640781426L, 1690623792L, 2021799277L, 2152645940L,
                  2295025838L, 2443437770L, 2779594451L, 2956155693L, 3476647774L,
                  3818127854L, 428309169L);
          List<Float> sparseValues = Arrays.asList(
                  1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f,
                  1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f);

          QueryResponseWithUnsignedIndices queryResponse = index.query(3, null, sparseIndices, sparseValues, null, "example-namespace", null, false, true);
          System.out.println(queryResponse);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
  	ctx := context.Background()

  	pc, err := pinecone.NewClient(pinecone.NewClientParams{
  		ApiKey: "YOUR_API_KEY",
  	})
  	if err != nil {
  		log.Fatalf("Failed to create Client: %v", err)
  	}

  	// To get the unique host for an index,
  	// see https://docs.pinecone.io/guides/manage-data/target-an-index
  	idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
  	if err != nil {
  		log.Fatalf("Failed to create IndexConnection for Host: %v", err)
  	}

  	sparseValues := pinecone.SparseValues{
  		Indices: []uint32{767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697},
  		Values:  []float32{1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0},
  	}

  	res, err := idxConnection.QueryByVectorValues(ctx, &pinecone.QueryByVectorValuesRequest{
  		SparseValues:    &sparseValues,
  		TopK:            3,
  		IncludeValues:   false,
  		IncludeMetadata: true,
  	})
  	if err != nil {
  		log.Fatalf("Error encountered when querying by vector: %v", err)
  	} else {
  		fmt.Printf(prettifyStruct(res))
  	}
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var index = pinecone.Index("docs-example");

  var queryResponse = await index.QueryAsync(new QueryRequest {
      Namespace = "example-namespace",
      TopK = 4,
      SparseVector = new SparseValues
      {
          Indices = [767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697],
          Values = new[] { 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f },
      },
      IncludeValues = false,
      IncludeMetadata = true
  });

  Console.WriteLine(queryResponse);
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/query" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "sparseVector": {
              "values": [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],
              "indices": [767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697]
          },
          "namespace": "example-namespace",
          "topK": 4,
          "includeMetadata": true,
          "includeValues": false
      }'
  ```
</CodeGroup>

The results will look as follows. The most similar records are scored highest.

<CodeGroup>
  ```python Python theme={null}
  {'matches': [{'id': 'vec2',
                'metadata': {'category': 'technology',
                             'quarter': 'Q4',
                             'chunk_text': "Analysts suggest that AAPL'''s "
                                            'upcoming Q4 product launch event '
                                            'might solidify its position in the '
                                            'premium smartphone market.'},
                'score': 10.9042969,
                'values': []},
               {'id': 'vec3',
                'metadata': {'category': 'technology',
                             'quarter': 'Q3',
                             'chunk_text': "AAPL'''s strategic Q3 partnerships "
                                            'with semiconductor suppliers could '
                                            'mitigate component risks and '
                                            'stabilize iPhone production'},
                'score': 6.48010254,
                'values': []},
               {'id': 'vec1',
                'metadata': {'category': 'technology',
                             'quarter': 'Q3',
                             'chunk_text': 'AAPL reported a year-over-year '
                                            'revenue increase, expecting '
                                            'stronger Q3 demand for its flagship '
                                            'phones.'},
                'score': 5.3671875,
                'values': []}],
   'namespace': 'example-namespace',
   'usage': {'read_units': 1}}
  ```

  ```javascript JavaScript theme={null}
  { 
    matches: [
              { 
                id: 'vec2',
                score: 10.9042969,
                values: [],
                metadata: {
                  chunk_text: "Analysts suggest that AAPL'''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
                  category: 'technology',
                  quarter: 'Q4'
                }
              },
              {
                id: 'vec3',
                score: 6.48010254,
                values: [],
                metadata: {
                  chunk_text: "AAPL'''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production.",
                  category: 'technology',
                  quarter: 'Q3'
                }
              },
              {
                id: 'vec1',
                score: 5.3671875,
                values: [],
                metadata: {
                    chunk_text: 'AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.',
                    category: 'technology',
                    quarter: 'Q3'
                }
              }
            ],
    namespace: 'example-namespace',
    usage: {readUnits: 1}
  }
  ```

  ```java Java theme={null}
  class QueryResponseWithUnsignedIndices {
      matches: [ScoredVectorWithUnsignedIndices {
          score: 10.34375
          id: vec2
          values: []
          metadata: fields {
            key: "category"
            value {
              string_value: "technology"
            }
          }
          fields {
            key: "chunk_text"
            value {
              string_value: "Analysts suggest that AAPL\'\\\'\'s upcoming Q4 product launch event might solidify its position in the premium smartphone market."
            }
          }
          fields {
            key: "quarter"
            value {
              string_value: "Q4"
            }
          }
          
          sparseValuesWithUnsignedIndices: SparseValuesWithUnsignedIndices {
              indicesWithUnsigned32Int: []
              values: []
          }
      }, ScoredVectorWithUnsignedIndices {
          score: 5.8638916
          id: vec3
          values: []
          metadata: fields {
            key: "category"
            value {
              string_value: "technology"
            }
          }
          fields {
            key: "chunk_text"
            value {
              string_value: "AAPL\'\\\'\'s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production"
            }
          }
          fields {
            key: "quarter"
            value {
              string_value: "Q3"
            }
          }
          
          sparseValuesWithUnsignedIndices: SparseValuesWithUnsignedIndices {
              indicesWithUnsigned32Int: []
              values: []
          }
      }, ScoredVectorWithUnsignedIndices {
          score: 5.3671875
          id: vec1
          values: []
          metadata: fields {
            key: "category"
            value {
              string_value: "technology"
            }
          }
          fields {
            key: "chunk_text"
            value {
              string_value: "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones."
            }
          }
          fields {
            key: "quarter"
            value {
              string_value: "Q3"
            }
          }
          
          sparseValuesWithUnsignedIndices: SparseValuesWithUnsignedIndices {
              indicesWithUnsigned32Int: []
              values: []
          }
      }]
      namespace: example-namespace
      usage: read_units: 1

  }
  ```

  ```go Go theme={null}
  {
    "matches": [
      {
        "vector": {
          "id": "vec2",
          "metadata": {
            "category": "technology",
            "quarter": "Q4",
            "chunk_text": "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market."
          }
        },
        "score": 10.904296
      },
      {
        "vector": {
          "id": "vec3",
          "metadata": {
            "category": "technology",
            "quarter": "Q3",
            "chunk_text": "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production"
          }
        },
        "score": 6.4801025
      },
      {
        "vector": {
          "id": "vec1",
          "metadata": {
            "category": "technology",
            "quarter": "Q3",
            "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones"
          }
        },
        "score": 5.3671875
      }
    ],
    "usage": {
      "read_units": 1
    },
    "namespace": "example-namespace"
  }
  ```

  ```csharp C# theme={null}
  {
    "results": [],
    "matches": [
      {
        "id": "vec2",
        "score": 10.904297,
        "values": [],
        "metadata": {
          "category": "technology",
          "chunk_text": "Analysts suggest that AAPL\u0027\u0027\u0027s upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
          "quarter": "Q4"
        }
      },
      {
        "id": "vec3",
        "score": 6.4801025,
        "values": [],
        "metadata": {
          "category": "technology",
          "chunk_text": "AAPL\u0027\u0027\u0027s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production",
          "quarter": "Q3"
        }
      },
      {
        "id": "vec1",
        "score": 5.3671875,
        "values": [],
        "metadata": {
          "category": "technology",
          "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
          "quarter": "Q3"
        }
      }
    ],
    "namespace": "example-namespace",
    "usage": {
      "readUnits": 1
    }
  }
  ```

  ```json curl theme={null}
  {
    "results": [],
    "matches": [
      {
        "id": "vec2",
        "score": 10.9042969,
        "values": [],
        "metadata": {
          "chunk_text": "Analysts suggest that AAPL'''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
          "category": "technology",
          "quarter": "Q4"
        }
      },
      {
        "id": "vec3",
        "score": 6.48010254,
        "values": [],
        "metadata": {
          "chunk_text": "AAPL'''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production.",
          "category": "technology",
          "quarter": "Q3"
        }
      },
      {
        "id": "vec1",
        "score": 5.3671875,
        "values": [],
        "metadata": {
            "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
            "category": "technology",
            "quarter": "Q3"
        }
      }
    ],
    "namespace": "example-namespace",
    "usage": {
      "readUnits": 1
    }
  }
  ```
</CodeGroup>

## Search with a record ID

When you search with a record ID, Pinecone uses the sparse vector associated with the record as the query. To search a sparse index with a record ID, use the [`query`](/reference/api/latest/data-plane/query) operation with the following parameters:

* `namespace`: The [namespace](/guides/index-data/indexing-overview#namespaces) to query. To use the default namespace, set to `"__default__"`.
* `id`: The unique record ID containing the sparse vector to use as the query.
* `top_k`: The number of results to return.
* `include_values`: Whether to include the vector values of the matching records in the response. Defaults to `false`.
* `include_metadata`: Whether to include the metadata of the matching records in the response. Defaults to `false`.

<Note>
  When querying with `top_k` over 1000, for optimal performance set `include_values=false` and `include_metadata=false` to return only IDs and scores.

  Since vectors values are retrieved from object storage, include them in your query results only when you need them (especially with higher `top_k` values). For details, see [Decrease latency](/guides/optimize/decrease-latency#avoid-including-vector-values-when-not-needed).
</Note>

For example, the following code uses an ID to search for the 3 records in the `example-namespace` namespace that best match the sparse vector in the record:

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  index.query(
      namespace="example-namespace",
      id="rec2", 
      top_k=3,
      include_metadata=True,
      include_values=False
  )
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  const queryResponse = await index.namespace('example-namespace').query({
      id: 'rec2',
      topK: 3,
      includeValues: false,
      includeMetadata: true,
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import io.pinecone.unsigned_indices_model.QueryResponseWithUnsignedIndices;

  public class QueryExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");
          QueryResponseWithUnsignedIndices queryRespone = index.queryByVectorId(3, "rec2", "example-namespace", null, false, true);
          System.out.println(queryResponse);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      vectorId := "rec2"
      res, err := idxConnection.QueryByVectorId(ctx, &pinecone.QueryByVectorIdRequest{
          VectorId:      vectorId,
          TopK:          3,
          IncludeValues: false,
          IncludeMetadata: true,
      })
      if err != nil {
          log.Fatalf("Error encountered when querying by vector ID `%v`: %v", vectorId, err)
      } else {
          fmt.Printf(prettifyStruct(res.Matches))
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var queryResponse = await index.QueryAsync(new QueryRequest {
      Id = "rec2",
      Namespace = "example-namespace",
      TopK = 3,
      IncludeValues = false,
      IncludeMetadata = true
  });

  Console.WriteLine(queryResponse);
  ```

  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/query" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "id": "rec2",
          "namespace": "example-namespace",
          "topK": 3,
          "includeMetadata": true,
          "includeValues": false
      }'
  ```
</CodeGroup>

## Filter by required terms

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and is available only on the `2025-10` version of the API. See [limitations](#limitations) for details.
</Note>

When [searching with text](#search-with-text), you can specify a list of terms that must be present in each lexical search result. This is especially useful for:

* **Precision filtering**: Ensuring specific entities or concepts appear in results
* **Quality control**: Filtering out results that don't contain essential keywords
* **Domain-specific searches**: Requiring domain-specific terminology in results
* **Entity-based filtering**: Ensuring specific people, places, or things are mentioned

To filter by required terms, add `match_terms` to your query, specifying the `terms` to require and the `strategy` to use. Currently, `all` is the only strategy supported (all terms must be present).

For example, the following request searches for records about Tesla's stock performance while ensuring both "Tesla" and "stock" appear in each result:

```bash curl theme={null}
PINECONE_API_KEY="YOUR_API_KEY"
INDEX_HOST="INDEX_HOST"

curl "https://$INDEX_HOST/records/namespaces/example-namespace/search" \
  -H "Content-Type: application/json" \
  -H "Api-Key: $PINECONE_API_KEY" \
  -H "X-Pinecone-Api-Version: unstable" \
  -d '{
        "query": {
          "inputs": { "text": "What is the current outlook for Tesla stock performance?" },
          "top_k": 3,
          "match_terms": {
            "terms": ["Tesla", "stock"],
            "strategy": "all"
          }
        },
        "fields": ["chunk_text"]
    }'
```

The response includes only records that contain both "Tesla" and "stock":

```json  theme={null}
{
  "result": {
    "hits": [
      {
        "_id": "tesla_q4_earnings",
        "_score": 9.82421875,
        "fields": {
          "chunk_text": "Tesla stock surged 8% in after-hours trading following strong Q4 earnings that exceeded analyst expectations. The company reported record vehicle deliveries and improved profit margins."
        }
      },
      {
        "_id": "tesla_competition_analysis",
        "_score": 7.49066162109375,
        "fields": {
          "chunk_text": "Tesla stock faces increasing competition from traditional automakers entering the electric vehicle market. However, analysts maintain that Tesla's technological lead and brand recognition provide significant advantages."
        }
      },
      {
        "_id": "tesla_production_update",
        "_score": 6.3671875,
        "fields": {
          "chunk_text": "Tesla stock performance is closely tied to production capacity at its Gigafactories. Recent expansion announcements suggest the company is positioning for continued growth in global markets."
        }
      }
    ]
  },
  "usage": {
    "embed_total_tokens": 18,
    "read_units": 1
  }
}
```

Without the `match_terms` filter, you might get results like:

* "Tesla cars are popular in California" (mentions Tesla but not stock)
* "Stock market volatility affects tech companies" (mentions stock but not Tesla)
* "Electric vehicle sales are growing" (neither Tesla nor stock)

### Limitations

* **Integrated indexes only**: Filtering by required terms is supported only for [indexes with integrated embedding](/guides/index-data/indexing-overview#integrated-embedding).
* **Post-processing filter**: The filtering happens after the initial query, so potential matches that weren't included in the initial `top_k` results won't appear in the final results
* **No phrase matching**: Terms are matched individually in any order and location.
* **No case-sensitivity**: Terms are normalized during processing.

---

# Source: https://docs.pinecone.io/troubleshooting/limitations-of-querying-by-id.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Limitations of querying by ID

When [querying by record ID](/guides/search/semantic-search#search-with-a-record-id), even with a high `topK`, the response is not guaranteed to include the record with the specified ID.

## Approximate nearest neighbor (ANN)

Approximate nearest neighbor algorithms are designed to quickly find the closest matches to a given data point within large datasets with reasonable accuracy rather than perfect precision. Depending on the data, ANN may have a slightly lower accuracy than Known Nearest Neighbor (KNN) algorithms, but will have significantly lower read costs and latency than KNN. This is one of the key features of ANN.

ANN algorithms assess broad data clusters to find matches. Some of these clusters might be ignored even if they contain relevant records simply because their overall similarity to the query is lower, because the algorithm aims to optimize the search by focusing on areas with a higher density of potential matches.

See our learning center for more information on [ANN algorithms](https://www.pinecone.io/learn/a-developers-guide-to-ann-algorithms/).

## Recommendations

### Perform a fetch instead of a query

Results from [Fetch](/guides/manage-data/fetch-data) are guaranteed to include the record with the specified ID.

### Use metadata filtering

A [metadata filter](/guides/index-data/indexing-overview#metadata) in an ANN search effectively narrows the dataset to a more relevant subset, fine-tuning the search process. By explicitly excluding less relevant clusters from the outset, the search is performed among a group of records more closely related to the query, thereby increasing the efficiency and accuracy of the search.

---

# Source: https://docs.pinecone.io/guides/manage-data/list-record-ids.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List record IDs

> List the IDS of records in an index namespace.

You can list the IDs of all records in a [namespace](/guides/index-data/indexing-overview#namespaces) or just the records with a common ID prefix.

Using `list` to get record IDs and not the associated data is a cheap and fast way to check [upserts](/guides/index-data/upsert-data).

<Note>
  The `list` endpoint is supported only for serverless indexes.
</Note>

## List the IDs of all records in a namespace

To list the IDs of all records in the namespace of a serverless index, pass only the `namespace` parameter:

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key='YOUR_API_KEY')

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  for ids in index.list(namespace='example-namespace'):
      print(ids)

  # Response:
  # ['doc1#chunk1', 'doc1#chunk2', 'doc1#chunk3']
  ```

  ```js JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';
  const pc = new Pinecone();

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST").namespace("example-namespace");

  const results = await index.listPaginated();
  console.log(results);
  // {
  //   vectors: [
  //     { id: 'doc1#01' }, { id: 'doc1#02' }, { id: 'doc1#03' },
  //     { id: 'doc1#04' }, { id: 'doc1#05' },  { id: 'doc1#06' },
  //     { id: 'doc1#07' }, { id: 'doc1#08' }, { id: 'doc1#09' },
  //     ...
  //   ],
  //   pagination: {
  //     next: 'eyJza2lwX3Bhc3QiOiJwcmVUZXN0LS04MCIsInByZWZpeCI6InByZVRlc3QifQ=='
  //   },
  //   namespace: 'example-namespace',
  //   usage: { readUnits: 1 }
  // }

  // Fetch the next page of results
  await index.listPaginated({ prefix: 'doc1#', paginationToken: results.pagination.next});
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import io.pinecone.proto.ListResponse;

  public class ListExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");
          // get the pagination token
          String paginationToken = index.list("example-namespace", 3).getPagination().getNext();
          // get vectors with limit 3 with the paginationToken obtained from the previous step
          ListResponse listResponse = index.list("example-namespace", 3, paginationToken);
      }
  }

  // Response:
  // vectors {
  //   id: "doc1#chunk1"
  // }
  // vectors {
  //   id: "doc1#chunk2"
  // }
  // vectors {
  //   id: "doc2#chunk1"
  // }
  // vectors {
  //   id: "doc3#chunk1"
  // }
  // pagination {
  //   next: "eyJza2lwX3Bhc3QiOiJhbHN0cm9lbWVyaWEtcGVydXZpYW4iLCJwcmVmaXgiOm51bGx9"
  // }
  // namespace: "example-namespace"
  // usage {
  //   read_units: 1
  // }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      limit := uint32(3)

      res, err := idxConnection.ListVectors(ctx, &pinecone.ListVectorsRequest{
          Limit:  &limit,
      })
      if len(res.VectorIds) == 0 {
          fmt.Println("No vectors found")
  	} else {
  		fmt.Printf(prettifyStruct(res))
  	}
  }

  // Response:
  // {
  //   "vector_ids": [
  //     "doc1#chunk1",
  //     "doc1#chunk2",
  //     "doc1#chunk3"
  //   ],
  //   "usage": {
  //     "read_units": 1
  //   },
  //   "next_pagination_token": "eyJza2lwX3Bhc3QiOiIwMDBkMTc4OC0zMDAxLTQwZmMtYjZjNC0wOWI2N2I5N2JjNDUiLCJwcmVmaXgiOm51bGx9"
  // }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var listResponse = await index.ListAsync(new ListRequest {
      Namespace = "example-namespace",
  });

  Console.WriteLine(listResponse);

  // Response:
  // {
  //   "vectors": [
  //     {
  //       "id": "doc1#chunk1"
  //     },
  //     {
  //       "id": "doc1#chunk2"
  //     },
  //     {
  //       "id": "doc1#chunk3"
  //     }
  //   ],
  //   "pagination": "eyJza2lwX3Bhc3QiOiIwMDBkMTc4OC0zMDAxLTQwZmMtYjZjNC0wOWI2N2I5N2JjNDUiLCJwcmVmaXgiOm51bGx9",
  //   "namespace": "example-namespace",
  //   "usage": {
  //     "readUnits": 1
  //   }
  // }
  ```

  ```shell curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X GET "https://$INDEX_HOST/vectors/list?namespace=example-namespace" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"

  # Response:
  # {
  #   "vectors": [
  #     { "id": "doc1#chunk1" },
  #     { "id": "doc1#chunk2" },
  #     { "id": "doc1#chunk3" },
  #     { "id": "doc1#chunk4" },
  #    ...
  #   ],
  #   "pagination": {
  #     "next": "c2Vjb25kY2FsbA=="
  #   },
  #   "namespace": "example-namespace",
  #   "usage": {
  #     "readUnits": 1
  #   }
  # }
  ```
</CodeGroup>

## List the IDs of records with a common prefix

ID prefixes enable you to query segments of content. Use the `list` endpoint to list all of the records with the common prefix. For more details, see [Use structured IDs](/guides/index-data/data-modeling#use-structured-ids).

## Paginate through results

The `list` endpoint returns up to 100 IDs per page at a time by default. If the `limit` parameter is passed, `list` returns up to that number of IDs per page instead. For example, if `limit=3`, up to 3 IDs be returned per page. Whenever there are additional IDs to return, the response also includes a `pagination_token` for fetching the next page of IDs.

### Implicit pagination

When using the Python SDK, `list` paginates automatically.

```python Python theme={null}
from pinecone.grpc import PineconeGRPC as Pinecone

pc = Pinecone(api_key='YOUR_API_KEY')

# To get the unique host for an index, 
# see https://docs.pinecone.io/guides/manage-data/target-an-index
index = pc.Index(host="INDEX_HOST")

for ids in index.list(namespace='example-namespace'):
    print(ids)

# Response:
# ['doc1#chunk1', 'doc1#chunk2', 'doc1#chunk3']
# ['doc1#chunk4', 'doc1#chunk5', 'doc1#chunk6']
# ...
```

### Manual pagination

When using the Node.js SDK, Java SDK, Go SDK, .NET SDK, or REST API, you must manually fetch each page of results. You can also manually paginate with the Python SDK using `list_paginated()`.

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key='YOUR_API_KEY')

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  namespace = 'example-namespace'

  # For manual control over pagination
  results = index.list_paginated(
      prefix='pref',
      limit=3,
      namespace='example-namespace'
  )
  print(results.namespace)
  print([v.id for v in results.vectors])
  print(results.pagination.next)
  print(results.usage)

  # Results:
  # ['10103-0', '10103-1', '10103-10']
  # eyJza2lwX3Bhc3QiOiIxMDEwMy0=
  # {'read_units': 1}
  ```

  ```js JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';
  const pc = new Pinecone();

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST").namespace("example-namespace");

  const results = await index.listPaginated({ prefix: 'doc1#', limit: 3 });
  console.log(results);

  // Response:
  // {
  //   vectors: [
  //     { id: 'doc1#01' }, { id: 'doc1#02' }, { id: 'doc1#03' }
  //   ],
  //   pagination: {
  //     next: 'eyJza2lwX3Bhc3QiOiJwcmVUZXN0LSCIsInByZWZpeCI6InByZVRlc3QifQ=='
  //   },
  //   namespace: 'example-namespace',
  //   usage: { readUnits: 1 }
  // }
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import io.pinecone.proto.ListResponse;

  public class ListExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");
          ListResponse listResponse = index.list("example-namespace", "doc1#" 2); /* Note: You must include an ID prefix to list vector IDs. */
          System.out.println(listResponse.getVectorsList());
          System.out.println(listResponse.getPagination());
      }
  }

  // Response:
  // vectors {
  //   id: "doc1#chunk1"
  // }
  // vectors {
  //   id: "doc1#chunk2"
  // }
  // pagination {
  //   next: "eyJza2lwX3Bhc3QiOiJhbHN0cm9lbWVyaWEtcGVydXZpYW4iLCJwcmVmaXgiOm51bGx9"
  // }
  // namespace: "example-namespace"
  // usage {
  //   read_units: 1
  // }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      limit := uint32(3)

      res, err := idxConnection.ListVectors(ctx, &pinecone.ListVectorsRequest{
          Limit:  &limit,
      })
      if len(res.VectorIds) == 0 {
          fmt.Println("No vectors found")
      } else {
          fmt.Printf(prettifyStruct(res))
      }
  }

  // Response:
  // {
  //   "vector_ids": [
  //     "doc1#chunk1",
  //     "doc1#chunk2",
  //     "doc1#chunk3"
  //   ],
  //   "usage": {
  //     "read_units": 1
  //   },
  //   "next_pagination_token": "eyJza2lwX3Bhc3QiOiIwMDBkMTc4OC0zMDAxLTQwZmMtYjZjNC0wOWI2N2I5N2JjNDUiLCJwcmVmaXgiOm51bGx9"
  // }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var listResponse = await index.ListAsync(new ListRequest {
      Namespace = "example-namespace",
      Prefix = "document1#",
  });

  Console.WriteLine(listResponse);

  // Response:
  // {
  //   "vectors": [
  //     {
  //       "id": "doc1#chunk1"
  //     },
  //     {
  //       "id": "doc1#chunk2"
  //     },
  //     {
  //       "id": "doc1#chunk3"
  //     }
  //   ],
  //   "pagination": "eyJza2lwX3Bhc3QiOiIwMDBkMTc4OC0zMDAxLTQwZmMtYjZjNC0wOWI2N2I5N2JjNDUiLCJwcmVmaXgiOm51bGx9",
  //   "namespace": "example-namespace",
  //   "usage": {
  //     "readUnits": 1
  //   }
  // }
  ```

  ```shell curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X GET "https://$INDEX_HOST/vectors/list?namespace=example-namespace" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"

  # Response:
  # {
  #   "vectors": [
  #     { "id": "doc1#chunk1" },
  #     { "id": "doc1#chunk2" },
  #     { "id": "doc1#chunk3" },
  #     { "id": "doc1#chunk4" },
  #    ...
  #   ],
  #   "pagination": {
  #     "next": "c2Vjb25kY2FsbA=="
  #   },
  #   "namespace": "example-namespace",
  #   "usage": {
  #     "readUnits": 1
  #   }
  # }
  ```
</CodeGroup>

Then, to get the next batch of IDs, use the returned `pagination_token`:

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key='YOUR_API_KEY')

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  namespace = 'example-namespace'

  results = index.list_paginated(
      prefix='pref',
      limit=3,
      namespace='example-namespace',
      pagination_token='eyJza2lwX3Bhc3QiOiIxMDEwMy0='
  )
  print(results.namespace)
  print([v.id for v in results.vectors])
  print(results.pagination.next)
  print(results.usage)

  # Response:
  # ['10103-0', '10103-1', '10103-10']
  # xndlsInByZWZpeCI6IjEwMTAzIn0==
  # {'read_units': 1}
  ```

  ```js JavaScript theme={null}
  await index.listPaginated({ prefix: 'doc1#', limit: 3, paginationToken: results.pagination.next});

  // Response:
  // {
  //   vectors: [
  //     { id: 'doc1#10' }, { id: 'doc1#11' }, { id: 'doc1#12' }
  //   ],
  //   pagination: {
  //     next: 'dfajlkjfdsoijeowjoDJFKLJldLIFf34KFNLDSndaklqoLQJORN45afdlkJ=='
  //   },
  //   namespace: 'example-namespace',
  //   usage: { readUnits: 1 }
  // }
  ```

  ```java Java theme={null}
  listResponse = index.list("example-namespace", "doc1#", "eyJza2lwX3Bhc3QiOiJ2MTg4IiwicHJlZml4IjpudWxsfQ==");
  System.out.println(listResponse.getVectorsList());

  // Response:
  // vectors {
  //   id: "doc1#chunk3"
  // }
  // vectors {
  //   id: "doc1#chunk4"
  // }
  // vectors {
  //   id: "doc1#chunk5"
  // }
  // vectors {
  //   id: "doc1#chunk6"
  // }
  // vectors {
  //   id: "doc1#chunk7"
  // }
  // vectors {
  //   id: "doc1#chunk8"
  // }
  // pagination {
  //   next: "eyJza2lwX3Bhc3QiOiJhbHN0cm9lbWVyaWEtcGVydXZpYW4iLCJwcmVmaXgiOm51bGx9"
  // }
  // namespace: "example-namespace"
  // usage {
  //   read_units: 1
  // }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      limit := uint32(3)
      paginationToken := "dfajlkjfdsoijeowjoDJFKLJldLIFf34KFNLDSndaklqoLQJORN45afdlkJ=="

      res, err := idxConnection.ListVectors(ctx, &pinecone.ListVectorsRequest{
          Limit:  &limit,
          PaginationToken: &paginationToken,
      })
      if len(res.VectorIds) == 0 {
          fmt.Println("No vectors found")
  	} else {
  	    fmt.Printf(prettifyStruct(res))
  	}
  }

  // Response:
  // {
  //   "vector_ids": [
  //     "doc1#chunk4",
  //     "doc1#chunk5",
  //     "doc1#chunk6"
  //   ],
  //   "usage": {
  //     "read_units": 1
  //   },
  //   "next_pagination_token": "eyJza2lwX3Bhc3QiOiIwMDBkMTc4OC0zMDAxLTQwZmMtYjZjNC0wOWI2N2I5N2JjNDUiLCJwcmVmaXgiOm51bGx9"
  // }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var listResponse = await index.ListAsync(new ListRequest {
      Namespace = "example-namespace",
      Prefix = "document1#",
      PaginationToken= "eyJza2lwX3Bhc3QiOiIwMDBkMTc4OC0zMDAxLTQwZmMtYjZjNC0wOWI2N2I5N2JjNDUiLCJwcmVmaXgiOm51bGx9",
  });

  Console.WriteLine(listResponse);

  // Response:
  // {
  //   "vectors": [
  //     {
  //       "id": "doc1#chunk4"
  //     },
  //     {
  //       "id": "doc1#chunk5"
  //     },
  //     {
  //       "id": "doc1#chunk6"
  //     }
  //   ],
  //   "pagination": "dfajlkjfdsoijeowjoDJFKLJldLIFf34KFNLDSndaklqoLQJORN45afdlkJ==",
  //   "namespace": "example-namespace",
  //   "usage": {
  //     "readUnits": 1
  //   }
  // }
  ```

  ```shell curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X GET "https://$INDEX_HOST/vectors/list?namespace=example-namespace&paginationToken=c2Vjb25kY2FsbA%3D%3D" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"

  # Response:
  # {
  #   "vectors": [
  #     { "id": "doc2#chunk1" },
  #     { "id": "doc2#chunk1" },
  #     { "id": "doc2#chunk1" },
  #     { "id": "doc2#chunk1" },
  #    ...
  #   ],
  #   "pagination": {
  #     "next": "mn23b4jB3Y9jpsS1"
  #   },
  #   "namespace": "example-namespace",
  #   "usage": {
  #     "readUnits": 1
  #   }
  # }
  ```
</CodeGroup>

When there are no more IDs to return, the response does not includes a `pagination_token`:

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key='YOUR_API_KEY')

  index = pc.Index(host="INDEX_HOST")

  namespace = 'example-namespace'

  results = index.list_paginated(
      prefix='10103',
      limit=3,
      pagination_token='xndlsInByZWZpeCI6IjEwMTAzIn0=='
  )

  print(results.namespace)
  print([v.id for v in results.vectors])
  print(results.pagination.next)
  print(results.usage)

  # Response:
  # ['10103-4', '10103-5', '10103-6']
  # {'read_units': 1}
  ```

  ```js JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';
  const pc = new Pinecone();

  const index = pc.index("INDEX_NAME", "INDEX_HOST").namespace("example-namespace");

  const results = await index.listPaginated({ prefix: 'doc1#' });
  console.log(results);

  // Response:
  // {
  //   vectors: [
  //     { id: 'doc1#19' }, { id: 'doc1#20' }, { id: 'doc1#21' }
  //   ],
  //   namespace: 'example-namespace',
  //   usage: { readUnits: 1 }
  // }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      limit := uint32(3)
      paginationToken := "eyJza2lwX3Bhc3QiOiIwMDBkMTc4OC0zMDAxLTQwZmMtYjZjNC0wOWI2N2I5N2JjNDUiLCJwcmVmaXgiOm51bGx9"

      res, err := idxConnection.ListVectors(ctx, &pinecone.ListVectorsRequest{
          Limit:  &limit,
          paginationToken: &paginationToken,
      })
      if len(res.VectorIds) == 0 {
          fmt.Println("No vectors found")
    	} else {
  	    fmt.Printf(prettifyStruct(res))
  	}
  }

  // Response:
  // {
  //   "vector_ids": [
  //     "doc1#chunk7",
  //     "doc1#chunk8",
  //     "doc1#chunk9"
  //   ],
  //   "usage": {
  //     "read_units": 1
  //   }
  // }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var index = pinecone.Index(host: "INDEX_HOST");

  var listResponse = await index.ListAsync(new ListRequest {
      Namespace = "example-namespace",
      Prefix = "document1#",
      PaginationToken= "dfajlkjfdsoijeowjoDJFKLJldLIFf34KFNLDSndaklqoLQJORN45afdlkJ==",
  });

  Console.WriteLine(listResponse);

  // Response:
  // {
  //   "vectors": [
  //     {
  //       "id": "doc1#chunk7"
  //     },
  //     {
  //       "id": "doc1#chunk8"
  //     },
  //     {
  //       "id": "doc1#chunk9"
  //     }
  //   ],
  //   "pagination": null,
  //   "namespace": "example-namespace",
  //   "usage": {
  //     "readUnits": 1
  //   }
  // }
  ```

  ```shell curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X GET "https://$INDEX_HOST/vectors/list?namespace=example-namespace&paginationToken=mn23b4jB3Y9jpsS1" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  # Response:
  # {
  #   "vectors": [
  #     { "id": "doc3#chunk1" },
  #     { "id": "doc5#chunk2" },
  #     { "id": "doc5#chunk3" },
  #     { "id": "doc5#chunk4" },
  #    ...
  #   ],
  #   "namespace": "example-namespace",
  #   "usage": {
  #     "readUnits": 1
  #   }
  # }
  ```
</CodeGroup>

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/list.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List vector IDs

> List the IDs of vectors in a single namespace of a serverless index. An optional prefix can be passed to limit the results to IDs with a common prefix.

Returns up to 100 IDs at a time by default in sorted order (bitwise "C" collation). If the `limit` parameter is set, `list` returns up to that number of IDs instead. Whenever there are additional IDs to return, the response also includes a `pagination_token` that you can use to get the next batch of IDs. When the response does not include a `pagination_token`, there are no more IDs to return.

For guidance and examples, see [List record IDs](https://docs.pinecone.io/guides/manage-data/list-record-ids).

**Note:** `list` is supported only for serverless indexes.

<RequestExample>
  ```shell curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X GET "https://$INDEX_HOST/vectors/list?namespace=example-namespace&prefix=doc1#&limit=3" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "vectors": [
      { "id": "doc1#chunk1" },
      { "id": "doc1#chunk2" },
      { "id": "doc1#chunk3" }
    ],
    "pagination": {
      "next": "c2Vjb25kY2FsbA=="
    },
    "namespace": "example-namespace",
    "usage": {
      "readUnits": 1
    }
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml get /vectors/list
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /vectors/list:
    get:
      tags:
        - Vector Operations
      summary: List vector IDs
      description: >-
        List the IDs of vectors in a single namespace of a serverless index. An
        optional prefix can be passed to limit the results to IDs with a common
        prefix.


        Returns up to 100 IDs at a time by default in sorted order (bitwise "C"
        collation). If the `limit` parameter is set, `list` returns up to that
        number of IDs instead. Whenever there are additional IDs to return, the
        response also includes a `pagination_token` that you can use to get the
        next batch of IDs. When the response does not include a
        `pagination_token`, there are no more IDs to return.


        For guidance and examples, see [List record
        IDs](https://docs.pinecone.io/guides/manage-data/list-record-ids).


        **Note:** `list` is supported only for serverless indexes.
      operationId: listVectors
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: query
          name: prefix
          description: The vector IDs to fetch. Does not accept values containing spaces.
          schema:
            type: string
          style: form
        - in: query
          name: limit
          description: Max number of IDs to return per page.
          schema:
            default: 100
            type: integer
            format: int64
          style: form
        - in: query
          name: paginationToken
          description: Pagination token to continue a previous listing operation.
          schema:
            type: string
          style: form
        - in: query
          name: namespace
          description: >-
            The namespace to list vectors from. If not provided, the default
            namespace is used.
          schema:
            type: string
          style: form
      responses:
        '200':
          description: A successful response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ListResponse'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    ListResponse:
      description: The response for the `list` operation.
      type: object
      properties:
        vectors:
          example:
            - id: document1#abb
            - id: document1#abc
          title: A list of ids
          type: array
          items:
            $ref: '#/components/schemas/ListItem'
        pagination:
          $ref: '#/components/schemas/Pagination'
        namespace:
          example: example-namespace
          description: The namespace of the vectors.
          type: string
        usage:
          $ref: '#/components/schemas/Usage'
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    ListItem:
      type: object
      properties:
        id:
          example: document1#abb
          type: string
    Pagination:
      type: object
      properties:
        next:
          example: Tm90aGluZyB0byBzZWUgaGVyZQo=
          type: string
    Usage:
      type: object
      properties:
        readUnits:
          example: 5
          description: The number of read units consumed by this operation.
          type: integer
          format: int64
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/admin/list_api_keys.md

# Source: https://docs.pinecone.io/reference/api/2025-10/admin-assistant/list_api_keys.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List API keys

> List all API keys in a project.

<RequestExample>
  ```bash curl theme={null}
  curl -X GET "https://api.pinecone.io/admin/projects" \
      -H "Authorization: Bearer $PINECONE_ACCESS_TOKEN" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "data": [
      {
        "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
        "name": "string",
        "project_id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
        "roles": [
          "ProjectEditor"
        ]
      }
    ]
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/admin_2025-10.oas.yaml get /admin/projects/{project_id}/api-keys
openapi: 3.0.3
info:
  title: Pinecone Admin API
  description: |
    Provides an API for managing a Pinecone organization and its resources.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - BearerAuth: []
tags:
  - name: API Keys
    description: Actions that manage API Keys.
  - name: Organizations
    description: Actions that manage organizations.
  - name: Projects
    description: Actions that manage projects.
paths:
  /admin/projects/{project_id}/api-keys:
    get:
      tags:
        - API Keys
      summary: List API keys
      description: List all API keys in a project.
      operationId: list_project_api_keys
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: project_id
          description: Project ID
          required: true
          schema:
            type: string
            format: uuid
          style: simple
      responses:
        '200':
          description: A list of API keys.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ListApiKeysResponse'
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
        4XX:
          description: Unexpected error on request.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
components:
  schemas:
    ListApiKeysResponse:
      type: object
      properties:
        data:
          type: array
          items:
            $ref: '#/components/schemas/APIKey'
      required:
        - data
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, or `UNPROCESSABLE_ENTITY`.        
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    APIKey:
      description: The details of an API key, without the secret.
      type: object
      properties:
        id:
          description: The unique ID of the API key.
          type: string
          format: uuid
        name:
          description: The name of the API key.
          type: string
        project_id:
          description: The ID of the project containing the API key.
          type: string
          format: uuid
        roles:
          description: The roles assigned to the API key.
          type: array
          items:
            example: ProjectEditor
            description: >-
              A role that can be assigned to an API key.

              Possible values: `ProjectEditor`, `ProjectViewer`,
              `ControlPlaneEditor`, `ControlPlaneViewer`, `DataPlaneEditor`, or
              `DataPlaneViewer`.
            x-enum:
              - ProjectEditor
              - ProjectViewer
              - ControlPlaneEditor
              - ControlPlaneViewer
              - DataPlaneEditor
              - DataPlaneViewer
            type: string
      required:
        - id
        - name
        - project_id
        - roles
  securitySchemes:
    BearerAuth:
      type: http
      scheme: bearer
      description: >
        An [access
        token](https://docs.pinecone.io/guides/organizations/manage-service-accounts#retrieve-an-access-token)
        must be provided in the `Authorization` header using the `Bearer`
        scheme.

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/assistant/list_assistants.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List assistants

> List of all assistants in a project.

For guidance and examples, see [Manage assistants](https://docs.pinecone.io/guides/assistant/manage-assistants#list-assistants-for-a-project).

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -X GET "https://api.pinecone.io/assistant/assistants" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "assistants": [
      {
        "name": "example-assistant",
        "instructions": "Use American English for spelling and grammar.",
        "metadata": {"team": "customer-support", "version": "1.0"},
        "status": "Ready",
        "host": "https://prod-1-data.ke.pinecone.io",
        "created_at": "2025-10-01T12:30:00Z",
        "updated_at": "2025-10-01T12:45:00Z"
      }
    ]
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/assistant_control_2025-10.oas.yaml get /assistants
openapi: 3.0.3
info:
  title: Pinecone Assistant Control Plane API
  description: >-
    Pinecone Assistant Engine is a context engine to store and retrieve relevant
    knowledge  from millions of documents at scale. This API supports creating
    and managing assistants. 
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io/assistant
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Assistants
    description: Actions that manage Assistants
paths:
  /assistants:
    get:
      tags:
        - Manage Assistants
      summary: List assistants
      description: >-
        List of all assistants in a project.


        For guidance and examples, see [Manage
        assistants](https://docs.pinecone.io/guides/assistant/manage-assistants#list-assistants-for-a-project).
      operationId: list_assistants
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      responses:
        '200':
          description: List all assistants in a project.
          content:
            application/json:
              schema:
                description: The list of assistants that exist in the project.
                type: object
                properties:
                  assistants:
                    type: array
                    items:
                      $ref: '#/components/schemas/Assistant'
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    Assistant:
      description: Describes the configuration and status of a Pinecone Assistant.
      type: object
      properties:
        name:
          example: example-assistant
          description: >
            The name of the assistant. Resource name must be 1-63 characters
            long, start and end with an alphanumeric character, and consist only
            of lower case alphanumeric characters or '-'.
          type: string
          minLength: 1
          maxLength: 63
        instructions:
          nullable: true
          description: >-
            Description or directive for the assistant to apply to all
            responses.
          type: string
        metadata:
          description: >-
            Optional metadata associated with the assistant. Metadata is a JSON
            object that can store custom organizational data, tags, and
            attributes.
          nullable: true
          type: object
        status:
          x-enum:
            - Initializing
            - Failed
            - Ready
            - Terminating
            - InitializationFailed
          type: string
        host:
          description: The host where the assistant is deployed.
          type: string
        created_at:
          type: string
          format: date-time
        updated_at:
          type: string
          format: date-time
      required:
        - name
        - status
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: Uploaded file can only currently be either a pdf or txt file
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: Pinecone API Key

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/list_collections.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List collections

> List all collections in a project.
Serverless indexes do not support collections.


<RequestExample>
  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X GET "https://api.pinecone.io/collections" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
      "collections": [
          {
              "name": "example-collection1",
              "status": "Ready",
              "environment": "us-east-1-aws",
              "size": 3081918,
              "vector_count": 99,
              "dimension": 3
          },
          {
              "name": "example-collection1",
              "status": "Ready",
              "environment": "us-east-1-aws",
              "size": 160087040000000,
              "vector_count": 10000000,
              "dimension": 1536
          }
      ]
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml get /collections
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /collections:
    get:
      tags:
        - Manage Indexes
      summary: List collections
      description: |
        List all collections in a project.
        Serverless indexes do not support collections.
      operationId: list_collections
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      responses:
        '200':
          description: List all the collections in your current project.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/CollectionList'
              examples:
                multiple-collections:
                  summary: Multiple collections with different states
                  value:
                    collections:
                      - dimension: 3
                        environment: us-east1-gcp
                        name: small-collection
                        size: 3126700
                        status: Ready
                        vector_count: 99
                      - dimension: 3
                        environment: us-east1-gcp
                        name: small-collection-new
                        size: 3126700
                        status: Initializing
                        vector_count: 99
                      - dimension: 1536
                        environment: us-east1-gcp
                        name: big-collection
                        size: 160087040000000
                        status: Ready
                        vector_count: 10000000
                no-collections:
                  summary: No collections created yet
                  value:
                    collections: []
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    CollectionList:
      description: The list of collections that exist in the project.
      type: object
      properties:
        collections:
          description: List of collections in the project
          type: array
          items:
            $ref: '#/components/schemas/CollectionModel'
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    CollectionModel:
      description: >-
        The CollectionModel describes the configuration and status of a Pinecone
        collection.
      type: object
      properties:
        name:
          example: example-collection
          description: The name of the collection.
          type: string
        size:
          example: 10000000
          description: The size of the collection in bytes.
          type: integer
          format: int64
        status:
          example: Initializing
          description: |-
            The status of the collection.
            Possible values: `Initializing`, `Ready`, or `Terminating`.
          x-enum:
            - Initializing
            - Ready
            - Terminating
          type: string
        dimension:
          example: 1536
          description: >-
            The dimension of the vectors stored in each record held in the
            collection.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        vector_count:
          example: 120000
          description: The number of records stored in the collection.
          type: integer
          format: int32
        environment:
          example: us-east1-gcp
          description: The environment where the collection is hosted.
          type: string
      required:
        - name
        - status
        - environment
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/assistant/list_files.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List Files

> List all files in an assistant, with an option to filter files with metadata.

For guidance and examples, see [Manage files](https://docs.pinecone.io/guides/assistant/manage-files#list-files-in-an-assistant).

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl -X GET "https://prod-1-data.ke.pinecone.io/assistant/files/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/assistant_data_2025-10.oas.yaml get /files/{assistant_name}
openapi: 3.0.3
info:
  title: Pinecone Assistant Data Plane API
  description: >-
    Pinecone Assistant Engine is a context engine to store and retrieve relevant
    knowledge from millions of documents at scale. This API supports
    interactions with assistants.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{assistant_host}
    variables:
      assistant_host:
        default: unknown
        description: host of the created assistant
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Assistants
    description: Actions that manage Assistants
paths:
  /files/{assistant_name}:
    get:
      tags:
        - Manage Assistants
      summary: List Files
      description: >-
        List all files in an assistant, with an option to filter files with
        metadata.


        For guidance and examples, see [Manage
        files](https://docs.pinecone.io/guides/assistant/manage-files#list-files-in-an-assistant).
      operationId: list_files
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: assistant_name
          description: The name of the assistant to list files for.
          required: true
          schema:
            type: string
          example: test-assistant
          style: simple
        - in: query
          name: filter
          description: Optional JSON-encoded metadata filter for files.
          schema:
            type: string
            format: json
          example: '{"genre":{"$eq":"comedy"}}'
          style: form
      responses:
        '200':
          description: >-
            This operation returns a list of all files that you have previously
            uploaded, and which are associated with the given assistant name.
          content:
            application/json:
              schema:
                description: The list of files that exist in the assistant
                type: object
                properties:
                  files:
                    type: array
                    items:
                      $ref: '#/components/schemas/AssistantFileModel'
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Assistant not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                assistant-not-found:
                  summary: Assistant not found.
                  value:
                    error:
                      code: NOT_FOUND
                      message: Assistant "example-assistant" not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    AssistantFileModel:
      description: The response format for a successful file upload request.
      type: object
      properties:
        name:
          type: string
        id:
          type: string
          format: uuid
        metadata:
          nullable: true
          type: object
        created_on:
          type: string
          format: date-time
        updated_on:
          type: string
          format: date-time
        status:
          description: >-
            The current state of the uploaded file. Possible values:

            - `Processing`: File is being processed (parsed, chunked, embedded)

            - `Available`: Processing completed successfully; file is ready for
            use

            - `Deleting`: Deletion has been initiated but not yet completed

            - `ProcessingFailed`: Processing failed with an error


            Note: Once a file is deleted, the API returns 404 Not Found instead
            of a file object.
          x-enum:
            - Processing
            - Available
            - Deleting
            - ProcessingFailed
          type: string
        percent_done:
          nullable: true
          description: The percentage of the file that has been processed
          type: number
          format: double
        signed_url:
          nullable: true
          example: https://storage.googleapis.com/bucket/file.pdf?...
          description: >-
            A [signed
            URL](https://cloud.google.com/storage/docs/access-control/signed-urls)
            that provides temporary, read-only access to the underlying file.
            Anyone with the link can access the file, so treat it as sensitive
            data. Expires after a short time.
          type: string
        error_message:
          nullable: true
          description: >-
            A message describing any error during file processing. Provided only
            if an error occurs.
          type: string
        multimodal:
          description: Indicates whether the file was processed as multimodal.
          type: boolean
      required:
        - id
        - name
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: Uploaded file can only currently be either a pdf or txt file
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: Pinecone API Key

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/list_imports.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List imports

> List all recent and ongoing import operations.

By default, `list_imports` returns up to 100 imports per page. If the `limit` parameter is set, `list` returns up to that number of imports instead. Whenever there are additional IDs to return, the response also includes a `pagination_token` that you can use to get the next batch of imports. When the response does not include a `pagination_token`, there are no more imports to return.

For guidance and examples, see [Import data](https://docs.pinecone.io/guides/index-data/import-data).

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Standard and Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

<RequestExample>
  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X GET "https://$INDEX_HOST/bulk/imports?paginationToken==Tm90aGluZyB0byBzZWUgaGVyZQo" \
    -H 'Api-Key: $PINECONE_API_KEY' \
    -H 'X-Pinecone-Api-Version: 2025-10'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "data": [
      {
        "id": "1",
        "uri": "s3://BUCKET_NAME/PATH/TO/DIR",
        "status": "Pending",
        "started_at": "2024-08-19T20:49:00.754Z",
        "finished_at": "2024-08-19T20:49:00.754Z",
        "percent_complete": 42.2,
        "records_imported": 1000000
      }
    ],
    "pagination": {
      "next": "Tm90aGluZyB0byBzZWUgaGVyZQo="
    }
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml get /bulk/imports
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /bulk/imports:
    get:
      tags:
        - Bulk Operations
      summary: List imports
      description: >-
        List all recent and ongoing import operations.


        By default, `list_imports` returns up to 100 imports per page. If the
        `limit` parameter is set, `list` returns up to that number of imports
        instead. Whenever there are additional IDs to return, the response also
        includes a `pagination_token` that you can use to get the next batch of
        imports. When the response does not include a `pagination_token`, there
        are no more imports to return.


        For guidance and examples, see [Import
        data](https://docs.pinecone.io/guides/index-data/import-data).
      operationId: listBulkImports
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: query
          name: limit
          description: Max number of operations to return per page.
          schema:
            default: 100
            type: integer
            format: int32
            minimum: 1
            maximum: 100
          example: 10
          style: form
        - in: query
          name: paginationToken
          description: Pagination token to continue a previous listing operation.
          schema:
            type: string
          style: form
      responses:
        '200':
          description: A list of import operations
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ListImportsResponse'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    ListImportsResponse:
      description: The response for the `list_imports` operation.
      type: object
      properties:
        data:
          type: array
          items:
            $ref: '#/components/schemas/ImportModel'
        pagination:
          $ref: '#/components/schemas/Pagination'
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    ImportModel:
      description: The model for an import operation.
      type: object
      properties:
        id:
          example: '101'
          description: Unique identifier for the import operation.
          type: string
          minLength: 1
          maxLength: 1000
        uri:
          description: The URI from where the data is imported.
          type: string
        status:
          example: Pending
          description: >-
            The status of the operation.

            Possible values: `Pending`, `InProgress`, `Failed`, `Completed`, or
            `Cancelled`.
          x-enum:
            - Pending
            - InProgress
            - Failed
            - Completed
            - Cancelled
          type: string
        createdAt:
          description: The start time of the import operation.
          type: string
          format: date-time
        finishedAt:
          description: The end time of the import operation.
          type: string
          format: date-time
        percentComplete:
          example: 42.2
          description: The progress made by the operation, as a percentage.
          type: number
          format: float
          minimum: 0
          maximum: 100
        recordsImported:
          example: 1000000
          description: The number of records successfully imported.
          type: integer
          format: int64
        error:
          description: The error message if the import process failed.
          type: string
    Pagination:
      type: object
      properties:
        next:
          example: Tm90aGluZyB0byBzZWUgaGVyZQo=
          type: string
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/list_index_backups.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List backups for an index

> List all backups for an index.

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_NAME="docs-example"

  curl -X GET "https://api.pinecone.io/indexes/$INDEX_NAME/backups" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -H "accept: application/json"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "data":
    [
      {
        "backup_id":"9947520e-d5a1-4418-a78d-9f464c9969da",
        "source_index_id":"8433941a-dae7-43b5-ac2c-d3dab4a56b2b",
        "source_index_name":"docs-example",
        "tags":{},
        "name":"example-backup",
        "description":"Monthly backup of production index",
        "status":"Pending",
        "cloud":"aws",
        "region":"us-east-1",
        "dimension":1024,
        "record_count":98,
        "namespace_count":3,
        "size_bytes":1069169,
        "created_at":"2025-03-11T18:29:50.549505Z"
        }
      ],
    "pagination":null
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml get /indexes/{index_name}/backups
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /indexes/{index_name}/backups:
    get:
      tags:
        - Manage Indexes
      summary: List backups for an index
      description: List all backups for an index.
      operationId: list_index_backups
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: index_name
          description: Name of the backed up index
          required: true
          schema:
            type: string
          style: simple
        - in: query
          name: limit
          description: The number of results to return per page.
          schema:
            default: 10
            type: integer
            minimum: 1
            maximum: 100
          style: form
        - in: query
          name: paginationToken
          description: The token to use to retrieve the next page of results.
          schema:
            type: string
          style: form
      responses:
        '200':
          description: >-
            This operation returns a list of all the backups that you have
            previously created, and which are associated with the given index.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/BackupList'
              examples:
                backups:
                  summary: A list containing backups.
                  value:
                    data:
                      - backup_id: bkp_123abc
                        cloud: aws
                        created_at: '2024-03-15T10:30:00.000Z'
                        description: Monthly backup of production index
                        dimension: 1536
                        metric: cosine
                        name: backup_2024_03_15
                        namespace_count: 3
                        record_count: 120000
                        region: us-east-1
                        schema:
                          fields:
                            genre:
                              filterable: true
                            title:
                              filterable: true
                        size_bytes: 10000000
                        source_index_id: idx_456
                        source_index_name: my-index
                        status: Ready
                        tags:
                          environment: production
                          type: monthly
                      - backup_id: bkp_789xyz
                        cloud: aws
                        created_at: '2024-03-20T15:45:00.000Z'
                        description: Pre-deployment safety backup
                        dimension: 1536
                        metric: cosine
                        name: backup_2024_03_20
                        namespace_count: 4
                        record_count: 125000
                        region: us-east-1
                        schema:
                          fields:
                            genre:
                              filterable: true
                            title:
                              filterable: true
                        size_bytes: 10500000
                        source_index_id: idx_456
                        source_index_name: my-index
                        status: Ready
                        tags:
                          environment: production
                          type: pre-deploy
                    pagination:
                      next: dXNlcl9pZD11c2VyXzE=
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Index not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-not-found:
                  summary: Index not found
                  value:
                    error:
                      code: NOT_FOUND
                      message: Index example-index not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    BackupList:
      description: The list of backups that exist in the project.
      type: object
      properties:
        data:
          description: List of backup objects
          type: array
          items:
            $ref: '#/components/schemas/BackupModel'
        pagination:
          $ref: '#/components/schemas/PaginationResponse'
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    BackupModel:
      description: >-
        The BackupModel describes the configuration and status of a Pinecone
        backup.
      type: object
      properties:
        backup_id:
          example: 670e8400-e29b-41d4-a716-446655440001
          description: Unique identifier for the backup.
          type: string
        source_index_name:
          example: my-index
          description: Name of the index from which the backup was taken.
          type: string
        source_index_id:
          example: 670e8400-e29b-41d4-a716-446655440000
          description: ID of the index.
          type: string
        name:
          example: backup-2025-02-04
          description: Optional user-defined name for the backup.
          type: string
        description:
          example: Backup before bulk update.
          description: Optional description providing context for the backup.
          type: string
        status:
          example: Ready
          description: Current status of the backup (e.g., Initializing, Ready, Failed).
          type: string
        cloud:
          example: aws
          description: Cloud provider where the backup is stored.
          type: string
        region:
          example: us-east-1
          description: Cloud region where the backup is stored.
          type: string
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If the 'vector_type' is
            'sparse', the metric must be 'dotproduct'. If the `vector_type` is
            `dense`, the metric defaults to 'cosine'.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        schema:
          $ref: '#/components/schemas/MetadataSchema'
        record_count:
          example: 120000
          description: Total number of records in the backup.
          type: integer
        namespace_count:
          example: 3
          description: Number of namespaces in the backup.
          type: integer
        size_bytes:
          example: 10000000
          description: Size of the backup in bytes.
          type: integer
        tags:
          $ref: '#/components/schemas/IndexTags'
        created_at:
          description: Timestamp when the backup was created.
          type: string
      required:
        - backup_id
        - source_index_name
        - source_index_id
        - status
        - cloud
        - region
    PaginationResponse:
      example:
        next: dXNlcl9pZD11c2VyXzE=
      description: The pagination object that is returned with paginated responses.
      type: object
      properties:
        next:
          example: dXNlcl9pZD11c2VyXzE=
          description: The token to use to retrieve the next page of results.
          type: string
      required:
        - next
    MetadataSchema:
      example:
        fields:
          description:
            filterable: true
          genre:
            filterable: true
          year:
            filterable: true
      description: >-
        Schema for the behavior of Pinecone's internal metadata index. By
        default, all metadata is indexed; when `schema` is present, only fields
        which are present in the `fields` object with a `filterable: true` are
        indexed. Note that `filterable: false` is not currently supported.
      type: object
      properties:
        fields:
          description: >-
            A map of metadata field names to their configuration. The field name
            must be a valid metadata field name. The field name must be unique.
          type: object
          additionalProperties:
            type: object
            properties:
              filterable:
                description: >-
                  Whether the field is filterable. If true, the field is indexed
                  and can be used in filters. Only true values are allowed.
                type: boolean
      required:
        - fields
    IndexTags:
      example:
        tag0: val0
        tag1: val1
      description: >-
        Custom user tags added to an index. Keys must be 80 characters or less.
        Values must be 120 characters or less. Keys must be alphanumeric, '_',
        or '-'.  Values must be alphanumeric, ';', '@', '_', '-', '.', '+', or '
        '. To unset a key, set the value to be an empty string.
      type: object
      additionalProperties:
        type: string
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/list_indexes.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List indexes

> List all indexes in a project.

<RequestExample>
  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X GET "https://api.pinecone.io/indexes" \
  -H "Api-Key: $PINECONE_API_KEY" \
  -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "indexes": [
      {
        "name": "example-serverless-dedicated-index",
        "vector_type": "dense",
        "metric": "cosine",
        "dimension": 1536,
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "host": "example-serverless-dedicated-index-bhnyigt.svc.aped-4627-b74a.pinecone.io",
        "spec": {
          "serverless": {
            "region": "us-east-1",
            "cloud": "aws",
            "read_capacity": {
              "mode": "Dedicated",
              "dedicated": {
                "node_type": "b1",
                "scaling": "Manual",
                "manual": {
                  "shards": 1,
                  "replicas": 2
                }
              },
              "status": {
                "state": "Scaling",
                "current_shards": 1,
                "current_replicas": 1
              }
            }
          }
        },
        "deletion_protection": "enabled",
        "tags": {
          "tag0": "value0",
          "tag1": "value1"
        }
      },
      {
        "name": "example-serverless-ondemand-index",
        "vector_type": "dense",
        "metric": "cosine",
        "dimension": 1024,
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "host": "example-serverless-ondemand-index-bhnyigt.svc.aped-4627-b74a.pinecone.io",
        "spec": {
          "serverless": {
            "region": "us-east-1",
            "cloud": "aws",
            "read_capacity": {
              "mode": "OnDemand",
              "status": {
                "state": "Ready",
                "current_shards": null,
                "current_replicas": null
              }
            }
          }
        },
        "deletion_protection": "enabled",
        "tags": {
          "tag1": "value1",
          "tag2": "value2"
        },
        "embed": {
          "model": "llama-text-embed-v2",
          "field_map": {
            "text": "text"
          },
          "dimension": 1024,
          "metric": "cosine",
          "write_parameters": {
            "dimension": 1024,
            "input_type": "passage",
            "truncate": "END"
          },
          "read_parameters": {
            "dimension": 1024,
            "input_type": "query",
            "truncate": "END"
          },
          "vector_type": "dense"
        }
      },
      {
        "name": "example-pod-index",
        "vector_type": "dense",
        "metric": "cosine",
        "dimension": 768,
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "host": "example-pod-index-bhnyigt.svc.us-east-1-aws.pinecone.io",
        "spec": {
          "pod": {
            "replicas": 1,
            "shards": 1,
            "pods": 1,
            "pod_type": "s1.x1",
            "environment": "us-east-1-aws"
          }
        },
        "deletion_protection": "disabled",
        "tags": null
      }
    ]
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml get /indexes
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /indexes:
    get:
      tags:
        - Manage Indexes
      summary: List indexes
      description: List all indexes in a project.
      operationId: list_indexes
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      responses:
        '200':
          description: >-
            This operation returns a list of all the indexes that you have
            previously created, and which are associated with the given project
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/IndexList'
              examples:
                multiple-indexes:
                  summary: >-
                    A list containing one serverless index and one pod-based
                    index.
                  value:
                    indexes:
                      - dimension: 384
                        host: semantic-search-c01b5b5.svc.us-west1-gcp.pinecone.io
                        metric: cosine
                        name: semantic-search
                        spec:
                          pod:
                            environment: us-west1-gcp
                            pod_type: p1.x1
                            pods: 4
                            replicas: 2
                            shards: 2
                        status:
                          ready: true
                          state: Ready
                        vector_type: dense
                      - dimension: 200
                        host: image-search-a31f9c1.svc.us-east1-gcp.pinecone.io
                        metric: dotproduct
                        name: image-search
                        spec:
                          serverless:
                            cloud: aws
                            read_capacity:
                              mode: OnDemand
                              status:
                                state: Ready
                            region: us-east-1
                        status:
                          ready: false
                          state: Initializing
                        vector_type: dense
                      - host: sparse-index-1a2b3c4d.svc.us-east1-gcp.pinecone.io
                        metric: dotproduct
                        name: sparse-index
                        spec:
                          serverless:
                            cloud: aws
                            read_capacity:
                              mode: OnDemand
                              status:
                                state: Ready
                            region: us-east-1
                        status:
                          ready: true
                          state: Ready
                        vector_type: sparse
                one-index:
                  summary: A list containing one serverless index.
                  value:
                    indexes:
                      - dimension: 1536
                        host: movie-embeddings-c01b5b5.svc.us-east1-gcp.pinecone.io
                        metric: cosine
                        name: movie-embeddings
                        spec:
                          serverless:
                            cloud: aws
                            read_capacity:
                              mode: OnDemand
                              status:
                                state: Ready
                            region: us-east-1
                        status:
                          ready: false
                          state: Initializing
                        vector_type: dense
                no-indexes:
                  summary: No indexes created yet.
                  value:
                    indexes: []
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    IndexList:
      description: The list of indexes that exist in the project.
      type: object
      properties:
        indexes:
          description: List of indexes in the project
          type: array
          items:
            $ref: '#/components/schemas/IndexModel'
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    IndexModel:
      description: >-
        The IndexModel describes the configuration and status of a Pinecone
        index.
      type: object
      properties:
        name:
          example: example-index
          description: >
            The name of the index. Resource name must be 1-45 characters long,
            start and end with an alphanumeric character, and consist only of
            lower case alphanumeric characters or '-'.
          type: string
          minLength: 1
          maxLength: 45
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If the 'vector_type' is
            'sparse', the metric must be 'dotproduct'. If the `vector_type` is
            `dense`, the metric defaults to 'cosine'.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        host:
          example: semantic-search-c01b5b5.svc.us-west1-gcp.pinecone.io
          description: The URL address where the index is hosted.
          type: string
        private_host:
          example: semantic-search-c01b5b5.svc.private.us-west1-gcp.pinecone.io
          description: The private endpoint URL of an index.
          type: string
        deletion_protection:
          $ref: '#/components/schemas/DeletionProtection'
        tags:
          $ref: '#/components/schemas/IndexTags'
        embed:
          $ref: '#/components/schemas/ModelIndexEmbed'
        spec:
          example:
            pod:
              environment: us-east-1-aws
              metadata_config:
                indexed:
                  - genre
                  - title
                  - imdb_rating
              pod_type: p1.x1
              pods: 1
              replicas: 1
              shards: 1
          description: The spec object defines how the index should be deployed.
          type: object
          oneOf:
            - title: Serverless
              type: object
              properties:
                serverless:
                  $ref: '#/components/schemas/ServerlessSpecResponse'
              required:
                - serverless
              additionalProperties: false
            - title: Pod-based
              type: object
              properties:
                pod:
                  $ref: '#/components/schemas/PodSpec'
              required:
                - pod
              additionalProperties: false
            - title: BYOC
              type: object
              properties:
                byoc:
                  $ref: '#/components/schemas/ByocSpecResponse'
              required:
                - byoc
              additionalProperties: false
        status:
          example:
            ready: true
            state: ScalingUpPodSize
          description: The current status of the index
          type: object
          properties:
            ready:
              description: Whether the index is ready for use
              type: boolean
            state:
              description: >-
                The state of the index.

                Possible values: `Initializing`, `InitializationFailed`,
                `ScalingUp`, `ScalingDown`, `ScalingUpPodSize`,
                `ScalingDownPodSize`, `Terminating`, `Ready`, or `Disabled`.
              x-enum:
                - Initializing
                - InitializationFailed
                - ScalingUp
                - ScalingDown
                - ScalingUpPodSize
                - ScalingDownPodSize
                - Terminating
                - Ready
                - Disabled
              type: string
          required:
            - ready
            - state
        vector_type:
          description: >-
            The index vector type. You can use 'dense' or 'sparse'. If 'dense',
            the vector dimension must be specified.  If 'sparse', the vector
            dimension should not be specified.
          default: dense
          type: string
      required:
        - name
        - metric
        - status
        - spec
        - host
        - vector_type
    DeletionProtection:
      description: >-
        Whether [deletion
        protection](http://docs.pinecone.io/guides/manage-data/manage-indexes#configure-deletion-protection)
        is enabled/disabled for the index.

        Possible values: `disabled` or `enabled`.
      default: disabled
      x-enum:
        - disabled
        - enabled
      type: string
    IndexTags:
      example:
        tag0: val0
        tag1: val1
      description: >-
        Custom user tags added to an index. Keys must be 80 characters or less.
        Values must be 120 characters or less. Keys must be alphanumeric, '_',
        or '-'.  Values must be alphanumeric, ';', '@', '_', '-', '.', '+', or '
        '. To unset a key, set the value to be an empty string.
      type: object
      additionalProperties:
        type: string
    ModelIndexEmbed:
      example:
        field_map:
          text: your-text-field
        metric: cosine
        model: multilingual-e5-large
        read_parameters:
          input_type: query
          truncate: NONE
        write_parameters:
          input_type: passage
      description: The embedding model and document fields mapped to embedding inputs.
      type: object
      properties:
        model:
          example: multilingual-e5-large
          description: The name of the embedding model used to create the index.
          type: string
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If not specified, the metric
            will be defaulted according to the model. Cannot be updated once
            set.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        vector_type:
          description: >-
            The index vector type. You can use 'dense' or 'sparse'. If 'dense',
            the vector dimension must be specified.  If 'sparse', the vector
            dimension should not be specified.
          default: dense
          type: string
        field_map:
          example:
            text: your-text-field
          description: >-
            Identifies the name of the text field from your document model that
            is embedded.
          type: object
        read_parameters:
          description: The read parameters for the embedding model.
          type: object
        write_parameters:
          description: The write parameters for the embedding model.
          type: object
      required:
        - model
    ServerlessSpecResponse:
      description: Configuration of a serverless index.
      type: object
      properties:
        cloud:
          example: aws
          description: |-
            The public cloud where you would like your index hosted.
            Possible values: `gcp`, `aws`, or `azure`.
          x-enum:
            - gcp
            - aws
            - azure
          type: string
        region:
          example: us-east-1
          description: The region where you would like your index to be created.
          type: string
        read_capacity:
          $ref: '#/components/schemas/ReadCapacityResponse'
        source_collection:
          example: movie-embeddings
          description: The name of the collection to be used as the source for the index.
          type: string
        schema:
          $ref: '#/components/schemas/MetadataSchema'
      required:
        - cloud
        - region
        - read_capacity
    PodSpec:
      example:
        environment: us-east1-gcp
        metadata_config:
          indexed:
            - genre
            - title
            - imdb_rating
        pod_type: p1.x1
        pods: 1
        replicas: 1
        shards: 1
        source_collection: movie-embeddings
      title: Pod-based
      description: Configuration needed to deploy a pod-based index.
      type: object
      properties:
        environment:
          example: us-east1-gcp
          description: The environment where the index is hosted.
          type: string
        replicas:
          description: >-
            The number of replicas. Replicas duplicate your index. They provide
            higher availability and throughput. Replicas can be scaled up or
            down as your needs change.
          default: 1
          type: integer
          format: int32
          minimum: 1
        shards:
          description: >-
            The number of shards. Shards split your data across multiple pods so
            you can fit more data into an index.
          default: 1
          type: integer
          format: int32
          minimum: 1
        pod_type:
          description: >-
            The type of pod to use. One of `s1`, `p1`, or `p2` appended with `.`
            and one of `x1`, `x2`, `x4`, or `x8`.
          default: p1.x1
          type: string
        pods:
          example: 1
          description: >-
            The number of pods to be used in the index. This should be equal to
            `shards` x `replicas`.'
          default: 1
          type: integer
          minimum: 1
        metadata_config:
          example:
            indexed:
              - genre
              - title
              - imdb_rating
          description: >-
            Configuration for the behavior of Pinecone's internal metadata
            index. By default, all metadata is indexed; when `metadata_config`
            is present, only specified metadata fields are indexed. These
            configurations are only valid for use with pod-based indexes.
          type: object
          properties:
            indexed:
              description: >-
                By default, all metadata is indexed; to change this behavior,
                use this property to specify an array of metadata fields that
                should be indexed.
              type: array
              items:
                type: string
        source_collection:
          example: movie-embeddings
          description: The name of the collection to be used as the source for the index.
          type: string
      required:
        - environment
        - pod_type
    ByocSpecResponse:
      title: BYOC
      description: Configuration of a BYOC index.
      type: object
      properties:
        environment:
          example: aws-us-east-1-b921
          description: The environment where the index is hosted.
          type: string
        read_capacity:
          $ref: '#/components/schemas/ReadCapacityResponse'
        schema:
          $ref: '#/components/schemas/MetadataSchema'
      required:
        - environment
        - read_capacity
    ReadCapacityResponse:
      description: Response containing read capacity configuration
      discriminator:
        propertyName: mode
        mapping:
          OnDemand: '#/components/schemas/ReadCapacityOnDemandSpecResponse'
          Dedicated: '#/components/schemas/ReadCapacityDedicatedSpecResponse'
      oneOf:
        - $ref: '#/components/schemas/ReadCapacityOnDemandSpecResponse'
        - $ref: '#/components/schemas/ReadCapacityDedicatedSpecResponse'
    MetadataSchema:
      example:
        fields:
          description:
            filterable: true
          genre:
            filterable: true
          year:
            filterable: true
      description: >-
        Schema for the behavior of Pinecone's internal metadata index. By
        default, all metadata is indexed; when `schema` is present, only fields
        which are present in the `fields` object with a `filterable: true` are
        indexed. Note that `filterable: false` is not currently supported.
      type: object
      properties:
        fields:
          description: >-
            A map of metadata field names to their configuration. The field name
            must be a valid metadata field name. The field name must be unique.
          type: object
          additionalProperties:
            type: object
            properties:
              filterable:
                description: >-
                  Whether the field is filterable. If true, the field is indexed
                  and can be used in filters. Only true values are allowed.
                type: boolean
      required:
        - fields
    ReadCapacityOnDemandSpecResponse:
      example:
        mode: OnDemand
        status:
          state: Ready
      title: On-demand
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
        status:
          $ref: '#/components/schemas/ReadCapacityStatus'
      required:
        - mode
        - status
      additionalProperties: false
    ReadCapacityDedicatedSpecResponse:
      example:
        dedicated:
          manual:
            replicas: 2
            shards: 2
          node_type: t1
          scaling: Manual
        mode: Dedicated
        status:
          current_replicas: 2
          current_shards: 2
          state: Ready
      title: Dedicated
      type: object
      properties:
        mode:
          description: >-
            The mode of the index. Possible values: `OnDemand` or `Dedicated`.
            Defaults to `OnDemand`. If set to `Dedicated`,
            `dedicated.node_type`, and `dedicated.scaling` must be specified.
          type: string
        dedicated:
          $ref: '#/components/schemas/ReadCapacityDedicatedConfig'
        status:
          $ref: '#/components/schemas/ReadCapacityStatus'
      required:
        - mode
        - dedicated
        - status
      additionalProperties: false
    ReadCapacityStatus:
      description: >-
        The current status of factors affecting the read capacity of a
        serverless index
      type: object
      properties:
        state:
          description: >-
            The `state` describes the overall status of factors relating to the
            read capacity of an index. 


            Available values:

            - `Ready` is the state most of the time

            - `Scaling` if the number of replicas or shards has been recently
            updated by calling the [configure index
            endpoint](https://docs.pinecone.io/reference/api/2025-10/control-plane/configure_index)

            - `Migrating` if the index is being migrated to a new `node_type`

            - `Error` if there is an error with the read capacity configuration.
            In that case, see `error_message` for more details.
          type: string
        current_replicas:
          description: >-
            The number of replicas. Each replica has dedicated  compute
            resources and data storage. Increasing this number  will increase
            the total throughput of the index.
          type: integer
          format: int32
        current_shards:
          description: >-
            The number of shards. Each shard has dedicated storage.  Increasing
            shards alleiviates index fullness. 
          type: integer
          format: int32
        error_message:
          description: >-
            An optional error message indicating any issues with your read
            capacity configuration
          type: string
      required:
        - state
    ReadCapacityDedicatedConfig:
      description: >-
        Configuration for dedicated read capacity. See  [this
        guide](https://docs.pinecone.io/guides/index-data/dedicated-read-nodes)
        for more details on  how to configure dedicated read capacity.
      type: object
      properties:
        node_type:
          description: >-
            The type of machines to use. Available options: `b1` and `t1`. `t1`
            includes increased processing power and memory.
          type: string
        scaling:
          description: The type of scaling strategy to use.
          type: string
        manual:
          $ref: '#/components/schemas/ScalingConfigManual'
    ScalingConfigManual:
      description: The config to use for manual read capacity scaling.
      type: object
      properties:
        replicas:
          description: >-
            The number of replicas to use. Replicas duplicate the compute
            resources and data of an index, allowing higher query throughput and
            availability. Setting replicas to 0 disables the index but can be
            used to reduce costs while usage is paused.
          type: integer
          format: int32
          minimum: 0
        shards:
          description: >-
            The number of shards to use. Shards determine the storage capacity
            of an index, with each shard providing 250 GB of storage.
          type: integer
          format: int32
          minimum: 1
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/inference/list_models.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List available models

> List the embedding and reranking models hosted by Pinecone. 

You can use hosted models as an integrated part of Pinecone operations or for standalone embedding and reranking. For more details, see [Vector embedding](https://docs.pinecone.io/guides/index-data/indexing-overview#vector-embedding) and [Rerank results](https://docs.pinecone.io/guides/search/rerank-results).

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://api.pinecone.io/models" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "models": [
      {
        "model": "llama-text-embed-v2",
        "short_description": "A high performance dense embedding model optimized for multilingual and cross-lingual text question-answering retrieval with support for long documents (up to 2048 tokens) and dynamic embedding size (Matryoshka Embeddings).",
        "type": "embed",
        "vector_type": "dense",
        "default_dimension": 1024,
        "modality": "text",
        "max_sequence_length": 2048,
        "max_batch_size": 96,
        "provider_name": "NVIDIA",
        "supported_metrics": [
          "Cosine",
          "DotProduct"
        ],
        "supported_dimensions": [
          384,
          512,
          768,
          1024,
          2048
        ],
        "supported_parameters": [
          {
            "parameter": "input_type",
            "required": true,
            "type": "one_of",
            "value_type": "string",
            "allowed_values": [
              "query",
              "passage"
            ]
          },
          {
            "parameter": "truncate",
            "required": false,
            "default": "END",
            "type": "one_of",
            "value_type": "string",
            "allowed_values": [
              "END",
              "NONE",
              "START"
            ]
          },
          {
            "parameter": "dimension",
            "required": false,
            "default": 1024,
            "type": "one_of",
            "value_type": "integer",
            "allowed_values": [
              384,
              512,
              768,
              1024,
              2048
            ]
          }
        ]
      },
      {
        "model": "multilingual-e5-large",
        "short_description": "A high-performance dense embedding model trained on a mixture of multilingual datasets. It works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)",
        "type": "embed",
        "vector_type": "dense",
        "default_dimension": 1024,
        "modality": "text",
        "max_sequence_length": 507,
        "max_batch_size": 96,
        "provider_name": "Microsoft",
        "supported_metrics": [
          "Cosine",
          "Euclidean"
        ],
        "supported_dimensions": [
          1024
        ],
        "supported_parameters": [
          {
            "parameter": "input_type",
            "required": true,
            "type": "one_of",
            "value_type": "string",
            "allowed_values": [
              "query",
              "passage"
            ]
          },
          {
            "parameter": "truncate",
            "required": false,
            "default": "END",
            "type": "one_of",
            "value_type": "string",
            "allowed_values": [
              "END",
              "NONE"
            ]
          }
        ]
      },
      {
        "model": "pinecone-sparse-english-v0",
        "short_description": "A sparse embedding model for converting text to sparse vectors for keyword or hybrid semantic/keyword search. Built on the innovations of the DeepImpact architecture.",
        "type": "embed",
        "vector_type": "sparse",
        "modality": "text",
        "max_sequence_length": 512,
        "max_batch_size": 96,
        "provider_name": "Pinecone",
        "supported_metrics": [
          "DotProduct"
        ],
        "supported_parameters": [
          {
            "parameter": "input_type",
            "required": true,
            "type": "one_of",
            "value_type": "string",
            "allowed_values": [
              "query",
              "passage"
            ]
          },
          {
            "parameter": "truncate",
            "required": false,
            "default": "END",
            "type": "one_of",
            "value_type": "string",
            "allowed_values": [
              "END",
              "NONE"
            ]
          },
          {
            "parameter": "return_tokens",
            "required": false,
            "default": false,
            "type": "any",
            "value_type": "boolean"
          }
        ]
      },
      {
        "model": "bge-reranker-v2-m3",
        "short_description": "A high-performance, multilingual reranking model that works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)",
        "type": "rerank",
        "modality": "text",
        "max_sequence_length": 1024,
        "max_batch_size": 100,
        "provider_name": "BAAI",
        "supported_parameters": [
          {
            "parameter": "truncate",
            "required": false,
            "default": "NONE",
            "type": "one_of",
            "value_type": "string",
            "allowed_values": [
              "END",
              "NONE"
            ]
          }
        ]
      },
      {
        "model": "cohere-rerank-3.5",
        "short_description": "Cohere's leading reranking model, balancing performance and latency for a wide range of enterprise search applications.",
        "type": "rerank",
        "modality": "text",
        "max_sequence_length": 40000,
        "max_batch_size": 200,
        "provider_name": "Cohere",
        "supported_parameters": [
          {
            "parameter": "max_chunks_per_doc",
            "required": false,
            "default": 3072,
            "type": "numeric_range",
            "value_type": "integer",
            "min": 1,
            "max": 3072
          }
        ]
      },
      {
        "model": "pinecone-rerank-v0",
        "short_description": "A state of the art reranking model that out-performs competitors on widely accepted benchmarks. It can handle chunks up to 512 tokens (1-2 paragraphs)",
        "type": "rerank",
        "modality": "text",
        "max_sequence_length": 512,
        "max_batch_size": 100,
        "provider_name": "Pinecone",
        "supported_parameters": [
          {
            "parameter": "truncate",
            "required": false,
            "default": "END",
            "type": "one_of",
            "value_type": "string",
            "allowed_values": [
              "END",
              "NONE"
            ]
          }
        ]
      }
    ]
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/inference_2025-10.oas.yaml get /models
openapi: 3.0.3
info:
  title: Pinecone Inference API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Inference
    description: Model inference
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /models:
    get:
      tags:
        - Inference
      summary: List available models
      description: >-
        List the embedding and reranking models hosted by Pinecone. 


        You can use hosted models as an integrated part of Pinecone operations
        or for standalone embedding and reranking. For more details, see [Vector
        embedding](https://docs.pinecone.io/guides/index-data/indexing-overview#vector-embedding)
        and [Rerank
        results](https://docs.pinecone.io/guides/search/rerank-results).
      operationId: list_models
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: query
          name: type
          description: Filter models by type ('embed' or 'rerank').
          schema:
            description: |-
              Filter models by type.
              Possible values: `embed` or `rerank`.
            x-enum:
              - embed
              - rerank
            type: string
          example: embed
          style: form
        - in: query
          name: vector_type
          description: >-
            Filter embedding models by vector type ('dense' or 'sparse'). Only
            relevant when `type=embed`.
          schema:
            description: |-
              Filter models by vector type.
              Possible values: `dense` or `sparse`.
            x-enum:
              - dense
              - sparse
            type: string
          example: sparse
          style: form
      responses:
        '200':
          description: The list of available models.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ModelInfoList'
              examples:
                multiple-models:
                  summary: Multiple available models.
                  value:
                    models:
                      - default_dimension: 256
                        max_batch_size: 96
                        max_sequence_length: 512
                        modality: text
                        model: example-embedding-model
                        provider_name: Embedding Model Provider
                        short_description: An example embedding model.
                        supported_dimensions:
                          - 256
                          - 512
                        supported_metrics:
                          - cosine
                          - euclidean
                        supported_parameters:
                          - allowed_values:
                              - value1
                              - value2
                            parameter: example_required_param
                            required: true
                            type: one_of
                            value_type: string
                          - allowed_values:
                              - value1
                              - value2
                            default: value1
                            parameter: example_param_with_default
                            required: false
                            type: one_of
                            value_type: string
                          - default: 5
                            max: 10
                            min: 0
                            parameter: example_numeric_range
                            required: false
                            type: numeric_range
                            value_type: integer
                        type: embed
                        vector_type: dense
                      - max_batch_size: 100
                        max_sequence_length: 1024
                        modality: text
                        model: example-reranking-model
                        provider_name: Reranking Model Provider
                        short_description: An example reranking model.
                        supported_parameters:
                          - default: true
                            parameter: example_any_value
                            required: false
                            type: any
                            value_type: boolean
                        type: rerank
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Model not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                model-not-found:
                  summary: Model not found
                  value:
                    error:
                      code: NOT_FOUND
                      message: Model example-model not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    ModelInfoList:
      description: The list of available models.
      type: object
      properties:
        models:
          description: List of available models.
          type: array
          items:
            $ref: '#/components/schemas/ModelInfo'
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, or `FORBIDDEN`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable error message describing the error.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    ModelInfo:
      description: >-
        Represents the model configuration including model type, supported
        parameters, and other model details.
      type: object
      properties:
        model:
          example: multilingual-e5-large
          description: The name of the model.
          type: string
        short_description:
          example: multilingual-e5-large
          description: A summary of the model.
          type: string
        type:
          example: embed
          description: The type of model (e.g. 'embed' or 'rerank').
          type: string
        vector_type:
          description: Whether the embedding model produces 'dense' or 'sparse' embeddings.
          type: string
        default_dimension:
          example: 1024
          description: >-
            The default embedding model dimension (applies to dense embedding
            models only).
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        modality:
          example: text
          description: The modality of the model (e.g. 'text').
          type: string
        max_sequence_length:
          example: 512
          description: The maximum tokens per sequence supported by the model.
          type: integer
          format: int32
          minimum: 1
        max_batch_size:
          example: 96
          description: The maximum batch size (number of sequences) supported by the model.
          type: integer
          format: int32
          minimum: 1
        provider_name:
          example: NVIDIA
          description: The name of the provider of the model.
          type: string
        supported_dimensions:
          description: >-
            The list of supported dimensions for the model (applies to dense
            embedding models only).
          type: array
          items:
            example: 1024
            type: integer
            format: int32
            minimum: 1
            maximum: 20000
        supported_metrics:
          $ref: '#/components/schemas/ModelInfoSupportedMetrics'
        supported_parameters:
          description: List of parameters supported by the model.
          type: array
          items:
            $ref: '#/components/schemas/ModelInfoSupportedParameter'
      required:
        - model
        - short_description
        - type
        - supported_parameters
    ModelInfoSupportedMetrics:
      description: The distance metrics supported by the model for similarity search.
      type: array
      items:
        $ref: '#/components/schemas/ModelInfoMetric'
    ModelInfoSupportedParameter:
      description: >-
        Describes a parameter supported by the model, including parameter value
        constraints.
      type: object
      properties:
        parameter:
          example: input_type
          description: The name of the parameter.
          type: string
        type:
          example: one_of
          description: >-
            The parameter type e.g. 'one_of', 'numeric_range', or 'any'.


            If the type is 'one_of', then 'allowed_values' will be set, and the
            value specified must be one of the allowed values. 'one_of' is only
            compatible with value_type 'string' or 'integer'.


            If 'numeric_range', then 'min' and 'max' will be set, then the value
            specified must adhere to the value_type and must fall within the
            `[min, max]` range (inclusive).


            If 'any' then any value is allowed, as long as it adheres to the
            value_type.
          type: string
        value_type:
          example: string
          description: >-
            The type of value the parameter accepts, e.g. 'string', 'integer',
            'float', or 'boolean'.
          type: string
        required:
          example: true
          description: Whether the parameter is required (true) or optional (false).
          type: boolean
        allowed_values:
          description: The allowed parameter values when the type is 'one_of'.
          type: array
          items:
            anyOf:
              - type: string
              - type: integer
        min:
          example: 1
          description: >-
            The minimum allowed value (inclusive) when the type is
            'numeric_range'.
          type: number
        max:
          example: 1
          description: >-
            The maximum allowed value (inclusive) when the type is
            'numeric_range'.
          type: number
        default:
          example: END
          description: The default value for the parameter when a parameter is optional.
          anyOf:
            - type: string
            - type: integer
              format: int32
            - type: number
              format: float
            - type: boolean
      required:
        - parameter
        - type
        - value_type
        - required
    ModelInfoMetric:
      description: >-
        A distance metric that the embedding model supports for similarity
        searches.

        Possible values: `cosine`, `euclidean`, or `dotproduct`.
      x-enum:
        - cosine
        - euclidean
        - dotproduct
      type: string
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/list_project_backups.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List backups for all indexes in a project

> List all backups for a project.

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -X GET "https://api.pinecone.io/backups" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -H "accept: application/json"
  ```
</RequestExample>

<ResponseExample>
  ```json curl  theme={null}
  {
    "data": [
      {
        "backup_id": "e12269b0-a29b-4af0-9729-c7771dec03e3",
        "source_index_id": "bcb5b3c9-903e-4cb6-8b37-a6072aeb874f",
        "source_index_name": "docs-example",
        "tags": null,
        "name": "example-backup",
        "description": null,
        "status": "Ready",
        "cloud": "aws",
        "region": "us-east-1",
        "dimension": 0,
        "record_count": 96,
        "namespace_count": 1,
        "size_bytes": 86393,
        "created_at": "2025-05-14T17:00:45.803146Z"
      },
      {
        "backup_id": "d686451d-1ede-4004-9f72-7d22cc799b6e",
        "source_index_id": "b49f27d1-1bf3-49c6-82b5-4ae46f00f0e6",
        "source_index_name": "docs-example2",
        "tags": null,
        "name": "example-backup2",
        "description": null,
        "status": "Ready",
        "cloud": "aws",
        "region": "us-east-1",
        "dimension": 1024,
        "record_count": 50,
        "namespace_count": 1,
        "size_bytes": 545171,
        "created_at": "2025-05-14T17:00:34.814371Z"
      },
      {
        "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
        "source_index_id": "f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
        "source_index_name": "docs-example3",
        "tags": {},
        "name": "example-backup3",
        "description": "Monthly backup of production index",
        "status": "Ready",
        "cloud": "aws",
        "region": "us-east-1",
        "dimension": 1024,
        "record_count": 98,
        "namespace_count": 3,
        "size_bytes": 1069169,
        "created_at": "2025-05-14T16:37:25.625540Z"
      }
    ],
    "pagination": null
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml get /backups
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /backups:
    get:
      tags:
        - Manage Indexes
      summary: List backups for all indexes in a project
      description: List all backups for a project.
      operationId: list_project_backups
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: query
          name: limit
          description: The number of results to return per page.
          schema:
            default: 10
            type: integer
            minimum: 1
            maximum: 100
          style: form
        - in: query
          name: paginationToken
          description: The token to use to retrieve the next page of results.
          schema:
            type: string
          style: form
      responses:
        '200':
          description: >-
            This operation returns a list of all the backups for the given index
            that you have previously created.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/BackupList'
              examples:
                backups:
                  summary: A list containing backups.
                  value:
                    data:
                      - backup_id: 670e8400-e29b-41d4-a716-446655440000
                        cloud: aws
                        created_at: '2024-03-15T10:30:00.000Z'
                        description: Monthly backup of production index
                        dimension: 1536
                        metric: cosine
                        name: backup_2024_03_15
                        namespace_count: 3
                        record_count: 120000
                        region: us-east-1
                        schema:
                          fields:
                            genre:
                              filterable: true
                            title:
                              filterable: true
                        size_bytes: 10000000
                        source_index_id: idx_456
                        source_index_name: my-index
                        status: Ready
                        tags:
                          environment: production
                          type: monthly
                      - backup_id: 670e8400-e29b-41d4-a716-446655440001
                        cloud: aws
                        created_at: '2024-03-20T15:45:00.000Z'
                        description: Pre-deployment safety backup
                        dimension: 1536
                        metric: cosine
                        name: backup_2024_03_20
                        namespace_count: 4
                        record_count: 125000
                        region: us-east-1
                        schema:
                          fields:
                            genre:
                              filterable: true
                            title:
                              filterable: true
                        size_bytes: 10500000
                        source_index_id: idx_789
                        source_index_name: my-index-2
                        status: Ready
                        tags:
                          environment: production
                          type: pre-deploy
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    BackupList:
      description: The list of backups that exist in the project.
      type: object
      properties:
        data:
          description: List of backup objects
          type: array
          items:
            $ref: '#/components/schemas/BackupModel'
        pagination:
          $ref: '#/components/schemas/PaginationResponse'
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    BackupModel:
      description: >-
        The BackupModel describes the configuration and status of a Pinecone
        backup.
      type: object
      properties:
        backup_id:
          example: 670e8400-e29b-41d4-a716-446655440001
          description: Unique identifier for the backup.
          type: string
        source_index_name:
          example: my-index
          description: Name of the index from which the backup was taken.
          type: string
        source_index_id:
          example: 670e8400-e29b-41d4-a716-446655440000
          description: ID of the index.
          type: string
        name:
          example: backup-2025-02-04
          description: Optional user-defined name for the backup.
          type: string
        description:
          example: Backup before bulk update.
          description: Optional description providing context for the backup.
          type: string
        status:
          example: Ready
          description: Current status of the backup (e.g., Initializing, Ready, Failed).
          type: string
        cloud:
          example: aws
          description: Cloud provider where the backup is stored.
          type: string
        region:
          example: us-east-1
          description: Cloud region where the backup is stored.
          type: string
        dimension:
          example: 1536
          description: The dimensions of the vectors to be inserted in the index.
          type: integer
          format: int32
          minimum: 1
          maximum: 20000
        metric:
          description: >-
            The distance metric to be used for similarity search. You can use
            'euclidean', 'cosine', or 'dotproduct'. If the 'vector_type' is
            'sparse', the metric must be 'dotproduct'. If the `vector_type` is
            `dense`, the metric defaults to 'cosine'.

            Possible values: `cosine`, `euclidean`, or `dotproduct`.
          x-enum:
            - cosine
            - euclidean
            - dotproduct
          type: string
        schema:
          $ref: '#/components/schemas/MetadataSchema'
        record_count:
          example: 120000
          description: Total number of records in the backup.
          type: integer
        namespace_count:
          example: 3
          description: Number of namespaces in the backup.
          type: integer
        size_bytes:
          example: 10000000
          description: Size of the backup in bytes.
          type: integer
        tags:
          $ref: '#/components/schemas/IndexTags'
        created_at:
          description: Timestamp when the backup was created.
          type: string
      required:
        - backup_id
        - source_index_name
        - source_index_id
        - status
        - cloud
        - region
    PaginationResponse:
      example:
        next: dXNlcl9pZD11c2VyXzE=
      description: The pagination object that is returned with paginated responses.
      type: object
      properties:
        next:
          example: dXNlcl9pZD11c2VyXzE=
          description: The token to use to retrieve the next page of results.
          type: string
      required:
        - next
    MetadataSchema:
      example:
        fields:
          description:
            filterable: true
          genre:
            filterable: true
          year:
            filterable: true
      description: >-
        Schema for the behavior of Pinecone's internal metadata index. By
        default, all metadata is indexed; when `schema` is present, only fields
        which are present in the `fields` object with a `filterable: true` are
        indexed. Note that `filterable: false` is not currently supported.
      type: object
      properties:
        fields:
          description: >-
            A map of metadata field names to their configuration. The field name
            must be a valid metadata field name. The field name must be unique.
          type: object
          additionalProperties:
            type: object
            properties:
              filterable:
                description: >-
                  Whether the field is filterable. If true, the field is indexed
                  and can be used in filters. Only true values are allowed.
                type: boolean
      required:
        - fields
    IndexTags:
      example:
        tag0: val0
        tag1: val1
      description: >-
        Custom user tags added to an index. Keys must be 80 characters or less.
        Values must be 120 characters or less. Keys must be alphanumeric, '_',
        or '-'.  Values must be alphanumeric, ';', '@', '_', '-', '.', '+', or '
        '. To unset a key, set the value to be an empty string.
      type: object
      additionalProperties:
        type: string
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/admin/list_projects.md

# Source: https://docs.pinecone.io/reference/api/2025-10/admin-assistant/list_projects.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List projects

> List all projects in an organization.

<RequestExample>
  ```bash curl theme={null}
  curl -X GET "https://api.pinecone.io/admin/projects" \
      -H "Authorization: Bearer $PINECONE_ACCESS_TOKEN" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "data": [
      {
        "id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
        "name": "example-project",
        "max_pods": 0,
        "force_encryption_with_cmek": true,
        "organization_id": "<string>",
        "created_at": "2023-11-07T05:31:56Z"
      }
    ]
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/admin_2025-10.oas.yaml get /admin/projects
openapi: 3.0.3
info:
  title: Pinecone Admin API
  description: |
    Provides an API for managing a Pinecone organization and its resources.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - BearerAuth: []
tags:
  - name: API Keys
    description: Actions that manage API Keys.
  - name: Organizations
    description: Actions that manage organizations.
  - name: Projects
    description: Actions that manage projects.
paths:
  /admin/projects:
    get:
      tags:
        - Projects
      summary: List projects
      description: List all projects in an organization.
      operationId: list_projects
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      responses:
        '200':
          description: A list of projects.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ProjectList'
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
        4XX:
          description: Unexpected error on request.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
components:
  schemas:
    ProjectList:
      type: object
      properties:
        data:
          type: array
          items:
            $ref: '#/components/schemas/Project'
      required:
        - data
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, or `UNPROCESSABLE_ENTITY`.        
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    Project:
      description: The details of a project.
      type: object
      properties:
        id:
          description: The unique ID of the project.
          type: string
          format: uuid
        name:
          description: The name of the project.
          type: string
          minLength: 1
          maxLength: 512
        max_pods:
          description: The maximum number of Pods that can be created in the project.
          type: integer
        force_encryption_with_cmek:
          description: >-
            Whether to force encryption with a customer-managed encryption key
            (CMEK).
          type: boolean
        organization_id:
          description: The unique ID of the organization that the project belongs to.
          type: string
        created_at:
          description: The date and time when the project was created.
          type: string
          format: date-time
      required:
        - id
        - name
        - max_pods
        - force_encryption_with_cmek
        - organization_id
  securitySchemes:
    BearerAuth:
      type: http
      scheme: bearer
      description: >
        An [access
        token](https://docs.pinecone.io/guides/organizations/manage-service-accounts#retrieve-an-access-token)
        must be provided in the `Authorization` header using the `Bearer`
        scheme.

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/control-plane/list_restore_jobs.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List restore jobs

> List all restore jobs for a project.

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://api.pinecone.io/restore-jobs" \
  	-H "X-Pinecone-Api-Version: 2025-10" \
  	-H "Api-Key: $PINECONE_API_KEY"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "data": [
      {
        "restore_job_id": "9857add2-99d4-4399-870e-aa7f15d8d326",
        "backup_id": "94a63aeb-efae-4f7a-b059-75d32c27ca57",
        "target_index_name": "restored-index",
        "target_index_id": "0d8aed24-adf8-4b77-8e10-fd674309dc85",
        "status": "Completed",
        "created_at": "2025-04-25T18:14:05.227526Z",
        "completed_at": "2025-04-25T18:14:11.074618Z",
        "percent_complete": 100
      },
      {
        "restore_job_id": "69acc1d0-9105-4fcb-b1db-ebf97b285c5e",
        "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
        "target_index_name": "restored-index2",
        "target_index_id": "e6c0387f-33db-4227-9e91-32181106e56b",
        "status": "Completed",
        "created_at": "2025-05-14T17:25:59.378989Z",
        "completed_at": "2025-05-14T17:26:23.997284Z",
        "percent_complete": 100
      }
    ],
    "pagination": null
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_control_2025-10.oas.yaml get /restore-jobs
openapi: 3.0.3
info:
  title: Pinecone Control Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Indexes
    description: Actions that manage indexes
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /restore-jobs:
    get:
      tags:
        - Manage Indexes
      summary: List restore jobs
      description: List all restore jobs for a project.
      operationId: list_restore_jobs
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: query
          name: limit
          description: The number of results to return per page.
          schema:
            default: 10
            type: integer
            minimum: 1
            maximum: 100
          style: form
        - in: query
          name: paginationToken
          description: The token to use to retrieve the next page of results.
          schema:
            type: string
          style: form
      responses:
        '200':
          description: >-
            This operation returns a list of all the restore jobs that you have
            previously created.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/RestoreJobList'
              examples:
                paginated-restore-jobs:
                  summary: A list containing restore jobs.
                  value:
                    data:
                      - backup_id: 670e8400-e29b-41d4-a716-446655440000
                        completed_at: '2024-03-15T10:35:00.000Z'
                        created_at: '2024-03-15T10:30:00.000Z'
                        percent_complete: 100
                        restore_job_id: 670e8400-e29b-41d4-a716-446655440001
                        status: Completed
                        target_index_id: idx_456
                        target_index_name: my-index
                    pagination:
                      next: dXNlcl9pZD11c2VyXzE=
                non-paginated-restore-jobs:
                  summary: A list containing restore jobs.
                  value:
                    data:
                      - backup_id: 670e8400-e29b-41d4-a716-446655440000
                        completed_at: '2024-03-15T10:35:00.000Z'
                        created_at: '2024-03-15T10:30:00.000Z'
                        percent_complete: 100
                        restore_job_id: 670e8400-e29b-41d4-a716-446655440001
                        status: Completed
                        target_index_id: idx_456
                        target_index_name: my-index
                    pagination: null
                no-restore-jobs:
                  summary: An empty list.
                  value:
                    data: []
                    pagination: null
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    RestoreJobList:
      description: The list of restore jobs that exist in the project.
      type: object
      properties:
        data:
          description: List of restore job objects
          type: array
          items:
            $ref: '#/components/schemas/RestoreJobModel'
        pagination:
          $ref: '#/components/schemas/PaginationResponse'
      required:
        - data
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, `UNPROCESSABLE_ENTITY`, or
                `PAYMENT_REQUIRED`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
                - PAYMENT_REQUIRED
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable description of the error
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    RestoreJobModel:
      description: The RestoreJobModel describes the status of a restore job.
      type: object
      properties:
        restore_job_id:
          example: 670e8400-e29b-41d4-a716-446655440001
          description: Unique identifier for the restore job
          type: string
        backup_id:
          example: 670e8400-e29b-41d4-a716-446655440000
          description: Backup used for the restore
          type: string
        target_index_name:
          example: sample-index
          description: Name of the index into which data is being restored
          type: string
        target_index_id:
          example: 670e8400-e29b-41d4-a716-446655440002
          description: ID of the index
          type: string
        status:
          example: Completed
          description: Status of the restore job
          type: string
        created_at:
          example: '2025-02-04T13:00:00.000Z'
          description: Timestamp when the restore job started
          type: string
          format: date-time
        completed_at:
          example: '2025-02-04T14:00:00.000Z'
          description: Timestamp when the restore job finished
          type: string
          format: date-time
        percent_complete:
          example: 42.2
          description: The progress made by the restore job out of 100
          type: number
          format: float
          minimum: 0
          maximum: 100
      required:
        - restore_job_id
        - backup_id
        - target_index_name
        - target_index_id
        - status
        - created_at
    PaginationResponse:
      example:
        next: dXNlcl9pZD11c2VyXzE=
      description: The pagination object that is returned with paginated responses.
      type: object
      properties:
        next:
          example: dXNlcl9pZD11c2VyXzE=
          description: The token to use to retrieve the next page of results.
          type: string
      required:
        - next
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/listnamespaces.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# List namespaces

> List all namespaces in a serverless index.

Up to 100 namespaces are returned at a time by default, in sorted order (bitwise “C” collation). If the `limit` parameter is set, up to that number of namespaces are returned instead. Whenever there are additional namespaces to return, the response also includes a `pagination_token` that you can use to get the next batch of namespaces. When the response does not include a `pagination_token`, there are no more namespaces to return.

For guidance and examples, see [Manage namespaces](https://docs.pinecone.io/guides/manage-data/manage-namespaces).

**Note:** This operation is not supported for pod-based indexes.

<RequestExample>
  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/namespaces" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "namespaces": [
      {
        "name": "example-namespace",
        "record_count": 20000
      },
      {
        "name": "example-namespace2",
        "record_count": 10500
      },
      ...
    ],
    "pagination": {
      "next": "Tm90aGluZyB0byBzZWUgaGVyZQo="
    }
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml get /namespaces
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /namespaces:
    get:
      tags:
        - Namespace Operations
      summary: List namespaces
      description: >-
        List all namespaces in a serverless index.


        Up to 100 namespaces are returned at a time by default, in sorted order
        (bitwise “C” collation). If the `limit` parameter is set, up to that
        number of namespaces are returned instead. Whenever there are additional
        namespaces to return, the response also includes a `pagination_token`
        that you can use to get the next batch of namespaces. When the response
        does not include a `pagination_token`, there are no more namespaces to
        return.


        For guidance and examples, see [Manage
        namespaces](https://docs.pinecone.io/guides/manage-data/manage-namespaces).


        **Note:** This operation is not supported for pod-based indexes.
      operationId: listNamespacesOperation
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: query
          name: limit
          description: Max number namespaces to return per page.
          schema:
            type: integer
            format: int32
          example: 10
          style: form
        - in: query
          name: paginationToken
          description: Pagination token to continue a previous listing operation.
          schema:
            type: string
          style: form
        - in: query
          name: prefix
          description: >-
            Prefix of the namespaces to list. Acts as a filter to return only
            namespaces that start with this prefix.
          schema:
            type: string
          example: prefixExample
          style: form
      responses:
        '200':
          description: A successful response
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ListNamespacesResponse'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    ListNamespacesResponse:
      type: object
      properties:
        namespaces:
          description: The list of namespaces belonging to this index.
          type: array
          items:
            $ref: '#/components/schemas/NamespaceDescription'
        pagination:
          $ref: '#/components/schemas/Pagination'
        total_count:
          example: 25
          description: The total number of namespaces in the index matching the prefix
          type: integer
          format: int32
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    NamespaceDescription:
      description: A description of a namespace, including the name and record count.
      type: object
      properties:
        name:
          example: example-namespace
          description: The name of the namespace.
          type: string
        record_count:
          example: 20000
          description: The total amount of records within the namespace.
          type: integer
          format: int64
        schema:
          example:
            fields:
              description:
                filterable: true
              genre:
                filterable: true
              year:
                filterable: true
          description: >-
            Schema for the behavior of Pinecone's internal metadata index. By
            default, all metadata is indexed; when `schema` is present, only
            fields which are present in the `fields` object with a `filterable:
            true` are indexed. Note that `filterable: false` is not currently
            supported.
          type: object
          properties:
            fields:
              description: >-
                A map of metadata field names to their configuration. The field
                name must be a valid metadata field name. The field name must be
                unique.
              type: object
              additionalProperties:
                type: object
                properties:
                  filterable:
                    description: >-
                      Whether the field is filterable. If true, the field is
                      indexed and can be used in filters. Only true values are
                      allowed.
                    type: boolean
          required:
            - fields
        indexed_fields:
          description: A list of all indexed metatadata fields in the namespace
          type: object
          properties:
            fields:
              type: array
              items:
                type: string
              example:
                - genre
                - year
                - author
    Pagination:
      type: object
      properties:
        next:
          example: Tm90aGluZyB0byBzZWUgaGVyZQo=
          type: string
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/integrations/llamaindex.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# LlamaIndex

> Using LlamaIndex and Pinecone to build semantic search and RAG applications

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

LlamaIndex is a framework for connecting data sources to LLMs, with its chief use case being the end-to-end development of retrieval augmented generation (RAG) applications. LlamaIndex provides the essential abstractions to more easily ingest, structure, and access private or domain-specific data in order to inject these safely and reliably into LLMs for more accurate text generation. It’s available in Python and Typescript.

Seamlessly integrate Pinecone vector database with LlamaIndex to build semantic search and RAG applications.

<PrimarySecondaryCTA secondaryHref={"#setup-guide"} secondaryLabel={"View setup guide"} />

## Setup guide

[View source](https://github.com/pinecone-io/examples/blob/master/learn/generation/llama-index/using-llamaindex-with-pinecone.ipynb)

[Open in Colab](https://colab.research.google.com/github/pinecone-io/examples/blob/master/learn/generation/llama-index/using-llamaindex-with-pinecone.ipynb)

[LlamaIndex](https://www.llamaindex.ai/) is a framework for connecting data sources to LLMs, with its chief use case being the end-to-end development of [RAG applications](https://www.pinecone.io/learn/retrieval-augmented-generation/). Compared to other similar frameworks, LlamaIndex offers a wide variety of tools for pre- and post-processing your data.

This guide shows you how to use LlamaIndex and Pinecone to both perform traditional semantic search and build a RAG pipeline. Specifically, you will:

* Load, transform, and vectorize sample data with LlamaIndex
* Index and store the vectorized data in Pinecone
* Search the data in Pinecone and use the results to augment an LLM call
* Evaluate the answer you get back from the LLM

<Note>
  This guide demonstrates only one way out of many that you can use LlamaIndex as part of a RAG pipeline. See LlamaIndex's section on [Advanced RAG](https://docs.llamaindex.ai/en/stable/optimizing/advanced%5Fretrieval/advanced%5Fretrieval.html) to learn more about what's possible.
</Note>

### Set up your environment

Before you begin, install some necessary libraries and set environment variables for your Pinecone and OpenAI API keys:

```Shell Shell theme={null}
# Install libraries
pip install -qU \
    "pinecone[grpc]"==5.1.0 \
    llama-index==0.11.4 \
    llama-index-vector-stores-pinecone==0.2.1 \
    llama-index-readers-file==0.2.0 \
    arxiv==2.1.3 \
    setuptools  # (Optional)
```

```Shell Shell theme={null}
# Set environment variables for API keys
export PINECONE_API_KEY=<your Pinecone API key available at app.pinecone.io>
export OPENAI_API_KEY=<your OpenAI API key, available at platform.openai.com/api-keys>
pinecone_api_key = os.environ.get('PINECONE_API_KEY')
openai_api_key = os.environ.get('OPENAI_API_KEY')
```

Also note that all code on this page is run on Python 3.11.

### Load the data

In this guide, you will use the [canonical HNSW paper](https://arxiv.org/pdf/1603.09320.pdf) by Yuri Malkov (PDF) as your sample dataset. Your first step is to download the PDF from arXiv.org and load it into a LlamaIndex loader called [PDF Loader](https://llamahub.ai/l/file-pdf?from=all). This Loader is available (along with many more) on the [LlamaHub](https://llamahub.ai/), which is a directory of data loaders.

```Python Python theme={null}
import arxiv
from pathlib import Path
from llama_index.readers.file import PDFReader

# Download paper to local file system (LFS)
# `id_list` contains 1 item that matches our PDF's arXiv ID
paper = next(arxiv.Client().results(arxiv.Search(id_list=["1603.09320"])))
paper.download_pdf(filename="hnsw.pdf")

# Instantiate `PDFReader` from LlamaHub
loader = PDFReader()

# Load HNSW PDF from LFS
documents = loader.load_data(file=Path('./hnsw.pdf'))

# Preview one of our documents
documents[0]
# Response:
# Document(id_='e25106d2-bde5-41f0-83fa-5cbfa8234bef', embedding=None, metadata={'page_label': '1', 'file_name': 'hnsw.pdf'}, excluded_embed_metadata_keys=[], excluded_llm_metadata_keys=[], relationships={}, text="IEEE TRANSACTIONS ON  JOURNAL NAME,  MANUS CRIPT ID  1 \n Efficient and robust approximate nearest \nneighbor search using Hierarchical Navigable \nSmall World graphs  \nYu. A. Malkov,  D. A. Yashunin  \nAbstract  — We present a new approach for the approximate K -nearest neighbor search based on navigable small world \ngraphs with controllable hierarchy (Hierarchical NSW , HNSW ) and tree alg o-\nrithms", start_char_idx=None, end_char_idx=None, text_template='{metadata_str}\n\n{content}', metadata_template='{key}: {value}', metadata_seperator='\n')
```

You can see above that each `Document` has a ton of useful information, but depending on which Loader you choose, you may have to clean your data. In this case, you need to remove things like remaining `\n` characters and broken, hyphenated words (e.g., `alg o-\nrithms` → `algorithms`).

```Python Python theme={null}
# Clean up our Documents' content
import re

def clean_up_text(content: str) -> str:
    """
    Remove unwanted characters and patterns in text input.

    :param content: Text input.
    
    :return: Cleaned version of original text input.
    """

    # Fix hyphenated words broken by newline
    content = re.sub(r'(\w+)-\n(\w+)', r'\1\2', content)

    # Remove specific unwanted patterns and characters
    unwanted_patterns = [
        "\\n", "  —", "——————————", "—————————", "—————",
        r'\\u[\dA-Fa-f]{4}', r'\uf075', r'\uf0b7'
    ]
    for pattern in unwanted_patterns:
        content = re.sub(pattern, "", content)

    # Fix improperly spaced hyphenated words and normalize whitespace
    content = re.sub(r'(\w)\s*-\s*(\w)', r'\1-\2', content)
    content = re.sub(r'\s+', ' ', content)

    return content

# Call function
cleaned_docs = []
for d in documents: 
    cleaned_text = clean_up_text(d.text)
    d.text = cleaned_text
    cleaned_docs.append(d)

# Inspect output
cleaned_docs[0].get_content()
# Response:
# "IEEE TRANSACTIONS ON JOURNAL NAME, MANUS CRIPT ID 1 Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphs Yu. A. Malkov, D. A. Yashunin Abstract We present a new approach for the approximate K-nearest neighbor search based on navigable small world graphs with controllable hierarchy (Hierarchical NSW , HNSW ) and tree algorithms."

# Great!
```

The value-add of using a file loader from LlamaHub is that your PDF is already broken down into LlamaIndex [Documents](https://docs.llamaindex.ai/en/stable/module%5Fguides/loading/documents%5Fand%5Fnodes/root.html#documents-nodes). Along with each Document object comes a [customizable](https://docs.llamaindex.ai/en/stable/module%5Fguides/loading/documents%5Fand%5Fnodes/usage%5Fdocuments.html#metadata) metadata dictionary and a hash ID, among other useful artifacts.

### Transform the data

#### Metadata

Now, if you look at one of your cleaned Document objects, you'll see that the default values in your metadata dictionary are not particularly useful.

```Python Python theme={null}
cleaned_docs[0].metadata
# Response:
# {'page_label': '1', 'file_name': 'hnsw.pdf'}
```

To add some metadata that would be more helpful, let's add author name and the paper's title. Note that whatever metadata you add to the metadata dictionary will apply to all [Nodes](https://docs.llamaindex.ai/en/stable/module%5Fguides/loading/documents%5Fand%5Fnodes/root.html#nodes), so you want to keep your additions high-level.

<Note>
  LlamaIndex also provides [advanced customizations](https://docs.llamaindex.ai/en/stable/module%5Fguides/loading/documents%5Fand%5Fnodes/usage%5Fdocuments.html#advanced-metadata-customization) for what metadata the LLM can see vs the embedding, etc.
</Note>

```Python Python theme={null}
# Iterate through `documents` and add our new key:value pairs
metadata_additions = {"authors": ["Yu. A. Malkov", "D. A. Yashunin"],
  "title": "Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphs"}

 # Update dict in place
[cd.metadata.update(metadata_additions) for cd in cleaned_docs]
 
# Let\'s confirm everything worked:
cleaned_docs[0].metadata
# Response:
# {'page_label': '1',
#      'file_name': 'hnsw.pdf',
#      'authors': ['Yu. A. Malkov', 'D. A. Yashunin'],
#      'title': 'Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphs'}

# Great!
```

#### Ingestion pipeline

The easiest way to turn your data into indexable vectors and put those into Pinecone is to make what's called an [Ingestion Pipeline](https://docs.llamaindex.ai/en/stable/module%5Fguides/loading/ingestion%5Fpipeline/root.html). Ingestion Pipelines are how you will build a pipeline that will take your list of Documents, parse them into Nodes (or “[chunks](https://www.pinecone.io/learn/chunking-strategies/)” in non-LlamaIndex contexts), vectorize each Node's content, and upsert them into Pinecone.

In the following pipeline, you'll use one of LlamaIndex's newer parsers: the [SemanticSplitterNodeParser](https://docs.llamaindex.ai/en/stable/module%5Fguides/loading/node%5Fparsers/modules.html#semanticsplitternodeparser), which uses OpenAI's [ada-002 embedding model](https://github.com/run-llama/llama%5Findex/blob/47b34d1fdfde2ded134a373b620c3e7a694e8380/llama%5Findex/embeddings/openai.py#L216) to split Documents into semantically coherent Nodes.

This step uses the OpenAI API key you set as an environment variable [earlier](#set-up-your-environment).

```Python Python theme={null}
import os

from llama_index.node_parser import SemanticSplitterNodeParser
from llama_index.embeddings import OpenAIEmbedding
from llama_index.ingestion import IngestionPipeline

# This will be the model we use both for Node parsing and for vectorization
embed_model = OpenAIEmbedding(api_key=openai_api_key)

# Define the initial pipeline
pipeline = IngestionPipeline(
    transformations=[
        SemanticSplitterNodeParser(
            buffer_size=1,
            breakpoint_percentile_threshold=95, 
            embed_model=embed_model,
            ),
        embed_model,
        ],
    )
```

Hold off on running this pipeline; you will modify it below.

### Upsert the data

Above, you defined an Ingestion Pipeline. There's one thing missing, though: a vector database into which you can upsert your transformed data.

LlamaIndex lets you declare a [VectorStore](https://docs.llamaindex.ai/en/stable/examples/vector%5Fstores/pinecone%5Fmetadata%5Ffilter.html) and add that right into the pipeline for super easy ingestion. Let's do that with Pinecone below.

This step uses the Pinecone API key you set as an environment variable [earlier](#set-up-your-environment).

```Python Python theme={null}
from pinecone.grpc import PineconeGRPC
from pinecone import ServerlessSpec

from llama_index.vector_stores.pinecone import PineconeVectorStore

# Initialize connection to Pinecone
pc = PineconeGRPC(api_key=pinecone_api_key)
index_name = "llama-integration-example"

# Create your index (can skip this step if your index already exists)
pc.create_index(
    index_name,
    dimension=1536,
    spec=ServerlessSpec(cloud="aws", region="us-east-1"),
)

# Initialize your index 
pinecone_index = pc.Index(index_name)

# Initialize VectorStore
vector_store = PineconeVectorStore(pinecone_index=pinecone_index)
```

With your PineconeVectorStore now initialized, you can pop that into your `pipeline` and run it.

```Python Python theme={null}
# Our pipeline with the addition of our PineconeVectorStore
pipeline = IngestionPipeline(
    transformations=[
        SemanticSplitterNodeParser(
            buffer_size=1,
            breakpoint_percentile_threshold=95, 
            embed_model=embed_model,
            ),
        embed_model,
        ],
        vector_store=vector_store  # Our new addition
    )

# Now we run our pipeline!
pipeline.run(documents=cleaned_docs)
```

Now ensure your index is up and running with some Pinecone-native methods like `.describe_index_stats()`:

```Python Python theme={null}
pinecone_index.describe_index_stats()

# Response:
# {'dimension': 1536,
# 'index_fullness': 0.0,
# 'namespaces': {'': {'vector_count': 46}},
# 'total_vector_count': 46}
```

Awesome, your index now has vectors in it. Since you have 46 vectors, you can infer that your `SemanticSplitterNodeParser` split your list of Documents into 46 Nodes.

#### Query the data

To fetch search results from Pinecone itself, you need to make a [VectorStoreIndex](https://docs.llamaindex.ai/en/stable/module%5Fguides/indexing/vector%5Fstore%5Findex.html) object and a [VectorIndexRetriever](https://github.com/run-llama/llama%5Findex/blob/main/llama%5Findex/indices/vector%5Fstore/retrievers/retriever.py#L21) object. You can then pass natural language queries to your Pinecone index and receive results.

```Python Python theme={null}
from llama_index.core import VectorStoreIndex
from llama_index.core.retrievers import VectorIndexRetriever

# Instantiate VectorStoreIndex object from your vector_store object
vector_index = VectorStoreIndex.from_vector_store(vector_store=vector_store)

# Grab 5 search results
retriever = VectorIndexRetriever(index=vector_index, similarity_top_k=5)

# Query vector DB
answer = retriever.retrieve('How does logarithmic complexity affect graph construction?')

# Inspect results
print([i.get_content() for i in answer])

# Response:
# ['some relevant search result 1', 'some relevant search result 1'...]
```

These search results can now be plugged into any downstream task you want.

One of the most common ways to use vector database search results is as additional context to augment a query sent to an LLM. This workflow is what's commonly referred to as a [RAG application](https://www.pinecone.io/learn/retrieval-augmented-generation/).

### Build a RAG app with the data

Building a RAG app with LlamaIndex is very simple.

In theory, you could create a simple [Query Engine](https://docs.llamaindex.ai/en/stable/module%5Fguides/deploying/query%5Fengine/usage%5Fpattern.html#usage-pattern) out of your `vector_index` object by calling `vector_index.as_query_engine().query(‘some query')`, but then you wouldn't be able to specify the number of Pinecone search results you'd like to use as context.

To control how many search results your RAG app uses from your Pinecone index, you will instead create your Query Engine using the [RetrieverQueryEngine](https://github.com/run-llama/llama%5Findex/blob/main/llama%5Findex/query%5Fengine/retriever%5Fquery%5Fengine.py#L21) class. This class allows you to pass in the `retriever` created above, which you configured to retrieve the top 5 search results.

```Python Python theme={null}
from llama_index.core.query_engine import RetrieverQueryEngine

# Pass in your retriever from above, which is configured to return the top 5 results
query_engine = RetrieverQueryEngine(retriever=retriever)

# Now you query:
llm_query = query_engine.query('How does logarithmic complexity affect graph construction?')

llm_query.response
# Response:
# 'Logarithmic complexity in graph construction affects the construction process by organizing the graph into different layers based on their length scale. This separation of links into layers allows for efficient and scalable routing in the graph. The construction algorithm starts from the top layer, which contains the longest links, and greedily traverses through the elements until a local minimum is reached. Then, the search switches to the lower layer with shorter links, and the process repeats. By keeping the maximum number of connections per element constant in all layers, the routing complexity in the graph scales logarithmically. This logarithmic complexity is achieved by assigning an integer level to each element, determining the maximum layer it belongs to. The construction algorithm incrementally builds a proximity graph for each layer, consisting of "short" links that approximate the Delaunay graph. Overall, logarithmic complexity in graph construction enables efficient and robust approximate nearest neighbor search.'
```

You can even inspect the context (Nodes) that informed your LLM's answer using the `.source_nodes` attribute. Let's inspect the first Node:

```Python Python theme={null}
llm_response_source_nodes = [i.get_content() for i in llm_query.source_nodes]

llm_response_source_nodes
# Response:
# ["AUTHOR ET AL.: TITL E 7 be auto-configured by using sample data. The construction process can be easily and efficiently parallelized with only few synchronization points (as demonstrated in Fig. 9) and no measurable effect on index quality. Construction speed/index q uality tradeoff is co ntrolled via the efConstruction parameter. The tradeoff between the search time and the index construction time is presented in Fig. 10 for a 10M SIFT dataset and shows that a reasonable quality index can be constructed for efConstruct ion=100 on a 4X 2.4 GHz 10-core X..."]
```

### Evaluate the data

Now that you've made a RAG app and queried your LLM, you need to evaluate its response.

With LlamaIndex, there are [many ways](https://docs.llamaindex.ai/en/module%5Fguides/evaluating/usage%5Fpattern.html#) to evaluate the results your RAG app generates. A great way to get started with evaluation is to confirm (or deny) that your LLM's responses are relevant, given the context retrieved from your vector database. To do this, you can use LlamaIndex's [RelevancyEvaluator](https://docs.llamaindex.ai/en/stable/examples/evaluation/relevancy%5Feval.html#relevancy-evaluator) class.

The great thing about this type of evaluation is that there is no need for [ground truth data](https://dtunkelang.medium.com/evaluating-search-using-human-judgement-fbb2eeba37d9) (i.e., labeled datasets to compare answers with).

```Python Python theme={null}
from llama_index.core.evaluation import RelevancyEvaluator

# (Need to avoid peripheral asyncio issues)
import nest_asyncio
nest_asyncio.apply()

# Define evaluator
evaluator = RelevancyEvaluator()

# Issue query
llm_response = query_engine.query(
    "How does logarithmic complexity affect graph construction?"
)

# Grab context used in answer query & make it pretty
llm_response_source_nodes = [i.get_content() for i in llm_response.source_nodes]

# # Take your previous question and pass in the response youwe got above
eval_result = evaluator.evaluate_response(query="How does logarithmic complexity affect graph construction?", response=llm_response)

# Print response
print(f'\nGiven the {len(llm_response_source_nodes)} chunks of content (below), is your \     
        LLM\'s response relevant? {eval_result.passing}\n \
        \n ----Contexts----- \n \
        \n{llm_response_source_nodes}')
# Response:
# "Given the 5 chunks of content (below), is your LLM's response relevant? True
         
#  ----Contexts----- 
         
# ['AUTHOR ET AL.: TITL E 7 be auto-configured by using sample data. The construction process can be easily and efficiently parallelized with only few synchronization points (as demonstrated in Fig...']"
```

You can see that there are various attributes you can inspect on your evaluator's result in order to ascertain what's going on behind the scenes. To get a quick binary True/False signal as to whether your LLM is producing relevant results given your context, inspect the `.passing` attribute.

Let's see what happens when we send a totally out of scope query through your RAG app. Issue a random query you know your RAG app won't be able to answer, given what's in your index:

```Python Python theme={null}
query = "Why did the chicken cross the road?"
response = query_engine.query(query)

print(response.response)
# Response:
# "I'm sorry, but I cannot answer that question based on the given context information."

# Evaluate
eval_result = evaluator.evaluate_response(query=query, response=response)

print(str(eval_result.passing))
# Response:
# False  # Our LLM is not taking our context into account, as expected :)
```

As expected, when you send an out-of-scope question through your RAG pipeline, your evaluator says the LLM's answer is not relevant to the retrieved context.

### Summary

As you have seen, LlamaIndex is a powerful framework to use when building semantic search and RAG applications – and we have only gotten to the tip of the iceberg! [Explore more](https://docs.llamaindex.ai/en/index.html) on your own and [let us know how it goes](https://community.pinecone.io/).

---

# Source: https://docs.pinecone.io/guides/operations/local-development.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Local development with Pinecone Local

> Develop locally with an in-memory Pinecone emulator.

Pinecone Local is an in-memory Pinecone emulator available as a Docker image.

This page shows you how to use Pinecone Local to develop your applications locally without connecting to your Pinecone account or incurring usage or storage fees.

<Warning>
  Pinecone Local is not suitable for production. See [Limitations](#limitations) for details.
</Warning>

<Note>
  This feature is in [public preview](/release-notes/feature-availability).
</Note>

## Limitations

Pinecone Local has the following limitations:

* Pinecone Local uses the `2025-01` API version, which is not the latest stable version.
* Pinecone Local is available in Docker only.
* Pinecone Local is an in-memory emulator and is not suitable for production. Records loaded into Pinecone Local do not persist after it is stopped.
* Pinecone Local does not authenticate client requests. API keys are ignored.
* Max number of records per index: 100,000.

Pinecone Local does not currently support the following features:

* [Import from object storage](/guides/index-data/import-data)
* [Backup/restore of serverless indexes](/guides/manage-data/backups-overview)
* [Collections for pod-based indexes](/guides/indexes/pods/understanding-collections)
* [Namespace management](/guides/manage-data/manage-namespaces)
* [Pinecone Inference](/reference/api/introduction#inference)
* [Pinecone Assistant](/guides/assistant/overview)

## 1. Start Pinecone Local

You can configure Pinecone Local as an index emulator or database emulator:

* **Index emulator** - This approach uses the `pinecone-index` Docker image to create and configure indexes on startup. This is recommended when you want to quickly experiment with reading and writing data without needing to manage the index lifecycle.

  <Note>
    With index emulation, you can only read and write data to the indexes created at startup. You cannot create new indexes, list indexes, or run other operations that do not involve reading and writing data.
  </Note>

* **Database emulator** - This approach uses the `pinecone-local` Docker image to emulate Pinecone Database more broadly. This is recommended when you want to test your production app or manually create and manage indexes.

### Index emulator

<Tabs>
  <Tab title="Docker Compose">
    Make sure [Docker](https://docs.docker.com/get-docker/) is installed and running on your local machine.

    Create a `docker-compose.yaml` file that defines a service for each index that Pinecone Local should create on startup. In this file, include the `pinecone-index` Docker image, a localhost port for the index to use, and other details:

    ```yaml  theme={null}
    services:
      dense-index:
        image: ghcr.io/pinecone-io/pinecone-index:latest
        container_name: dense-index
        environment:
          PORT: 5081
          INDEX_TYPE: serverless
          VECTOR_TYPE: dense
          DIMENSION: 2 
          METRIC: cosine
        ports:
          - "5081:5081"
        platform: linux/amd64
      sparse-index:
        image: ghcr.io/pinecone-io/pinecone-index:latest
        container_name: sparse-index
        environment:
          PORT: 5082
          INDEX_TYPE: serverless
          VECTOR_TYPE: sparse
          DIMENSION: 0
          METRIC: dotproduct
        ports:
          - "5082:5082"
        platform: linux/amd64
    ```

    For each index, update the environment variables as needed:

    * `PORT`: Specify the port number for the index to listen on.
    * `INDEX_TYPE`: Specify the type of Pinecone index to create. Accepted values: `serverless` or `pod`.
    * `VECTOR_TYPE`: Specify the [type of vectors](/guides/index-data/indexing-overview#indexes) you will store in the index. Accepted values: `dense` or `sparse`.

      <Note>Sparse is supported only with serverless indexes.</Note>
    * `DIMENSION`: Specify the dimension of vectors you will store in the index.

      <Note>For sparse indexes, this must be set to `0`.</Note>
    * `METRIC`: Specify the [distance metric](/guides/index-data/indexing-overview#distance-metrics) for calculating the similarity between vectors in the index. Accepted values for dense indexes: `cosine`, `euclidean`, or `dotproduct`. Accepted value for sparse indexes: `dotproduct`.

    To start Pinecone Local, run the following command:

    ```shell  theme={null}
    docker compose up -d
    ```

    You'll see a message with details about each index.
  </Tab>

  <Tab title="Docker CLI">
    Make sure [Docker](https://docs.docker.com/get-docker/) is installed and running on your local machine.

    Download the latest `pinecone-index` Docker image:

    ```shell  theme={null}
    docker pull ghcr.io/pinecone-io/pinecone-index:latest
    ```

    Start Pinecone Local with one or more indexes:

    ```shell  theme={null}
    docker run -d \
    --name dense-index \
    -e PORT=5081 \
    -e INDEX_TYPE=serverless \
    -e VECTOR_TYPE=dense \
    -e DIMENSION=2 \
    -e METRIC=cosine \
    -p 5081:5081 \
    --platform linux/amd64 \
    ghcr.io/pinecone-io/pinecone-index:latest
    ```

    ```shell  theme={null}
    docker run -d \
    --name sparse-index \
    -e PORT=5082 \
    -e INDEX_TYPE=serverless \
    -e VECTOR_TYPE=sparse \
    -e DIMENSION=0 \
    -e METRIC=dotproduct \
    -p 5082:5082 \
    --platform linux/amd64 \
    ghcr.io/pinecone-io/pinecone-index:latest
    ```

    For each index, update the environment variables as needed:

    * `PORT`: Specify the port number for the index to listen on.
    * `INDEX_TYPE`: Specify the type of Pinecone index to create. Accepted values: `serverless` or `pod`.
    * `VECTOR_TYPE`: Specify the [type of vectors](/guides/index-data/indexing-overview#indexes) you will store in the index. Accepted values: `dense` or `sparse`.

      <Note>Sparse is supported only with serverless indexes.</Note>
    * `DIMENSION`: Specify the dimension of vectors you will store in the index.

      <Note>For sparse indexes, this must be set to `0`.</Note>
    * `METRIC`: Specify the [distance metric](/guides/index-data/indexing-overview#distance-metrics) for calculating the similarity between vectors in the index. Accepted values for dense indexes: `cosine`, `euclidean`, or `dotproduct`. Accepted value for sparse indexes: `dotproduct`.
  </Tab>
</Tabs>

### Database emulator

<Tabs>
  <Tab title="Docker Compose">
    Make sure [Docker](https://docs.docker.com/get-docker/) is installed and running on your local machine.

    Create a `docker-compose.yaml` file that defines a service for Pinecone local, including the `pinecone-local` Docker image, the host and port that Pinecone Local will run on and the range of ports that will be available for indexes:

    ```yaml  theme={null}
    services:
      pinecone:
        image: ghcr.io/pinecone-io/pinecone-local:latest
        environment: 
          PORT: 5080
          PINECONE_HOST: localhost
        ports: 
          - "5080-5090:5080-5090"
        platform: linux/amd64
    ```

    To start Pinecone Local, run the following command:

    ```shell  theme={null}
    docker compose up -d
    ```

    You'll see a message with details about the Pinecone Local instance.
  </Tab>

  <Tab title="Docker CLI">
    Make sure [Docker](https://docs.docker.com/get-docker/) is installed and running on your local machine.

    Download the latest `pinecone-local` Docker image:

    ```shell  theme={null}
    docker pull ghcr.io/pinecone-io/pinecone-local:latest
    ```

    Start Pinecone Local:

    ```shell  theme={null}
    docker run -d \
    --name pinecone-local \
    -e PORT=5080 \
    -e PINECONE_HOST=localhost \
    -p 5080-5090:5080-5090 \
    --platform linux/amd64 \
    ghcr.io/pinecone-io/pinecone-local:latest
    ```

    This command defines the host and port that Pinecone Local will run on, as well as the range of ports that will be available for indexes.
  </Tab>
</Tabs>

## 2. Develop your app

Running code against Pinecone Local is just like running code against your Pinecone account, with the following differences:

* Pinecone Local does not authenticate client requests. API keys are ignored.

* The latest version of Pinecone Local uses [Pinecone API version](/reference/api/versioning) `2025-01` and requires [Python SDK](/reference/sdks/python/overview) `v6.x` or later, [Node.js SDK](/reference/sdks/node/overview) `v5.x` or later, [Java SDK](/reference/sdks/java/overview) `v4.x` or later, [Go SDK](/reference/sdks/go/overview) `v3.x` or later, and [.NET SDK](/reference/sdks/dotnet/overview) `v3.x` or later.

<Note>
  Be sure to review the [limitations](#limitations) of Pinecone Local before using it for development or testing.
</Note>

**Example**

The following example assumes that you have [started Pinecone Local without indexes](/guides/operations/local-development#database-emulator). It initializes a client, creates a [dense index](/guides/index-data/indexing-overview#dense-indexes) and a [sparse index](/guides/index-data/indexing-overview#sparse-indexes), upserts records into the indexes, checks their record counts, and queries the indexes.

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC, GRPCClientConfig
  from pinecone import ServerlessSpec

  # Initialize a client.
  # API key is required, but the value does not matter.
  # Host and port of the Pinecone Local instance
  # is required when starting without indexes. 
  pc = PineconeGRPC(
      api_key="pclocal", 
      host="http://localhost:5080" 
  )                                    

  # Create two indexes, one dense and one sparse
  dense_index_name = "dense-index"
  sparse_index_name = "sparse-index"

  if not pc.has_index(dense_index_name):  
      dense_index_model = pc.create_index(
          name=dense_index_name,
          vector_type="dense",
          dimension=2,
          metric="cosine",
          spec=ServerlessSpec(cloud="aws", region="us-east-1"),
          deletion_protection="disabled",
          tags={"environment": "development"}
      )

  print("Dense index model:\n", dense_index_model)

  if not pc.has_index(sparse_index_name):  
      sparse_index_model = pc.create_index(
          name=sparse_index_name,
          vector_type="sparse",
          metric="dotproduct",
          spec=ServerlessSpec(cloud="aws", region="us-east-1"),
          deletion_protection="disabled",
          tags={"environment": "development"}
      )

  print("\nSparse index model:\n", sparse_index_model)

  # Target each index, disabling tls
  dense_index_host = pc.describe_index(name=dense_index_name).host
  dense_index = pc.Index(host=dense_index_host, grpc_config=GRPCClientConfig(secure=False))
  sparse_index_host = pc.describe_index(name=sparse_index_name).host
  sparse_index = pc.Index(host=sparse_index_host, grpc_config=GRPCClientConfig(secure=False))

  # Upsert records into the dense index
  dense_index.upsert(
      vectors=[
          {
              "id": "vec1", 
              "values": [1.0, -2.5],
              "metadata": {"genre": "drama"}
          },
          {
              "id": "vec2", 
              "values": [3.0, -2.0],
              "metadata": {"genre": "documentary"}
          },
          {
              "id": "vec3", 
              "values": [0.5, -1.5],
              "metadata": {"genre": "documentary"}
          }
      ],
      namespace="example-namespace"
  )

  # Upsert records into the sparse index
  sparse_index.upsert(
      namespace="example-namespace",
      vectors=[
          {
              "id": "vec1",
              "sparse_values": {
                  "values": [1.7958984, 0.41577148, 2.828125, 2.8027344, 2.8691406, 1.6533203, 5.3671875, 1.3046875, 0.49780273, 0.5722656, 2.71875, 3.0820312, 2.5019531, 4.4414062, 3.3554688],
                  "indices": [822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191]
              },
              "metadata": {
                  "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
                  "category": "technology",
                  "quarter": "Q3"
              }
          },
          {
              "id": "vec2",
              "sparse_values": {
                  "values": [0.4362793, 3.3457031, 2.7714844, 3.0273438, 3.3164062, 5.6015625, 2.4863281, 0.38134766, 1.25, 2.9609375, 0.34179688, 1.4306641, 0.34375, 3.3613281, 1.4404297, 2.2558594, 2.2597656, 4.8710938, 0.5605469],
                  "indices": [131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697]
              },
              "metadata": {
                  "chunk_text": "Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
                  "category": "technology",
                  "quarter": "Q4"
              }
          },
          {
              "id": "vec3",
              "sparse_values": {
                  "values": [2.6875, 4.2929688, 3.609375, 3.0722656, 2.1152344, 5.78125, 3.7460938, 3.7363281, 1.2695312, 3.4824219, 0.7207031, 0.0826416, 4.671875, 3.7011719, 2.796875, 0.61621094],
                  "indices": [8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697]
              },
              "metadata": {
                  "chunk_text": "AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production",
                  "category": "technology",
                  "quarter": "Q3"
              }
          }
      ]
  )

  # Check the number of records in each index
  print("\nDense index stats:\n", dense_index.describe_index_stats())
  print("\nSparse index stats:\n", sparse_index.describe_index_stats())

  # Query the dense index with a metadata filter
  dense_response = dense_index.query(
      namespace="example-namespace",
      vector=[3.0, -2.0],
      filter={"genre": {"$eq": "documentary"}},
      top_k=1,
      include_values=False,
      include_metadata=True
  )

  print("\nDense query response:\n", dense_response)

  # Query the sparse index with a metadata filter
  sparse_response = sparse_index.query(
      namespace="example-namespace",
      sparse_vector={
        "values": [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],
        "indices": [767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697]
      }, 
      filter={
          "quarter": {"$eq": "Q4"}
      },
      top_k=1,
      include_values=False,
      include_metadata=True
  )

  print("/nSparse query response:\n", sparse_response)

  # Delete the indexes
  pc.delete_index(name=dense_index_name)
  pc.delete_index(name=sparse_index_name)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  // Initialize a client.
  // API key is required, but the value does not matter.
  // Host and port of the Pinecone Local instance
  // is required when starting without indexes.
  const pc = new Pinecone({ 
      apiKey: 'pclocal', 
      controllerHostUrl: 'http://localhost:5080' 
  });

  // Create two indexes, one dense and one sparse
  const denseIndexName = 'dense-index';
  const sparseIndexName = 'sparse-index';

  const denseIndexModel = await pc.createIndex({
    name: denseIndexName,
    vectorType: 'dense',
    dimension: 2,
    metric: 'cosine',
    spec: {
      serverless: {
        cloud: 'aws',
        region: 'us-east-1'
      }
    },
    deletionProtection: 'disabled',
    tags: { environment: 'development' }, 
  });

  console.log('Dense index model:', denseIndexModel);

  const sparseIndexModel = await pc.createIndex({
      name: sparseIndexName,
      vectorType: 'sparse',
      metric: 'dotproduct',
      spec: {
        serverless: {
          cloud: 'aws',
          region: 'us-east-1'
        }
      },
      deletionProtection: 'disabled',
      tags: { environment: 'development' }, 
    });
    
  console.log('\nSparse index model:', sparseIndexModel);
    
  // Target each index
  const denseIndexHost = (await pc.describeIndex(denseIndexName)).host;
  const denseIndex = await pc.index(denseIndexName, 'http://' + denseIndexHost);
  const sparseIndexHost = (await pc.describeIndex(sparseIndexName)).host;
  const sparseIndex = await pc.index(sparseIndexName, 'http://' + sparseIndexHost);

  // Upsert records into the dense index
  await denseIndex.namespace('example-namespace').upsert([
      {
          id: 'vec1', 
          values: [1.0, -2.5],
          metadata: { genre: 'drama' },
      },
      {
          id: 'vec2', 
          values: [3.0, -2.0],
          metadata: { genre: 'documentary' },
      },
      {
          id: 'vec3', 
          values: [0.5, -1.5],
          metadata: { genre: 'documentary' },
      }
  ]);

  // Upsert records into the sparse index
  await sparseIndex.namespace('example-namespace').upsert([
      {
          id: 'vec1',
          sparseValues: {
              indices: [822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191],
              values: [1.7958984, 0.41577148, 2.828125, 2.8027344, 2.8691406, 1.6533203, 5.3671875, 1.3046875, 0.49780273, 0.5722656, 2.71875, 3.0820312, 2.5019531, 4.4414062, 3.3554688]
          },
          metadata: { 
              chunk_text: 'AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.', 
              category: 'technology',
              quarter: 'Q3' 
          }
      },
      {
          id: 'vec2',
          sparseValues: {
              indices: [131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697],
              values: [0.4362793, 3.3457031, 2.7714844, 3.0273438, 3.3164062, 5.6015625, 2.4863281, 0.38134766, 1.25, 2.9609375, 0.34179688, 1.4306641, 0.34375, 3.3613281, 1.4404297, 2.2558594, 2.2597656, 4.8710938, 0.5605469]
          },
          metadata: { 
              chunk_text: "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market.", 
              category: 'technology',
              quarter: 'Q4' 
          }
      },
      {
          id: 'vec3',
          sparseValues: {
              indices: [8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697],
              values: [2.6875, 4.2929688, 3.609375, 3.0722656, 2.1152344, 5.78125, 3.7460938, 3.7363281, 1.2695312, 3.4824219, 0.7207031, 0.0826416, 4.671875, 3.7011719, 2.796875, 0.61621094]
          },
          metadata: { 
              chunk_text: "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production", 
              category: 'technology',
              quarter: 'Q3' 
          }
      }
  ]);
    
  // Check the number of records in each index
  console.log('\nDense index stats:', await denseIndex.describeIndexStats());
  console.log('\nSparse index stats:', await sparseIndex.describeIndexStats());

  // Query the dense index with a metadata filter
  const denseQueryResponse = await denseIndex.namespace('example-namespace').query({
      vector: [3.0, -2.0],
      filter: {
          'genre': {'$eq': 'documentary'}
      },
      topK: 1,
      includeValues: false,
      includeMetadata: true,
  });

  console.log('\nDense query response:', denseQueryResponse);

  const sparseQueryResponse = await sparseIndex.namespace('example-namespace').query({
      sparseVector: {
          indices: [767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697],
          values: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
      },
      topK: 1,
      includeValues: false,
      includeMetadata: true
  });

  console.log('\nSparse query response:', sparseQueryResponse);

  // Delete the index
  await pc.deleteIndex(denseIndexName);
  await pc.deleteIndex(sparseIndexName);
  ```

  ```java Java theme={null}
  import com.google.protobuf.Struct;
  import com.google.protobuf.Value;
  import io.pinecone.clients.Index;
  import io.pinecone.clients.Pinecone;
  import io.pinecone.proto.DescribeIndexStatsResponse;
  import org.openapitools.db_control.client.model.DeletionProtection;
  import io.pinecone.unsigned_indices_model.QueryResponseWithUnsignedIndices;

  import java.util.*;

  public class PineconeLocalExample {
      public static void main(String[] args) {
          // Initialize a client.
          // API key is required, but the value does not matter.
          // When starting without indexes, disable TLS and
          // provide the host and port of the Pinecone Local instance.
          String host = "http://localhost:5080";
          Pinecone pc = new Pinecone.Builder("pclocal")
                  .withHost(host)
                  .withTlsEnabled(false)
                  .build();

          // Create two indexes, one dense and one sparse
          String denseIndexName = "dense-index";
          String sparseIndexName = "sparse-index";
          HashMap<String, String> tags = new HashMap<>();
          tags.put("environment", "development");
          pc.createServerlessIndex(
                  denseIndexName,
                  "cosine",
                  2,
                  "aws",
                  "us-east-1",
                  DeletionProtection.DISABLED,
                  tags
          );
          pc.createSparseServelessIndex(
                  sparseIndexName,
                  "aws",
                  "us-east-1",
                  DeletionProtection.DISABLED,
                  tags,
                  "sparse"
          );

          // Get index connection objects
          Index denseIndexConnection = pc.getIndexConnection(denseIndexName);
          Index sparseIndexConnection = pc.getIndexConnection(sparseIndexName);

          // Upsert records into the dense index
          Struct metaData1 = Struct.newBuilder()
                  .putFields("genre", Value.newBuilder().setStringValue("drama").build())
                  .build();
          Struct metaData2 = Struct.newBuilder()
                  .putFields("genre", Value.newBuilder().setStringValue("documentary").build())
                  .build();
          Struct metaData3 = Struct.newBuilder()
                  .putFields("genre", Value.newBuilder().setStringValue("documentary").build())
                  .build();

          denseIndexConnection.upsert("vec1", Arrays.asList(1.0f, -2.5f),  null, null, metaData1, "example-namespace");
          denseIndexConnection.upsert("vec2", Arrays.asList(3.0f, -2.0f),  null, null, metaData2, "example-namespace");
          denseIndexConnection.upsert("vec3", Arrays.asList(0.5f, -1.5f),  null, null, metaData3, "example-namespace");

          // Upsert records into the sparse index
          ArrayList<Long> indices1 = new ArrayList<>(Arrays.asList(
                  822745112L, 1009084850L, 1221765879L, 1408993854L, 1504846510L,
                  1596856843L, 1640781426L, 1656251611L, 1807131503L, 2543655733L,
                  2902766088L, 2909307736L, 3246437992L, 3517203014L, 3590924191L
          ));

          ArrayList<Float> values1 = new ArrayList<>(Arrays.asList(
                  1.7958984f, 0.41577148f, 2.828125f, 2.8027344f, 2.8691406f,
                  1.6533203f, 5.3671875f, 1.3046875f, 0.49780273f, 0.5722656f,
                  2.71875f, 3.0820312f, 2.5019531f, 4.4414062f, 3.3554688f
          ));

          Struct sparseMetaData1 = Struct.newBuilder()
                  .putFields("chunk_text", Value.newBuilder().setStringValue("AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.").build())
                  .putFields("category", Value.newBuilder().setStringValue("technology").build())
                  .putFields("quarter", Value.newBuilder().setStringValue("Q3").build())
                  .build();

          ArrayList<Long> indices2 = new ArrayList<>(Arrays.asList(
                  131900689L, 592326839L, 710158994L, 838729363L, 1304885087L,
                  1640781426L, 1690623792L, 1807131503L, 2066971792L, 2428553208L,
                  2548600401L, 2577534050L, 3162218338L, 3319279674L, 3343062801L,
                  3476647774L, 3485013322L, 3517203014L, 4283091697L
          ));

          ArrayList<Float> values2 = new ArrayList<>(Arrays.asList(
                  0.4362793f, 3.3457031f, 2.7714844f, 3.0273438f, 3.3164062f,
                  5.6015625f, 2.4863281f, 0.38134766f, 1.25f, 2.9609375f,
                  0.34179688f, 1.4306641f, 0.34375f, 3.3613281f, 1.4404297f,
                  2.2558594f, 2.2597656f, 4.8710938f, 0.5605469f
          ));

          Struct sparseMetaData2 = Struct.newBuilder()
                  .putFields("chunk_text", Value.newBuilder().setStringValue("Analysts suggest that AAPL'\\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.").build())
                  .putFields("category", Value.newBuilder().setStringValue("technology").build())
                  .putFields("quarter", Value.newBuilder().setStringValue("Q4").build())
                  .build();

          ArrayList<Long> indices3 = new ArrayList<>(Arrays.asList(
                  8661920L, 350356213L, 391213188L, 554637446L, 1024951234L,
                  1640781426L, 1780689102L, 1799010313L, 2194093370L, 2632344667L,
                  2641553256L, 2779594451L, 3517203014L, 3543799498L,
                  3837503950L, 4283091697L
          ));

          ArrayList<Float> values3 = new ArrayList<>(Arrays.asList(
                  2.6875f, 4.2929688f, 3.609375f, 3.0722656f, 2.1152344f,
                  5.78125f, 3.7460938f, 3.7363281f, 1.2695312f, 3.4824219f,
                  0.7207031f, 0.0826416f, 4.671875f, 3.7011719f, 2.796875f,
                  0.61621094f
          ));

          Struct sparseMetaData3 = Struct.newBuilder()
                  .putFields("chunk_text", Value.newBuilder().setStringValue("AAPL'\\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production").build())
                  .putFields("category", Value.newBuilder().setStringValue("technology").build())
                  .putFields("quarter", Value.newBuilder().setStringValue("Q3").build())
                  .build();

          sparseIndexConnection.upsert("vec1", Collections.emptyList(), indices1, values1, sparseMetaData1, "example-namespace");
          sparseIndexConnection.upsert("vec2", Collections.emptyList(), indices2, values2, sparseMetaData2, "example-namespace");
          sparseIndexConnection.upsert("vec3", Collections.emptyList(), indices3, values3, sparseMetaData3, "example-namespace");

          // Check the number of records each the index
          DescribeIndexStatsResponse denseIndexStatsResponse = denseIndexConnection.describeIndexStats(null);
          System.out.println("Dense index stats:");
          System.out.println(denseIndexStatsResponse);
          DescribeIndexStatsResponse sparseIndexStatsResponse = sparseIndexConnection.describeIndexStats(null);
          System.out.println("Sparse index stats:");
          System.out.println(sparseIndexStatsResponse);

          // Query the dense index with a metadata filter
          List<Float> queryVector = Arrays.asList(1.0f, 1.5f);
          QueryResponseWithUnsignedIndices denseQueryResponse = denseIndexConnection.query(1, queryVector, null, null, null, "example-namespace", null, false, true);
          System.out.println("Dense query response:");
          System.out.println(denseQueryResponse);

          // Query the sparse index with a metadata filter
          List<Long> sparseIndices = Arrays.asList(
                  767227209L, 1640781426L, 1690623792L, 2021799277L, 2152645940L,
                  2295025838L, 2443437770L, 2779594451L, 2956155693L, 3476647774L,
                  3818127854L, 428309169L);
          List<Float> sparseValues = Arrays.asList(
                  1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f,
                  1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f);

          QueryResponseWithUnsignedIndices sparseQueryResponse = sparseIndexConnection.query(1, null, sparseIndices, sparseValues, null, "example-namespace", null, false, true);
          System.out.println("Sparse query response:");
          System.out.println(sparseQueryResponse);

          // Delete the indexes
          pc.deleteIndex(denseIndexName);
          pc.deleteIndex(sparseIndexName);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  	"google.golang.org/protobuf/types/known/structpb"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
  	ctx := context.Background()

  	// Initialize a client.
  	// No API key is required.
  	// Host and port of the Pinecone Local instance
  	// is required when starting without indexes.
  	pc, err := pinecone.NewClientBase(pinecone.NewClientBaseParams{
  		Host: "http://localhost:5080",
  	})
  	if err != nil {
  		log.Fatalf("Failed to create Client: %v", err)
  	}

  	// Create two indexes, one dense and one sparse
  	denseIndexName := "dense-index"
  	denseVectorType := "dense"
  	dimension := int32(2)
  	denseMetric := pinecone.Cosine
  	deletionProtection := pinecone.DeletionProtectionDisabled
  	denseIdx, err := pc.CreateServerlessIndex(ctx, &pinecone.CreateServerlessIndexRequest{
  		Name:               denseIndexName,
  		VectorType:         &denseVectorType,
  		Dimension:          &dimension,
  		Metric:             &denseMetric,
  		Cloud:              pinecone.Aws,
  		Region:             "us-east-1",
  		DeletionProtection: &deletionProtection,
  		Tags:               &pinecone.IndexTags{"environment": "development"},
  	})
  	if err != nil {
  		log.Fatalf("Failed to create serverless index: %v", denseIdx.Name)
  	} else {
  		fmt.Printf("Successfully created serverless index: %v\n", denseIdx.Name)
  	}

  	sparseIndexName := "sparse-index"
  	sparseVectorType := "sparse"
  	sparseMetric := pinecone.Dotproduct
  	sparseIdx, err := pc.CreateServerlessIndex(ctx, &pinecone.CreateServerlessIndexRequest{
  		Name:               sparseIndexName,
  		VectorType:         &sparseVectorType,
  		Metric:             &sparseMetric,
  		Cloud:              pinecone.Aws,
  		Region:             "us-east-1",
  		DeletionProtection: &deletionProtection,
  		Tags:               &pinecone.IndexTags{"environment": "development"},
  	})
  	if err != nil {
  		log.Fatalf("Failed to create serverless index: %v", sparseIdx.Name)
  	} else {
  		fmt.Printf("\nSuccessfully created serverless index: %v\n", sparseIdx.Name)
  	}

  	// Get the index hosts
  	denseIdxModel, err := pc.DescribeIndex(ctx, denseIndexName)
  	if err != nil {
  		log.Fatalf("Failed to describe index \"%v\": %v", denseIndexName, err)
  	}

  	sparseIdxModel, err := pc.DescribeIndex(ctx, sparseIndexName)
  	if err != nil {
  		log.Fatalf("Failed to describe index \"%v\": %v", sparseIndexName, err)
  	}

  	// Target the indexes.
  	// Make sure to prefix the hosts with http:// to let the SDK know to disable tls.
  	denseIdxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "http://" + denseIdxModel.Host, Namespace: "example-namespace"})
  	if err != nil {
  		log.Fatalf("Failed to create IndexConnection for Host: %v", err)
  	}

  	sparseIdxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "http://" + sparseIdxModel.Host, Namespace: "example-namespace"})
  	if err != nil {
  		log.Fatalf("Failed to create IndexConnection for Host: %v", err)
  	}

  	// Upsert records into the dense index
  	denseMetadataMap1 := map[string]interface{}{
  		"genre": "drama",
  	}

  	denseMetadata1, err := structpb.NewStruct(denseMetadataMap1)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	denseMetadataMap2 := map[string]interface{}{
  		"genre": "documentary",
  	}

  	denseMetadata2, err := structpb.NewStruct(denseMetadataMap2)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	denseMetadataMap3 := map[string]interface{}{
  		"genre": "documentary",
  	}

  	denseMetadata3, err := structpb.NewStruct(denseMetadataMap3)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	denseVectors := []*pinecone.Vector{
  		{
  			Id:       "vec1",
  			Values:   &[]float32{1.0, -2.5},
  			Metadata: denseMetadata1,
  		},
  		{
  			Id:       "vec2",
  			Values:   &[]float32{3.0, -2.0},
  			Metadata: denseMetadata2,
  		},
  		{
  			Id:       "vec3",
  			Values:   &[]float32{0.5, -1.5},
  			Metadata: denseMetadata3,
  		},
  	}

  	denseCount, err := denseIdxConnection.UpsertVectors(ctx, denseVectors)
  	if err != nil {
  		log.Fatalf("Failed to upsert vectors: %v", err)
  	} else {
  		fmt.Printf("\nSuccessfully upserted %d vector(s)!\n", denseCount)
  	}

  	// Upsert records into the sparse index
  	sparseValues1 := pinecone.SparseValues{
  		Indices: []uint32{822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191},
  		Values:  []float32{1.7958984, 0.41577148, 2.828125, 2.8027344, 2.8691406, 1.6533203, 5.3671875, 1.3046875, 0.49780273, 0.5722656, 2.71875, 3.0820312, 2.5019531, 4.4414062, 3.3554688},
  	}

  	sparseMetadataMap1 := map[string]interface{}{
  		"chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones",
  		"category":   "technology",
  		"quarter":    "Q3",
  	}

  	sparseMetadata1, err := structpb.NewStruct(sparseMetadataMap1)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	sparseValues2 := pinecone.SparseValues{
  		Indices: []uint32{131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697},
  		Values:  []float32{0.4362793, 3.3457031, 2.7714844, 3.0273438, 3.3164062, 5.6015625, 2.4863281, 0.38134766, 1.25, 2.9609375, 0.34179688, 1.4306641, 0.34375, 3.3613281, 1.4404297, 2.2558594, 2.2597656, 4.8710938, 0.560546},
  	}

  	sparseMetadataMap2 := map[string]interface{}{
  		"chunk_text": "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
  		"category":   "technology",
  		"quarter":    "Q4",
  	}

  	sparseMetadata2, err := structpb.NewStruct(sparseMetadataMap2)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	sparseValues3 := pinecone.SparseValues{
  		Indices: []uint32{8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697},
  		Values:  []float32{2.6875, 4.2929688, 3.609375, 3.0722656, 2.1152344, 5.78125, 3.7460938, 3.7363281, 1.2695312, 3.4824219, 0.7207031, 0.0826416, 4.671875, 3.7011719, 2.796875, 0.61621094},
  	}

  	sparseMetadataMap3 := map[string]interface{}{
  		"chunk_text": "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production",
  		"category":   "technology",
  		"quarter":    "Q3",
  	}

  	sparseMetadata3, err := structpb.NewStruct(sparseMetadataMap3)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	sparseVectors := []*pinecone.Vector{
  		{
  			Id:           "vec1",
  			SparseValues: &sparseValues1,
  			Metadata:     sparseMetadata1,
  		},
  		{
  			Id:           "vec2",
  			SparseValues: &sparseValues2,
  			Metadata:     sparseMetadata2,
  		},
  		{
  			Id:           "vec3",
  			SparseValues: &sparseValues3,
  			Metadata:     sparseMetadata3,
  		},
  	}

  	sparseCount, err := sparseIdxConnection.UpsertVectors(ctx, sparseVectors)
  	if err != nil {
  		log.Fatalf("Failed to upsert vectors: %v", err)
  	} else {
  		fmt.Printf("\nSuccessfully upserted %d vector(s)!\n", sparseCount)
  	}

  	// Check the number of records in each index
  	denseStats, err := denseIdxConnection.DescribeIndexStats(ctx)
  	if err != nil {
  		log.Fatalf("Failed to describe index: %v", err)
  	} else {
  		fmt.Printf("\nDense index stats: %+v\n", prettifyStruct(*denseStats))
  	}

  	sparseStats, err := sparseIdxConnection.DescribeIndexStats(ctx)
  	if err != nil {
  		log.Fatalf("Failed to describe index: %v", err)
  	} else {
  		fmt.Printf("\nSparse index stats: %+v\n", prettifyStruct(*sparseStats))
  	}

  	// Query the dense index with a metadata filter
  	queryVector := []float32{3.0, -2.0}

  	queryMetadataMap := map[string]interface{}{
  		"genre": map[string]interface{}{
  			"$eq": "documentary",
  		},
  	}

  	metadataFilter, err := structpb.NewStruct(queryMetadataMap)
  	if err != nil {
  		log.Fatalf("Failed to create metadata map: %v", err)
  	}

  	denseRes, err := denseIdxConnection.QueryByVectorValues(ctx, &pinecone.QueryByVectorValuesRequest{
  		Vector:          queryVector,
  		TopK:            1,
  		MetadataFilter:  metadataFilter,
  		IncludeValues:   false,
  		IncludeMetadata: true,
  	})
  	if err != nil {
  		log.Fatalf("Error encountered when querying by vector: %v", err)
  	} else {
  		fmt.Printf("\nDense query response: %v\n", prettifyStruct(denseRes))
  	}

  	// Query the sparse index with a metadata filter

  	sparseValues := pinecone.SparseValues{
  		Indices: []uint32{767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697},
  		Values:  []float32{1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0},
  	}

  	sparseRes, err := sparseIdxConnection.QueryByVectorValues(ctx, &pinecone.QueryByVectorValuesRequest{
  		SparseValues:    &sparseValues,
  		TopK:            1,
  		IncludeValues:   false,
  		IncludeMetadata: true,
  	})
  	if err != nil {
  		log.Fatalf("Error encountered when querying by vector: %v", err)
  	} else {
  		fmt.Printf("\nSparse query response: %v\n", prettifyStruct(sparseRes))
  	}
  	// Delete the indexes
  	err = pc.DeleteIndex(ctx, denseIndexName)
  	if err != nil {
  		log.Fatalf("Failed to delete index: %v", err)
  	} else {
  		fmt.Printf("\nIndex \"%v\" deleted successfully\n", denseIndexName)
  	}

  	err = pc.DeleteIndex(ctx, sparseIndexName)
  	if err != nil {
  		log.Fatalf("Failed to delete index: %v", err)
  	} else {
  		fmt.Printf("\nIndex \"%v\" deleted successfully\n", sparseIndexName)
  	}
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  // Initialize a client.
  // API key is required, but the value does not matter.
  // When starting without indexes, disable TLS and
  // provide the host and port of the Pinecone Local instance.
  var pc = new PineconeClient("pclocal",
      new ClientOptions
      {
          BaseUrl = "http://localhost:5080",
          IsTlsEnabled = false
      }
  );

  // Create two indexes, one dense and one sparse
  var denseIndexName = "dense-index";
  var sparseIndexName = "sparse-index";

  var createDenseIndexRequest = await pc.CreateIndexAsync(new CreateIndexRequest
  {
      Name = denseIndexName,
      VectorType = VectorType.Dense,
      Dimension = 2,
      Metric = MetricType.Cosine,
      Spec = new ServerlessIndexSpec
      {
          Serverless = new ServerlessSpec
          {
              Cloud = ServerlessSpecCloud.Aws,
              Region = "us-east-1"
          }
      },
      DeletionProtection = DeletionProtection.Disabled,
      Tags = new Dictionary<string, string> 
      {  
          { "environment", "development" }
      }
  });

  Console.WriteLine("Dense index model:" + createDenseIndexRequest);

  var createSparseIndexRequest = await pc.CreateIndexAsync(new CreateIndexRequest
  {
      Name = sparseIndexName,
      VectorType = VectorType.Sparse,
      Metric = MetricType.Dotproduct,
      Spec = new ServerlessIndexSpec
      {
          Serverless = new ServerlessSpec
          {
              Cloud = ServerlessSpecCloud.Aws,
              Region = "us-east-1"
          }
      },
      DeletionProtection = DeletionProtection.Disabled,
      Tags = new Dictionary<string, string> 
      {  
          { "environment", "development" }
      }
  });

  Console.WriteLine("\nSparse index model:" + createSparseIndexRequest);

  // Target the indexes
  var denseIndex = pc.Index(denseIndexName);
  var sparseIndex = pc.Index(sparseIndexName);

  // Upsert records into the dense index
  var denseUpsertResponse = await denseIndex.UpsertAsync(new UpsertRequest()
  {
      Namespace = "example-namespace",
      Vectors = new List<Vector>
      {
          new Vector 
          { 
              Id = "vec1", 
              Values = new ReadOnlyMemory<float>([1.0f, -2.5f]),
              Metadata = new Metadata {
                  ["genre"] = new("drama"),
              },
          },
          new Vector 
          { 
              Id = "vec2", 
              Values = new ReadOnlyMemory<float>([3.0f, -2.0f]), 
              Metadata = new Metadata {
                  ["genre"] = new("documentary"),
              },
          },
          new Vector 
          { 
              Id = "vec3", 
              Values = new ReadOnlyMemory<float>([0.5f, -1.5f]),
              Metadata = new Metadata {
                  ["genre"] = new("documentary"),
              } 
          }
      }
  });
  Console.WriteLine($"\nUpserted {denseUpsertResponse.UpsertedCount} dense vectors");

  // Upsert records into the sparse index
  var sparseVector1 = new Vector
  {
      Id = "vec1",
      SparseValues = new SparseValues
      {
          Indices = new uint[] { 822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191 },
          Values = new ReadOnlyMemory<float>([1.7958984f, 0.41577148f, 2.828125f, 2.8027344f, 2.8691406f, 1.6533203f, 5.3671875f, 1.3046875f, 0.49780273f, 0.5722656f, 2.71875f, 3.0820312f, 2.5019531f, 4.4414062f, 3.3554688f])
      },
      Metadata = new Metadata {
          ["chunk_text"] = new("AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones."),
          ["category"] = new("technology"),
          ["quarter"] = new("Q3"),
      },
  };

  var sparseVector2 = new Vector
  {
      Id = "vec2",
      SparseValues = new SparseValues
      {
          Indices = new uint[] { 131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697 },
          Values = new ReadOnlyMemory<float>([0.4362793f, 3.3457031f, 2.7714844f, 3.0273438f, 3.3164062f, 5.6015625f, 2.4863281f, 0.38134766f, 1.25f, 2.9609375f, 0.34179688f, 1.4306641f, 0.34375f, 3.3613281f, 1.4404297f, 2.2558594f, 2.2597656f, 4.8710938f, 0.5605469f])
      },
      Metadata = new Metadata {
          ["chunk_text"] = new("Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market."),
          ["category"] = new("technology"),
          ["quarter"] = new("Q4"),
      },
  };

  var sparseVector3 = new Vector
  {
      Id = "vec3",
      SparseValues = new SparseValues
      {
          Indices = new uint[] { 8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697 },
          Values = new ReadOnlyMemory<float>([2.6875f, 4.2929688f, 3.609375f, 3.0722656f, 2.1152344f, 5.78125f, 3.7460938f, 3.7363281f, 1.2695312f, 3.4824219f, 0.7207031f, 0.0826416f, 4.671875f, 3.7011719f, 2.796875f, 0.61621094f])
      },
      Metadata = new Metadata {
          ["chunk_text"] = new("AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production"),
          ["category"] = new("technology"),
          ["quarter"] = new("Q3"),
      },    
  };

  var sparseUpsertResponse = await sparseIndex.UpsertAsync(new UpsertRequest
  {
      Vectors = new List<Vector> { sparseVector1, sparseVector2, sparseVector3 },
      Namespace = "example-namespace"
  });
  Console.WriteLine($"\nUpserted {sparseUpsertResponse.UpsertedCount} sparse vectors");

  // Check the number of records in each index
  var denseIndexStatsResponse = await denseIndex.DescribeIndexStatsAsync(new DescribeIndexStatsRequest());
  Console.WriteLine("\nDense index stats:" + denseIndexStatsResponse);
  var sparseIndexStatsResponse = await sparseIndex.DescribeIndexStatsAsync(new DescribeIndexStatsRequest());
  Console.WriteLine("\nSparse index stats:" + sparseIndexStatsResponse);

  // Query the dense index with a metadata filter
  var denseQueryResponse = await denseIndex.QueryAsync(new QueryRequest
  {
      Vector = new ReadOnlyMemory<float>([3.0f, -2.0f]),
      TopK = 1,
      Namespace = "example-namespace",
      Filter = new Metadata
      {
          ["genre"] = new Metadata
          {
              ["$eq"] = "documentary",
          }
      },
      IncludeValues = false,
      IncludeMetadata = true
  });
  Console.WriteLine("\nDense query response:" + denseQueryResponse);

  // Query the sparse index with a metadata filter
  var sparseQueryResponse = await sparseIndex.QueryAsync(new QueryRequest {
      Namespace = "example-namespace",
      TopK = 1,
      SparseVector = new SparseValues
      {
          Indices = [767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697],
          Values = new[] { 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f },
      },
      Filter = new Metadata
      {
          ["quarter"] = new Metadata
          {
              ["$eq"] = "Q4",
          }
      },
      IncludeValues = false,
      IncludeMetadata = true
  });
  Console.WriteLine("\nSparse query response:" + sparseQueryResponse);

  // Delete the indexes
  await pc.DeleteIndexAsync(denseIndexName);
  await pc.DeleteIndexAsync(sparseIndexName);
  ```

  ```shell curl theme={null}
  PINECONE_LOCAL_HOST="localhost:5080"
  DENSE_INDEX_HOST="localhost:5081"
  SPARSE_INDEX_HOST="localhost:5082"

  # Create two indexes, one dense and one sparse
  curl -X POST "http://$PINECONE_LOCAL_HOST/indexes" \
      -H "Accept: application/json" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "name": "dense-index",
              "vector_type": "dense",
              "dimension": 2,
              "metric": "cosine",
              "spec": {
                  "serverless": {
                      "cloud": "aws",
                      "region": "us-east-1"
                  }
              },
              "tags": {
                  "environment": "development"
              },
              "deletion_protection": "disabled"
          }'

  curl -X POST "http://$PINECONE_LOCAL_HOST/indexes" \
      -H "Accept: application/json" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "name": "sparse-index",
              "vector_type": "sparse",
              "metric": "dotproduct",
              "spec": {
                  "serverless": {
                      "cloud": "aws",
                      "region": "us-east-1"
                  }
              },
              "tags": {
                  "environment": "development"
              },
              "deletion_protection": "disabled"
          }'

  # Upsert records into the dense index
  curl -X POST "http://$DENSE_INDEX_HOST/vectors/upsert" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "namespace": "example-namespace",
              "vectors": [
                  {
                      "id": "vec1", 
                      "values": [1.0, -2.5],
                      "metadata": {"genre": "drama"}
                  },
                  {
                      "id": "vec2", 
                      "values": [3.0, -2.0],
                      "metadata": {"genre": "documentary"}
                  },
                  {
                      "id": "vec3", 
                      "values": [0.5, -1.5],
                      "metadata": {"genre": "documentary"}
                  }
              ]
          }'

  # Upsert records into the sparse index
  curl -X POST "http://$SPARSE_INDEX_HOST/vectors/upsert" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "namespace": "example-namespace",
              "vectors": [
                  {
                      "id": "vec1",
                      "sparseValues": {
                          "values": [1.7958984, 0.41577148, 2.828125, 2.8027344, 2.8691406, 1.6533203, 5.3671875, 1.3046875, 0.49780273, 0.5722656, 2.71875, 3.0820312, 2.5019531, 4.4414062, 3.3554688],
                          "indices": [822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191]
                      },
                      "metadata": {
                          "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
                          "category": "technology",
                          "quarter": "Q3"
                      }
                  },
                  {
                      "id": "vec2",
                      "sparseValues": {
                          "values": [0.4362793, 3.3457031, 2.7714844, 3.0273438, 3.3164062, 5.6015625, 2.4863281, 0.38134766, 1.25, 2.9609375, 0.34179688, 1.4306641, 0.34375, 3.3613281, 1.4404297, 2.2558594, 2.2597656, 4.8710938, 0.5605469],
                          "indices": [131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697]
                      },
                      "metadata": {
                          "chunk_text": "Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
                          "category": "technology",
                          "quarter": "Q4"
                      }
                  },
                  {
                      "id": "vec3",
                      "sparseValues": {
                          "values": [2.6875, 4.2929688, 3.609375, 3.0722656, 2.1152344, 5.78125, 3.7460938, 3.7363281, 1.2695312, 3.4824219, 0.7207031, 0.0826416, 4.671875, 3.7011719, 2.796875, 0.61621094],
                          "indices": [8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697]
                      },
                      "metadata": {
                          "chunk_text": "AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production",
                          "category": "technology",
                          "quarter": "Q3"
                      }
                  }
              ]
          }'

  # Check the number of records in each index
  curl -X POST "http://$DENSE_INDEX_HOST/describe_index_stats" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{}'

  curl -X POST "http://$SPARSE_INDEX_HOST/describe_index_stats" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{}'

  # Query the dense index with a metadata filter
  curl "http://$DENSE_INDEX_HOST/query" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "vector": [3.0, -2.0],
              "filter": {"genre": {"$eq": "documentary"}},
              "topK": 1,
              "includeMetadata": true,
              "includeValues": false,
              "namespace": "example-namespace"
          }'

  # Query the sparse index with a metadata filter
  curl "http://$SPARSE_INDEX_HOST/query" \
      -H "Content-Type: application/json" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "sparseVector": {
                  "values": [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],
                  "indices": [767227209, 1640781426, 1690623792, 2021799277, 2152645940, 2295025838, 2443437770, 2779594451, 2956155693, 3476647774, 3818127854, 4283091697]
              },
              "filter": {"quarter": {"$eq": "Q4"}},
              "namespace": "example-namespace",
              "topK": 1,
              "includeMetadata": true,
              "includeValues": false
          }'

  # Delete the index
  curl -X DELETE "http://$PINECONE_LOCAL_HOST/indexes/dense-index" \
      -H "X-Pinecone-Api-Version: 2025-10"

  curl -X DELETE "http://$PINECONE_LOCAL_HOST/indexes/sparse-index" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

## 3. Stop Pinecone Local

<Warning>
  Pinecone Local is an in-memory emulator. Records loaded into Pinecone Local do not persist after Pinecone Local is stopped.
</Warning>

To stop and remove the resources for Pinecone Local, run the following command:

<CodeGroup>
  ```shell Docker Compose theme={null}
  docker compose down
  ```

  ```shell Docker CLI theme={null}
  # If you started Pinecone Local with indexes:
  docker stop dense-index sparse-index
  docker rm dense-index sparse-index

  # If you started Pinecone Local without indexes:
  docker stop pinecone-local
  docker rm pinecone-local
  ```
</CodeGroup>

## Moving from Pinecone Local to your Pinecone account

When you're ready to run your application against your Pinecone account, be sure to do the following:

* Update your application to [use your Pinecone API key](/reference/api/authentication).
* Update your application to [target your Pinecone indexes](/guides/manage-data/target-an-index).
* [Use Pinecone's import feature](/guides/index-data/import-data) to efficiently load large amounts of data into your indexes and then [use batch upserts](/guides/index-data/upsert-data#upsert-in-batches) for ongoing writes.
* Follow Pinecone's [production best practices](/guides/production/production-checklist).

---

# Source: https://docs.pinecone.io/troubleshooting/login-code-issues.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Login code issues

If the email token you received from Pinecone is not accepted when logging in there may be a few different reasons why:

## The code has expired

A code is only valid for 10 hours, so if you enter it after that time it will no longer be accepted.

## The code has already been used

If you're using a shared email account or distribution list, please check with your teammates to see if anyone else has used your code.

## A subsequent request was made

Similar to the first reason, if you're using a shared account, it's possible that someone else requested a code after you did, rendering the first code invalid.

## Your computer's system clock time is offset

User authentication with a verification code relies on your device’s system clock to verify the time. If your computer’s clock is more than 10 minutes off from your time zone, the login will fail. If you see the below error message, please set your system clock to the correct time and time zone before trying again.

```
Please check your computer's system clock time. See https://docs.pinecone.io/troubleshooting/login-code-issues for more information.
```

## Your anti-spam filter followed the links in the email to check their validity

If your anti-spam filter followed the links in the email to check their validity, and one of them submitted the code as part of the URL, please check with your anti-spam system admin or vendor to see if this might be the cause.

---

# Source: https://docs.pinecone.io/guides/projects/manage-api-keys.md

# Source: https://docs.pinecone.io/guides/assistant/admin/manage-api-keys.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage API keys

> Create and manage API keys with custom permissions.

Each Pinecone [project](/guides/projects/understanding-projects) has one or more API keys. In order to [make calls to the Pinecone API](/guides/get-started/quickstart), you must provide a valid API key for the relevant Pinecone project.

This page shows you how to [create](#create-an-api-key), [view](#view-api-keys), [change permissions for](#change-api-key-permissions), and [delete](#delete-an-api-key) API keys.

<Warning>
  If you use custom API key permissions, ensure that you [target your index by host](/guides/manage-data/target-an-index#target-by-index-host-recommended) when performing data operations such as `upsert` and `query`.
</Warning>

## Create an API key

You can create a new API key for your project, as follows:

<Tabs>
  <Tab title="Pinecone console">
    1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).

    2. Select your project.

    3. Go to **API keys**.

    4. Click **Create API key**.

    5. Enter an **API key name**.

    6. Select the **Permissions** to grant to the API key. For a description of the permission roles, see [API key permissions](/guides/production/security-overview#api-keys).

       <Note>
         Users on the Starter plan can set the permissions to **All** only. To customize the permissions further, [upgrade to the Standard or Enterprise plan](/guides/organizations/manage-billing/upgrade-billing-plan).
       </Note>

    7. Click **Create key**.

    8. Copy and save the generated API key in a secure place for future use.

       <Warning>
         You will not be able to see the API key again after you close the dialog.
       </Warning>

    9. Click **Close**.
  </Tab>

  <Tab title="Code">
    <Note>
      An [access token](/guides/organizations/manage-service-accounts#retrieve-an-access-token) must be provided to complete this action through the Admin API. The Admin API is in [public preview](/release-notes/feature-availability).
    </Note>

    <CodeGroup>
      ```bash curl theme={null}
      PINECONE_PROJECT_ID="YOUR_PROJECT_ID"
      PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"

      curl -X POST "https://api.pinecone.io/admin/projects/$PINECONE_PROJECT_ID/api-keys" \
           -H "X-Pinecone-Api-Version: 2025-10" \
           -H "Authorization: Bearer $PINECONE_ACCESS_TOKEN" \
           -d '{
                 "name": "example-api-key",
                 "roles": ["ProjectEditor"]
               }'
      ```

      ```bash CLI theme={null}
      # Target the project for which you want to create an API key.
      pc target -o "example-org" -p "example-project"
      # Create the API key
      pc api-key create -n "example-api-key" --roles ProjectEditor
      ```
    </CodeGroup>

    The example returns a response like the following:

    <CodeGroup>
      ```json curl theme={null}
      {
        "key": {
          "id": "62b0dbfe-3489-4b79-b850-34d911527c88",
          "name": "example-api-key",
          "project_id": "32c8235a-5220-4a80-a9f1-69c24109e6f2",
          "roles": [
            "ProjectEditor"
          ],
          "created_at": "2025-10-20T23:40:27.069075Z"
        },
        "value": "..."
      }
      ```

      ```text CLI theme={null}
      ATTRIBUTE     VALUE
      Name          example-api-key
      ID            62b0dbfe-3489-4b79-b850-34d911527c88
      Value         ...
      Project ID    32c8235a-5220-4a80-a9f1-69c24109e6f2
      Roles         ProjectEditor
      ```
    </CodeGroup>
  </Tab>
</Tabs>

## View project API keys

You can [view the API keys](/reference/api/latest/admin/list_api_keys) for your project:

<Tabs>
  <Tab title="Pinecone console">
    1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).
    2. Select your project.
    3. Go to the **API keys** tab.

    You will see a list of all API keys for the project, including their names, IDs, and permissions.
  </Tab>

  <Tab title="Code">
    <Note>
      An [access token](/guides/organizations/manage-service-accounts#retrieve-an-access-token) must be provided to complete this action through the Admin API. The Admin API is in [public preview](/release-notes/feature-availability).
    </Note>

    <CodeGroup>
      ```bash curl theme={null}
      PINECONE_PROJECT_ID="32c8235a-5220-4a80-a9f1-69c24109e6f2"
      PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"

      curl -X GET "https://api.pinecone.io/admin/projects/$PINECONE_PROJECT_ID/api-keys" \
           -H "Authorization: Bearer $PINECONE_ACCESS_TOKEN" \
           -H "X-Pinecone-Api-Version: 2025-10"
      ```

      ```bash CLI theme={null}
      PINECONE_PROJECT_ID="32c8235a-5220-4a80-a9f1-69c24109e6f2"

      pc api-key list -i $PINECONE_PROJECT_ID
      ```
    </CodeGroup>

    The example returns a response like the following:

    <CodeGroup>
      ```json curl theme={null}
      {
        "data": [
          {
            "id": "62b0dbfe-3489-4b79-b850-34d911527c88",
            "name": "example-api-key",
            "project_id": "32c8235a-5220-4a80-a9f1-69c24109e6f2",
            "roles": [
              "ProjectEditor"
            ],
            "created_at": "2025-10-20T23:39:43.665754Z"
          },
          {
            "id": "0d0d3678-81b4-4e0d-a4f0-70ba488acfb7",
            "name": "example-api-key-2",
            "project_id": "32c8235a-5220-4a80-a9f1-69c24109e6f2",
            "roles": [
              "ProjectEditor"
            ],
            "created_at": "2025-10-20T23:43:13.176422Z"
          }
        ]
      }
      ```

      ```text CLI theme={null}
      Organization: example-organization (ID: -NM7af6f234168c4e44a)
      Project: example-project (ID: 32c8235a-5220-4a80-a9f1-69c24109e6f2)

      API Keys

      NAME                 ID                                      PROJECT ID                              ROLES
      example-api-key      62b0dbfe-3489-4b79-b850-34d911527c88    32c8235a-5220-4a80-a9f1-69c24109e6f2    ProjectEditor
      example-api-key-2    0d0d3678-81b4-4e0d-a4f0-70ba488acfb7    32c8235a-5220-4a80-a9f1-69c24109e6f2    ProjectEditor
      ```
    </CodeGroup>
  </Tab>
</Tabs>

## View API key details

You can [view the details of an API key](/reference/api/latest/admin/fetch_api_key):

<Tabs>
  <Tab title="Pinecone console">
    1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).
    2. Select your project.
    3. Go to the **API keys** tab.
    4. In the row of the API key you want to change, in the **Actions** column, click **ellipsis (...) menu > Settings**.

    You will see the API key's name, ID, and permissions.
  </Tab>

  <Tab title="Code">
    <Note>
      An [access token](/guides/organizations/manage-service-accounts#retrieve-an-access-token) must be provided to complete this action through the Admin API. The Admin API is in [public preview](/release-notes/feature-availability).
    </Note>

    <CodeGroup>
      ```bash curl theme={null}
      PINECONE_API_KEY_ID="62b0dbfe-3489-4b79-b850-34d911527c88"
      PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"

      curl -X GET "https://api.pinecone.io/admin/api-keys/$PINECONE_API_KEY_ID" \
           -H "Authorization: Bearer $PINECONE_ACCESS_TOKEN" \
           -H "accept: application/json" \
           -H "X-Pinecone-Api-Version: 2025-10"
      ```

      ```bash CLI theme={null}
      PINECONE_API_KEY_ID="62b0dbfe-3489-4b79-b850-34d911527c88"

      pc api-key describe -i $PINECONE_API_KEY_ID
      ```
    </CodeGroup>

    The example returns a response like the following:

    <CodeGroup>
      ```json curl theme={null}
      {
        "id": "62b0dbfe-3489-4b79-b850-34d911527c88",
        "name": "example-api-key",
        "project_id": "32c8235a-5220-4a80-a9f1-69c24109e6f2",
        "roles": [
          "ProjectEditor"
        ],
        "created_at": "2025-10-22T19:27:21.202955Z"
      }
      ```

      ```text CLI theme={null}
      ATTRIBUTE     VALUE
      Name          example-api-key
      ID            62b0dbfe-3489-4b79-b850-34d911527c88
      Project ID    32c8235a-5220-4a80-a9f1-69c24109e6f2
      Roles         ProjectEditor
      ```
    </CodeGroup>
  </Tab>
</Tabs>

## Update an API key

<Note>
  Users on the Starter plan cannot change API key permissions once they are set. Instead, [create a new API key](#create-an-api-key) or [upgrade to the Standard or Enterprise plan](/guides/organizations/manage-billing/upgrade-billing-plan).
</Note>

If you are a [project owner](/guides/projects/understanding-projects#project-roles), you can update the name and roles of an API key:

<Tabs>
  <Tab title="Pinecone console">
    1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).

    2. Select your project.

    3. Go to the **API keys** tab.

    4. In the row of the API key you want to change, in the **Actions** column, click **ellipsis (...) menu > Settings**.

    5. Change the name and/or permissions for the API key as needed.

       For information about the different API key permissions, refer to [Understanding security - API keys](/guides/production/security-overview#api-keys).

    6. Click **Update**.
  </Tab>

  <Tab title="Code">
    <Note>
      An [access token](/guides/organizations/manage-service-accounts#retrieve-an-access-token) must be provided to complete this action through the Admin API. The Admin API is in [public preview](/release-notes/feature-availability).
    </Note>

    <CodeGroup>
      ```bash curl theme={null}
      PINECONE_API_KEY_ID="62b0dbfe-3489-4b79-b850-34d911527c88"
      PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"

      curl -X PATCH "https://api.pinecone.io/admin/api-keys/$PINECONE_API_KEY_ID" \
           -H "X-Pinecone-Api-Version: 2025-10" \
           -H "Authorization: Bearer $PINECONE_ACCESS_TOKEN" \
           -d '{
                 "name": "new-api-key-name",
                 "roles": ["ProjectEditor"]
               }'
      ```

      ```bash CLI theme={null}
      PINECONE_API_KEY_ID="62b0dbfe-3489-4b79-b850-34d911527c88"

      # Target the organization that contains the API key.
      pc target -o "example-org"
      # Update the API key name.
      pc api-key update -i $PINECONE_API_KEY_ID -n "new-api-key-name"
      ```
    </CodeGroup>

    The example returns a response like the following:

    <CodeGroup>
      ```json curl theme={null}
      {
        "id": "62b0dbfe-3489-4b79-b850-34d911527c88",
        "name": "new-api-key-name",
        "project_id": "32c8235a-5220-4a80-a9f1-69c24109e6f2",
        "roles": [
          "ProjectEditor"
        ],
        "created_at": "2025-10-22T19:27:21.202955Z"
      }
      ```

      ```text CLI theme={null}
      ATTRIBUTE     VALUE
      Name          new-api-key-name
      ID            62b0dbfe-3489-4b79-b850-34d911527c88
      Project ID    32c8235a-5220-4a80-a9f1-69c24109e6f2
      Roles         ProjectEditor
      ```
    </CodeGroup>
  </Tab>
</Tabs>

## Delete an API key

If you are a [project owner](/guides/projects/understanding-projects#project-roles), you can delete your API key:

<Tabs>
  <Tab title="Pinecone console">
    1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).
    2. Select your project.
    3. Go to the **API keys** tab.
    4. In the row of the API key you want to change, in the **Actions** column, click **ellipsis (...) menu > Delete**.
    5. Enter the **API key name**.
    6. Click **Confirm deletion**.

       <Warning>
         Deleting an API key is irreversible and will immediately disable any applications using the API key.
       </Warning>
  </Tab>

  <Tab title="Code">
    <Note>
      An [access token](/guides/organizations/manage-service-accounts#retrieve-an-access-token) must be provided to complete this action through the Admin API. The Admin API is in [public preview](/release-notes/feature-availability).
    </Note>

    <CodeGroup>
      ```bash curl theme={null}
      PINECONE_API_KEY_ID="62b0dbfe-3489-4b79-b850-34d911527c88"
      PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"

      curl -X DELETE "https://api.pinecone.io/admin/api-keys/$PINECONE_API_KEY_ID" \
           -H "X-Pinecone-Api-Version: 2025-10" \
           -H "Authorization: Bearer $PINECONE_ACCESS_TOKEN"
      ```

      ```bash CLI theme={null}
      PINECONE_API_KEY_ID="62b0dbfe-3489-4b79-b850-34d911527c88"

      # Delete the API key. Use --skip-confirmation to skip
      # the confirmation prompt.
      pc api-key delete -i $PINECONE_API_KEY_ID
      ```
    </CodeGroup>

    The example returns a response like the following:

    <CodeGroup>
      ```text curl theme={null}
      No response payload
      ```

      ```text CLI theme={null}
      [WARN] This operation will delete API key example-api-key from project example-project.
      [WARN] Any integrations that authenticate with this API key will immediately stop working.
      [WARN] This action cannot be undone.
      Do you want to continue? (y/N): y
      [INFO] You chose to continue delete.
      [SUCCESS] API key example-api-key deleted
      ```
    </CodeGroup>
  </Tab>
</Tabs>

---

# Source: https://docs.pinecone.io/guides/assistant/manage-assistants.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage assistants

> View, update, and delete, and check the status of assistants.

## List assistants for a project

You can [get the name, status, and metadata for each assistant](/reference/api/latest/assistant/list_assistants) in your project as in the following example:

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  pc = Pinecone(api_key="YOUR_API_KEY")

  assistants = pc.assistant.list_assistants()
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistants = await pc.listAssistants();
  console.log(assistants);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -X GET "https://api.pinecone.io/assistant/assistants" \
    -H "Api-Key: $PINECONE_API_KEY"
  ```
</CodeGroup>

This operation returns a response like the following:

```JSON  theme={null}
{
  "assistants": [
    {
      "name": "example-assistant",
      "instructions": "Use American English for spelling and grammar.",
      "metadata": {"team": "customer-support", "version": "1.0"},
      "status": "Initializing",
      "created_at": "2023-11-07T05:31:56Z",
      "updated_at": "2023-11-07T05:31:56Z"
    }
  ]
}
```

You can use the `name` value to [check the status of an assistant](/guides/assistant/manage-assistants#get-the-status-of-an-assistant).

<Tip>
  You can list assistants using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/assistant/-/files).
</Tip>

## Get the status of an assistant

You can [get the status and metadata for your assistant](/reference/api/latest/assistant/describe_assistant) as in the following example:

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  pc = Pinecone(api_key="YOUR_API_KEY")

  assistant = pc.assistant.describe_assistant(
      assistant_name="example-assistant", 
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistant = await pc.describeAssistant('example-assistant');
  console.log(assistant);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl -X GET "https://api.pinecone.io/assistant/assistants/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY"
  ```
</CodeGroup>

This operation returns a response like the following:

```JSON  theme={null}
{
  "name": "example-assistant",
  "instructions": "Use American English for spelling and grammar.",
  "metadata": {"team": "customer-support", "version": "1.0"},
  "status": "Initializing",
  "created_at": "2023-11-07T05:31:56Z",
  "updated_at": "2023-11-07T05:31:56Z"
}
```

The `status` field has the following possible values:

* Initializing
* Failed
* Ready
* Terminating

<Tip>
  You can check the status of an assistant using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/assistant).
</Tip>

## Change an assistant's chat model

The chat model is the underlying large language model (LLM) that powers the assistant's responses. You can change the chat model for an existing assistant through the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/assistant):

1. On the **Assistants** page, select the assistant you want to update.
2. In the sidebar on the right, select **Settings** (gear icon).
3. Select the **Chat model**.

## Add instructions to an assistant

You can [add or update the instructions](/reference/api/latest/assistant/update_assistant) for an existing assistant. Instructions are a short description or directive for the assistant to apply to all of its responses. For example, you can update the instructions to reflect the assistant's role or purpose.

<Note>
  Instructions (maximum size 16 KB) are included in every chat API call. Longer instructions increase input token costs for each request and consume more of the LLM's context window, reducing available space for retrieved context and conversation history.
</Note>

For example:

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone

  pc = Pinecone(api_key=YOUR_API_KEY)

  assistant = pc.assistant.update_assistant(
      assistant_name="example-assistant", 
      instructions="Use American English for spelling and grammar.",
      metadata={"team": "customer-support", "version": "1.1"} # Optional metadata (max 16KB) for organizing assistants.
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await pc.updateAssistant('example-assistant', {
    instructions: 'Use American English for spelling and grammar.',
    metadata: { team: 'customer-support', version: '1.1' }, // Optional metadata (max 16KB) for organizing assistants.
  });
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -X PATCH "https://api.pinecone.io/assistant/assistants/example-assistant" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
    "instructions": "Use American English for spelling and grammar.",
    "metadata": {"team": "customer-support", "version": "1.1"}
  }'
  ```
</CodeGroup>

The example above returns a result like the following:

```JSON  theme={null}
{
    "name":"example-assistant",
    "instructions":"Use American English for spelling and grammar.",
    "metadata":{"team": "customer-support", "version": "1.1"},
    "status":"Ready",
    "created_at":"2024-06-14T14:58:06.573004549Z",
    "updated_at":"2024-10-01T19:44:32.813235817Z"
}
```

<Tip>
  You can add or update instructions for an assistant using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/assistant).
</Tip>

## Delete an assistant

You can [delete an assistant](/reference/api/latest/assistant/delete_assistant) as in the following example:

<Warning>
  Deleting an assistant also deletes all files uploaded to the assistant.
</Warning>

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.assistant.delete_assistant(
      assistant_name="example-assistant", 
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await pc.deleteAssistant('example-assistant');
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl -X DELETE "https://api.pinecone.io/assistant/assistants/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY"
  ```
</CodeGroup>

<Tip>
  You can delete an assistant using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/assistant).
</Tip>

---

# Source: https://docs.pinecone.io/guides/manage-cost/manage-cost.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage cost

> Learn strategies for managing cost in Pinecone.

For the latest pricing details, see our [pricing page](https://www.pinecone.io/pricing/).

For help estimating total cost, see [Understanding cost](/guides/manage-cost/understanding-cost). To view or download a detailed report of your current usage and costs, see [Monitor usage and costs](/guides/manage-cost/monitor-usage-and-costs#monitor-organization-level-usage).

## Set monthly spend alerts

You can set up email alerts to monitor your organization's monthly spending. These alerts notify designated recipients when spending reaches specified thresholds. The alerts automatically reset at the start of each monthly billing cycle.

To set a spend alert:

1. Go to [Settings > Spend alerts](https://app.pinecone.io/organizations/-/settings/spend-alerts) in the Pinecone console
2. Click **+ Add Alert**.
3. Enter the dollar amount for the spend alert.
4. Enter the email addresses to send the alert to. [Organization owners](/guides/organizations/understanding-organizations#organization-roles) are listed by default.
5. Click **Create**.

To edit a spend alert:

1. In the row of the spend alert you want to edit, click **ellipsis (...) menu > Edit**.
2. Change the dollar amount and/or email addresses for the spend alert.
3. Click **Update**.

<Note>
  **Auto-spend spike alert**: To protect from unexpected cost increases, Pinecone sends an alert when spending exceeds double your previous month's invoice amount. While the alert threshold is fixed and the alert cannot be deleted, you can modify which email addresses receive the alert and enable or disable the alert notifications.
</Note>

## List by ID prefix

By using a [hierarchical ID schema](/guides/index-data/data-modeling#use-structured-ids), you can retrieve records without performing a query. To do so, you can use [`list`](/reference/api/latest/data-plane/list) to retrieve records by ID prefix, then use `fetch` to retrieve the records you need. This can reduce costs, because [`query` consumes more RUs when scanning a larger namespace](/guides/manage-cost/understanding-cost#query), while [`fetch` consumes a fixed ratio of RUs to records retrieved](/guides/manage-cost/understanding-cost#fetch).

## Use namespaces for multitenancy

If your application requires you to isolate the data of each customer/user, consider [implementing multitenancy with serverless indexes and namespaces](/guides/index-data/implement-multitenancy). With serverless indexes, you pay only for the amount of data stored and operations performed. For queries in particular, the cost is partly based on the total number of records that must be scanned, so using namespaces can significantly reduce query costs.

## Commit to annual spend

Pinecone offers a discount for customers who enroll in an annual plan with an upfront payment. Eligibility requires a minimum annual commitment of \$8,000.

Throughout the contract term, the discount applies to List Price for all Pinecone services, excluding Storage and Support. Any usage that exceeds the prepaid credits will be billed at full List Price and will not qualify for the contractual discount.

Customers on the Standard and Enterprise pay-as-you-go plans may upgrade directly by navigating in the Pinecone console to **[Settings > Billing > Plans](https://app.pinecone.io/organizations/-/settings/billing/plans)**.

<Note>
  The self-serve annual plan is not available through cloud marketplace billing. To purchase an annual plan through a cloud marketplace, contact [ar@pinecone.io](mailto:ar@pinecone.io).
</Note>

## Talk to support

Users on Standard and Enterprise plans can [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket) for help in optimizing costs.

## See also

* [Understanding cost](/guides/manage-cost/understanding-cost)
* [Monitor usage and costs](/guides/manage-cost/monitor-usage-and-costs)

---

# Source: https://docs.pinecone.io/guides/assistant/manage-files.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage files

> List, check status, and delete files from your assistant.

<Note>
  File upload limitations depend on the plan you are using. For more information, see [Pricing and limitations](/guides/assistant/pricing-and-limits#limits).
</Note>

## List files in an assistant

### View all files

You can [get the status, ID, and metadata for each file in your assistant](/reference/api/latest/assistant/list_files), as in the following example:

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  pc = Pinecone(api_key="YOUR_API_KEY")

  # Get your assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant", 
  )

  # List files in your assistant.
  files = assistant.list_files()
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const files = await assistant.listFiles();
  console.log(files);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl -X GET "https://prod-1-data.ke.pinecone.io/assistant/files/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY"
  ```
</CodeGroup>

This operation returns a response like the following:

```JSON  theme={null}
{
  "files": [
    {
      "status": "Available",
      "id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
      "name": "example_file.txt",
      "size": 1073470,
      "metadata": {},
      "updated_on": "2025-07-16T16:46:40.787204651Z",
      "created_on": "2025-07-16T16:45:59.414273474Z",
      "percent_done": 1.0,
      "signed_url": null,
      "error_message": null
    }
  ]
}
```

You can use the `id` value to [check the status of an individual file](#get-the-status-of-a-file).

<Tip>
  You can list file in an assistant using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/assistant). Select the assistant and view the files in the Assistant playground.
</Tip>

### View a filtered list of files

Metadata filter expressions can be included when listing files. This will limit the list of files to only those matching the filter expression. Use the `filter` parameter to specify the metadata filter expression.

For more information about filtering with metadata, see [Understanding files](/guides/assistant/files-overview#metadata-query-language).

The following example lists files that are a manuscript:

<CodeGroup>
  ```Python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  pc = Pinecone(api_key="YOUR_API_KEY")

  # Get your assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant", 
  )

  # List files in your assistant that match the metadata filter.
  files = assistant.list_files(filter={"document_type":"manuscript"})
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const files = await assistant.listFiles({
    filter: { document_type: 'manuscript' },
  });
  console.log(files);

  // You can also use filter operators:
  // const files = await assistant.listFiles({
  //   filter: { document_type: { '$ne': 'manuscript' } },
  // });
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"
  ENCODED_METADATA="%7B%22document_type%22%3A%20%22manuscript%22%7D" # URL encoded metadata - See w3schools.com/tags/ref_urlencode.ASP

  curl -X GET "https://prod-1-data.ke.pinecone.io/assistant/files/$ASSISTANT_NAME?filter=$ENCODED_METADATA" \
    -H "Api-Key: $PINECONE_API_KEY"
  ```
</CodeGroup>

## Get the status of a file

You can [get the status and metadata for your assistant](/reference/api/latest/assistant/describe_file), as in the following example:

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  pc = Pinecone(api_key="YOUR_API_KEY")

  # Get an assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant", 
  )

  # Describe a file. 
  # To get a signed URL in the response, set `include_url` to `True`.
  file = assistant.describe_file(file_id="3c90c3cc-0d44-4b50-8888-8dd25736052a", include_url=True)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const fileId = "3c90c3cc-0d44-4b50-8888-8dd25736052a";

  // Describe a file. Returns a signed URL by default. 
  const file = await assistant.describeFile(fileId)
  // To exclude signed URL, set `includeUrl` to `false`.
  // const includeUrl = false;
  // const file = await assistant.describeFile(fileId, includeUrl)
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"
  FILE_ID="3c90c3cc-0d44-4b50-8888-8dd25736052a"

  # Describe a file. 
  # To get a signed URL in the response, set `include_url` to `true`.
  curl -X GET "https://prod-1-data.ke.pinecone.io/assistant/files/$ASSISTANT_NAME/$FILE_ID?include_url=true" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

This operation returns a response like the following:

```JSON  theme={null}
{
  "status": "Available",
  "id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
  "name": "example_file.txt",
  "size": 1073470,
  "metadata": {},
  "updated_on": "2025-07-16T16:46:40.787204651Z",
  "created_on": "2025-07-16T16:45:59.414273474Z",
  "percent_done": 1.0,
  "signed_url": "https://storage.googleapis.com/...",
  "error_message": null
}                          
```

<Warning>
  [`signed_url`](https://cloud.google.com/storage/docs/access-control/signed-urls) provides temporary, read-only access to the relevant file. Anyone with the link can access the file, so treat it as sensitive data. Expires in one hour.
</Warning>

<Tip>
  You can check the status a file using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/assistant). In the Assistant playground, click the file for more details.
</Tip>

## Delete a file

You can [delete a file](/reference/api/latest/assistant/delete_file) from an assistant.

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  pc = Pinecone(api_key="YOUR_API_KEY")

  # Get your assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant", 
  )

  # Delete a file from your assistant.
  assistant.delete_file(file_id="3c90c3cc-0d44-4b50-8888-8dd25736052a")
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);

  const file = await assistant.deleteFile("070513b3-022f-4966-b583-a9b12e0290ff")
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"
  FILE_ID="3c90c3cc-0d44-4b50-8888-8dd25736052a"

  curl -X DELETE "https://prod-1-data.ke.pinecone.io/assistant/files/$ASSISTANT_NAME/$FILE_ID" \
    -H "Api-Key: $PINECONE_API_KEY"
  ```
</CodeGroup>

* Once a file is deleted, you cannot recover it.
* You can delete a file while it is still processing. You do not need to wait for processing to complete.
* You can also delete a file from an assistant using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/assistant). In the Assistant playground, find the file and click the **ellipsis (...) menu > Delete**.

---

# Source: https://docs.pinecone.io/guides/manage-data/manage-indexes.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage serverless indexes

> List, describe, and configure serverless indexes.

This page shows you how to manage your existing serverless indexes.

## List indexes

Use the [`list_indexes`](/reference/api/latest/control-plane/list_indexes) operation to get a complete description of all indexes in a project:

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  index_list = pc.list_indexes()

  print(index_list)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  const indexList = await pc.listIndexes();

  console.log(indexList);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.*;

  public class ListIndexesExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          IndexList indexList = pc.listIndexes();
          System.out.println(indexList);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
      bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idxs, err := pc.ListIndexes(ctx)
      if err != nil {
          log.Fatalf("Failed to list indexes: %v", err)
      } else {
          for _, index := range idxs {
              fmt.Printf("index: %v\n", prettifyStruct(index))
          }
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var indexList = await pinecone.ListIndexesAsync();

  Console.WriteLine(indexList);
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X GET "https://api.pinecone.io/indexes" \
  -H "Api-Key: $PINECONE_API_KEY" \
  -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

The response will look like this:

<CodeGroup>
  ```python Python theme={null}
  [{
      "name": "docs-example-sparse",
      "metric": "dotproduct",
      "host": "docs-example-sparse-govk0nt.svc.aped-4627-b74a.pinecone.io",
      "spec": {
          "serverless": {
              "cloud": "aws",
              "region": "us-east-1"
          }
      },
      "status": {
          "ready": true,
          "state": "Ready"
      },
      "vector_type": "sparse",
      "dimension": null,
      "deletion_protection": "disabled",
      "tags": {
          "environment": "development"
      }
  }, {
      "name": "docs-example-dense",
      "metric": "cosine",
      "host": "docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io",
      "spec": {
          "serverless": {
              "cloud": "aws",
              "region": "us-east-1"
          }
      },
      "status": {
          "ready": true,
          "state": "Ready"
      },
      "vector_type": "dense",
      "dimension": 1536,
      "deletion_protection": "disabled",
      "tags": {
          "environment": "development"
      }
  }]
  ```

  ```javascript JavaScript theme={null}
  {
    indexes: [
      {
        name: 'docs-example-sparse',
        dimension: undefined,
        metric: 'dotproduct',
        host: 'docs-example-sparse-govk0nt.svc.aped-4627-b74a.pinecone.io',
        deletionProtection: 'disabled',
        tags: { environment: 'development', example: 'tag' },
        embed: undefined,
        spec: { pod: undefined, serverless: { cloud: 'aws', region: 'us-east-1' } },
        status: { ready: true, state: 'Ready' },
        vectorType: 'sparse'
      },
      {
        name: 'docs-example-dense',
        dimension: 1536,
        metric: 'cosine',
        host: 'docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io',
        deletionProtection: 'disabled',
        tags: { environment: 'development', example: 'tag' },
        embed: undefined,
        spec: { pod: undefined, serverless: { cloud: 'aws', region: 'us-east-1' } },
        status: { ready: true, state: 'Ready' },
        vectorType: 'dense'
      }
    ]
  }
  ```

  ```java Java theme={null}
  class IndexList {
      indexes: [class IndexModel {
          name: docs-example-sparse
          dimension: null
          metric: dotproduct
          host: docs-example-sparse-govk0nt.svc.aped-4627-b74a.pinecone.io
          deletionProtection: disabled
          tags: {environment=development}
          embed: null
          spec: class IndexModelSpec {
              pod: null
              serverless: class ServerlessSpec {
                  cloud: aws
                  region: us-east-1
                  additionalProperties: null
              }
              additionalProperties: null
          }
          status: class IndexModelStatus {
              ready: true
              state: Ready
              additionalProperties: null
          }
          vectorType: sparse
          additionalProperties: null
      }, class IndexModel {
          name: docs-example-dense
          dimension: 1536
          metric: cosine
          host: docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io
          deletionProtection: disabled
          tags: {environment=development}
          embed: null
          spec: class IndexModelSpec {
              pod: null
              serverless: class ServerlessSpec {
                  cloud: aws
                  region: us-east-1
                  additionalProperties: null
              }
              additionalProperties: null
          }
          status: class IndexModelStatus {
              ready: true
              state: Ready
              additionalProperties: null
          }
          vectorType: dense
          additionalProperties: null
      }]
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  index: {
    "name": "docs-example-sparse",
    "host": "docs-example-sparse-govk0nt.svc.aped-4627-b74a.pinecone.io",
    "metric": "dotproduct",
    "vector_type": "sparse",
    "deletion_protection": "disabled",
    "dimension": null,
    "spec": {
      "serverless": {
        "cloud": "aws",
        "region": "us-east-1"
      }
    },
    "status": {
      "ready": true,
      "state": "Ready"
    },
    "tags": {
      "environment": "development"
    }
  }
  index: {
    "name": "docs-example-dense",
    "host": "docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io",
    "metric": "cosine",
    "vector_type": "dense",
    "deletion_protection": "disabled",
    "dimension": 1536,
    "spec": {
      "serverless": {
        "cloud": "aws",
        "region": "us-east-1"
      }
    },
    "status": {
      "ready": true,
      "state": "Ready"
    },
    "tags": {
      "environment": "development"
    }
  }
  ```

  ```csharp C# theme={null}
  {
    "indexes": [
      {
        "name": "docs-example-sparse",
        "metric": "dotproduct",
        "host": "docs-example-sparse-govk0nt.svc.aped-4627-b74a.pinecone.io",
        "deletion_protection": "disabled",
        "tags": {
          "environment": "development"
        },
        "spec": {
          "serverless": {
            "cloud": "aws",
            "region": "us-east-1"
          }
        },
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "vector_type": "sparse"
      },
      {
        "name": "docs-example-dense",
        "dimension": 1536,
        "metric": "cosine",
        "host": "docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io",
        "deletion_protection": "disabled",
        "tags": {
          "environment": "development"
        },
        "spec": {
          "serverless": {
            "cloud": "aws",
            "region": "us-east-1"
          }
        },
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "vector_type": "dense"
      }
    ]
  }
  ```

  ```json curl theme={null}
  {
    "indexes": [
      {
        "name": "docs-example-sparse",
        "vector_type": "sparse",
        "metric": "dotproduct",
        "dimension": null,
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "host": "docs-example-sparse-govk0nt.svc.aped-4627-b74a.pinecone.io",
        "spec": {
          "serverless": {
            "region": "us-east-1",
            "cloud": "aws"
          }
        },
        "deletion_protection": "disabled",
        "tags": {
          "environment": "development"
        }
      },
      {
        "name": "docs-example-dense",
        "vector_type": "dense",
        "metric": "cosine",
        "dimension": 1536,
        "status": {
          "ready": true,
          "state": "Ready"
        },
        "host": "docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io",
        "spec": {
          "serverless": {
            "region": "us-east-1",
            "cloud": "aws"
          }
        },
        "deletion_protection": "disabled",
        "tags": {
          "environment": "development"
        }
      }
    ]
  }
  ```
</CodeGroup>

With the Python SDK, you can use the `.names()` helper function to iterate over the index names in the `list_indexes()` response, for example:

```Python Python theme={null}
from pinecone.grpc import PineconeGRPC as Pinecone
from pinecone import ServerlessSpec

for index_name in pc.list_indexes().names:
    print(index_name)
```

## Describe an index

Use the [`describe_index`](/reference/api/latest/control-plane/describe_index/) endpoint to get a complete description of a specific index:

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.describe_index(name="docs-example")
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await pc.describeIndex('docs-example');
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.*;

  public class DescribeIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOURE_API_KEY").build();
          IndexModel indexModel = pc.describeIndex("docs-example");
          System.out.println(indexModel);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
      bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idx, err := pc.DescribeIndex(ctx, "docs-example")
      if err != nil {
          log.Fatalf("Failed to describe index \"%v\": %v", idx.Name, err)
      } else {
          fmt.Printf("index: %v\n", prettifyStruct(idx))
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var indexModel = await pinecone.DescribeIndexAsync("docs-example");

  Console.WriteLine(indexModel);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X GET "https://api.pinecone.io/indexes/docs-example" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

The response will look like this:

<CodeGroup>
  ```Python Python theme={null}
  {'deletion_protection': 'disabled',
   'dimension': 1536,
   'host': 'docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io',
   'metric': 'cosine',
   'name': 'docs-example-dense',
   'spec': {'serverless': {'cloud': 'aws', 'region': 'us-east-1'}},
   'status': {'ready': True, 'state': 'Ready'},
   'tags': {'environment': 'development'},
   'vector_type': 'dense'}
  ```

  ```javaScript JavaScript theme={null}
  {
    name: 'docs-example-dense',
    dimension: 1536,
    metric: 'cosine',
    host: 'docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io',
    deletionProtection: 'disabled',
    tags: { environment: 'development', example: 'tag' },
    embed: undefined,
    spec: { pod: undefined, serverless: { cloud: 'aws', region: 'us-east-1' } },
    status: { ready: true, state: 'Ready' },
    vectorType: 'dense'
  }
  ```

  ```java Java theme={null}
  class IndexModel {
      name: docs-example-dense
      dimension: 1536
      metric: cosine
      host: docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io
      deletionProtection: disabled
      tags: {environment=development}
      embed: null
      spec: class IndexModelSpec {
          pod: null
          serverless: class ServerlessSpec {
              cloud: aws
              region: us-east-1
              additionalProperties: null
          }
          additionalProperties: null
      }
      status: class IndexModelStatus {
          ready: true
          state: Ready
          additionalProperties: null
      }
      vectorType: dense
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  index: {
    "name": "docs-example-dense",
    "host": "docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io",
    "metric": "cosine",
    "vector_type": "dense",
    "deletion_protection": "disabled",
    "dimension": 1536,
    "spec": {
      "serverless": {
        "cloud": "aws",
        "region": "us-east-1"
      }
    },
    "status": {
      "ready": true,
      "state": "Ready"
    },
    "tags": {
      "environment": "development"
    }
  }
  ```

  ```csharp C# theme={null}
  {
    "name": "docs-example-dense",
    "dimension": 1536,
    "metric": "cosine",
    "host": "docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io",
    "deletion_protection": "disabled",
    "tags": {
      "environment": "development"
    },
    "spec": {
      "serverless": {
        "cloud": "aws",
        "region": "us-east-1"
      }
    },
    "status": {
      "ready": true,
      "state": "Ready"
    },
    "vector_type": "dense"
  }
  ```

  ```json curl theme={null}
  {
    "name": "docs-example-dense",
    "vector_type": "dense",
    "metric": "cosine",
    "dimension": 1536,
    "status": {
      "ready": true,
      "state": "Ready"
    },
    "host": "docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io",
    "spec": {
      "serverless": {
        "region": "us-east-1",
        "cloud": "aws"
      }
    },
    "deletion_protection": "disabled",
    "tags": {
      "environment": "development"
    }
  }
  ```
</CodeGroup>

<Warning>
  **Do not target an index by name in production.**

  When you target an index by name for data operations such as `upsert` and `query`, the SDK gets the unique DNS host for the index using the `describe_index` operation. This is convenient for testing but should be avoided in production because `describe_index` uses a different API than data operations and therefore adds an additional network call and point of failure. Instead, you should get an index host once and cache it for reuse or specify the host directly.
</Warning>

## Delete an index

Use the [`delete_index`](reference/api/latest/control-plane/delete_index) operation to delete an index and all of its associated resources.

<CodeGroup>
  ```python Python theme={null}
  # pip install "pinecone[grpc]"
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.delete_index(name="docs-example")
  ```

  ```javascript JavaScript theme={null}
  // npm install @pinecone-database/pinecone
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({
    apiKey: 'YOUR_API_KEY'
  });

  await pc.deleteIndex('docs-example');
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;

  public class DeleteIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          pc.deleteIndex("docs-example");
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      indexName := "docs-example"

      err = pc.DeleteIndex(ctx, indexName)
      if err != nil {
          log.Fatalf("Failed to delete index: %v", err)
      } else {
          fmt.Println("Index \"%v\" deleted successfully", indexName)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  await pinecone.DeleteIndexAsync("docs-example");
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X DELETE "https://api.pinecone.io/indexes/docs-example" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

If deletion protection is enabled on an index, requests to delete it will fail and return a `403 - FORBIDDEN` status with the following error:

```
Deletion protection is enabled for this index. Disable deletion protection before retrying.
```

Before you can delete such an index, you must first [disable deletion protection](/guides/manage-data/manage-indexes#configure-deletion-protection).

<Tip>
  You can delete an index using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/indexes). For the index you want to delete, click the three dots to the right of the index name, then click **Delete**.
</Tip>

## Associate an embedding model

[Integrated inference](/guides/index-data/indexing-overview#integrated-embedding) lets you upsert and search without extra steps for embedding data and reranking results.

To configure an existing serverless index for an embedding model, use the [`configure_index`](/reference/api/latest/control-plane/configure_index) operation as follows:

* Set `embed.model` to one of [Pinecone's hosted embedding models](/guides/index-data/create-an-index#embedding-models).
* Set `embed.field_map` to the name of the field in your source document that contains the data for embedding.

<Warning>
  The `vector_type`, `metric`, and `dimension` of the index must be supported by the specified embedding model.
</Warning>

<CodeGroup>
  ```python Python theme={null}
  # pip install --upgrade pinecone
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.configure_index(
      name="docs-example",
      embed={
        "model":"llama-text-embed-v2",
        "field_map":{"text": "chunk_text"}
      }
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await pc.configureIndex('docs-example', {
    embed: {
      model: 'llama-text-embed-v2',
      fieldMap: { text: 'chunk_text' },
    },
  });
  ```

  ```json curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s -X PATCH "https://api.pinecone.io/indexes/docs-example" \
      -H "Content-Type: application/json" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
        "embed": {
            "model": "llama-text-embed-v2",
            "field_map": {
                "text": "chunk_text"
            }
        }
    }'
  ```
</CodeGroup>

## Configure deletion protection

<Note>
  This feature requires [Pinecone API version](/reference/api/versioning) `2024-07`, [Python SDK](/reference/sdks/python/overview) v5.0.0, [Node.js SDK](/reference/sdks/node/overview) v3.0.0, [Java SDK](/reference/sdks/java/overview) v2.0.0, or [Go SDK](/reference/sdks/go/overview) v1.0.0 or later.
</Note>

### Enable deletion protection

You can prevent an index and its data from accidental deleting when [creating a new index](/guides/index-data/create-an-index) or after its been created. In both cases, you set the `deletion_protection` parameter to `enabled`.

<Warning>
  Enabling deletion protection does *not* prevent [namespace deletions](/guides/manage-data/manage-namespaces#delete-a-namespace).
</Warning>

To enable deletion protection when creating a new index:

<CodeGroup>
  ```python Python theme={null}
  # pip install "pinecone[grpc]"
  # Serverless index
  from pinecone.grpc import PineconeGRPC as Pinecone
  from pinecone import ServerlessSpec

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.create_index(
    name="docs-example",
    dimension=1536,
    metric="cosine",
    spec=ServerlessSpec(
      cloud="aws",
      region="us-east-1"
    ),
    deletion_protection="enabled"
  )
  ```

  ```javascript JavaScript theme={null}
  // npm install @pinecone-database/pinecone
  // Serverles index
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({
    apiKey: 'YOUR_API_KEY'
  });

  await pc.createIndex({
    name: 'docs-example',
    dimension: 1536,
    metric: 'cosine',
    spec: {
      serverless: {
        cloud: 'aws',
        region: 'us-east-1'
      }
    },
    deletionProtection: 'enabled',
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.IndexModel;
  import org.openapitools.db_control.client.model.DeletionProtection;

  // Serverless index
  public class CreateServerlessIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          pc.createServerlessIndex("docs-example", "cosine", 1536, "aws", "us-east-1", DeletionProtection.enabled);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // Serverless index
    	indexName := "docs-example"
  	vectorType := "dense"
      dimension := int32(1536)
      metric := pinecone.Cosine
  	deletionProtection := pinecone.DeletionProtectionDisabled

  	idx, err := pc.CreateServerlessIndex(ctx, &pinecone.CreateServerlessIndexRequest{
          Name:               indexName,
          VectorType:         &vectorType,
          Dimension:          &dimension,
          Metric:             &metric,
          Cloud:              pinecone.Aws,
          Region:             "us-east-1",
          DeletionProtection: &deletionProtection,
          Tags:               &pinecone.IndexTags{ "environment": "development" },
      })
      if err != nil {
          log.Fatalf("Failed to create serverless index: %v", err)
      } else {
          fmt.Printf("Successfully created serverless index: %v", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // Serverless index
  var createIndexRequest = await pinecone.CreateIndexAsync(new CreateIndexRequest
  {
      Name = "docs-example",
      Dimension = 1536,
      Metric = MetricType.Cosine,
      Spec = new ServerlessIndexSpec
      {
          Serverless = new ServerlessSpec
          {
              Cloud = ServerlessSpecCloud.Aws,
              Region = "us-east-1",
          }
      },
      DeletionProtection = DeletionProtection.Enabled
  });
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  # Serverless index
  curl -s "https://api.pinecone.io/indexes" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
           "name": "docs-example",
           "dimension": 1536,
           "metric": "cosine",
           "spec": {
              "serverless": {
                 "cloud": "aws",
                 "region": "us-east-1"
              }
           },
           "deletion_protection": "enabled"
        }'
  ```
</CodeGroup>

To enable deletion protection when configuring an existing index:

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.configure_index(
     name="docs-example", 
     deletion_protection="enabled"
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const client = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await client.configureIndex('docs-example', { deletionProtection: 'enabled' });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.*;

  public class ConfigureIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          pc.configureServerlessIndex("docs-example", DeletionProtection.ENABLED);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idx, err := pc.ConfigureIndex(ctx, "docs-example", pinecone.ConfigureIndexParams{DeletionProtection: "enabled"})
    	if err != nil {
          log.Fatalf("Failed to configure index \"%v\": %v", idx.Name, err)
      } else {
          fmt.Printf("Successfully configured index \"%v\"", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var indexMetadata = await pinecone.ConfigureIndexAsync("docs-example", new ConfigureIndexRequest
  {
      DeletionProtection = DeletionProtection.Enabled,
  });
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s -X PATCH "https://api.pinecone.io/indexes/docs-example-curl" \
      -H "Content-Type: application/json" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
          "deletion_protection": "enabled"
          }'
  ```
</CodeGroup>

When deletion protection is enabled on an index, requests to delete the index fail and return a `403 - FORBIDDEN` status with the following error:

```
Deletion protection is enabled for this index. Disable deletion protection before retrying.
```

### Disable deletion protection

Before you can [delete an index](#delete-an-index) with deletion protection enabled, you must first disable deletion protection as follows:

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.configure_index(
     name="docs-example", 
     deletion_protection="disabled"
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const client = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await client.configureIndex('docs-example', { deletionProtection: 'disabled' });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.*;

  public class ConfigureIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          pc.configureServerlessIndex("docs-example", DeletionProtection.DISABLED);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idx, err := pc.ConfigureIndex(ctx, "docs-example", pinecone.ConfigureIndexParams{DeletionProtection: "disabled"})
    	if err != nil {
          log.Fatalf("Failed to configure index \"%v\": %v", idx.Name, err)
      } else {
          fmt.Printf("Successfully configured index \"%v\"", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var configureIndexRequest = await pinecone.ConfigureIndexAsync("docs-example", new ConfigureIndexRequest
  {
      DeletionProtection = DeletionProtection.Disabled,
  });
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s -X PATCH "https://api.pinecone.io/indexes/docs-example-curl" \
      -H "Content-Type: application/json" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
          "deletion_protection": "disabled"
          }'
  ```
</CodeGroup>

## Configure index tags

Tags are key-value pairs that you can use to categorize and identify the index.

### Add tags

To add tags to an index, use the `tags` parameter when [creating a new index](/guides/index-data/create-an-index) or configuring an existing index.

To add tags when creating a new index:

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone
  from pinecone import ServerlessSpec

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.create_index(
      name="docs-example",
      dimension=1536,
      metric="cosine",
      spec=ServerlessSpec(
          cloud="aws",
          region="us-east-1"
      ),
      deletion_protection="disabled",
      tags={
          "example": "tag", 
          "environment": "development"
      }
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const client = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await pc.createIndex({
    name: 'docs-example',
    dimension: 1536,
    metric: 'cosine',
    spec: {
      serverless: {
        cloud: 'aws',
        region: 'us-east-1'
      }
    },
    deletionProtection: 'disabled',
    tags: { example: 'tag', environment: 'development' }, 
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.IndexModel;
  import org.openapitools.db_control.client.model.DeletionProtection;
  import java.util.HashMap;

  // Serverless index
  public class CreateServerlessIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          HashMap<String, String> tags = new HashMap<>();
          tags.put("tag", "development");
          pc.createServerlessIndex("docs-example", "cosine", 1536, "aws", "us-east-1", DeletionProtection.DISABLED, tags);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // Serverless index
      idx, err := pc.CreateServerlessIndex(ctx, &pinecone.CreateServerlessIndexRequest{
          Name:               "docs-example",
          Dimension:          1536,
          Metric:             pinecone.Cosine,
          Cloud:              pinecone.Aws,
          Region:             "us-east-1",
          DeletionProtection: "disabled",
          Tags:               &pinecone.IndexTags{ "example": "tag", "environment": "development" },
      })
    	if err != nil {
          log.Fatalf("Failed to create serverless index: %v", err)
      } else {
          fmt.Printf("Successfully created serverless index: %v", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var createIndexRequest = await pinecone.CreateIndexAsync(new CreateIndexRequest
  {
      Name = "docs-example",
      Dimension = 1536,
      Metric = MetricType.Cosine,
      Spec = new ServerlessIndexSpec
      {
          Serverless = new ServerlessSpec
          {
              Cloud = ServerlessSpecCloud.Aws,
              Region = "us-east-1"
          }
      },
      DeletionProtection = DeletionProtection.Disabled,
      Tags = new Dictionary<string, string> 
      { 
          { "example", "tag" }, 
          { "environment", "development" }
      }
  });
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  # Serverless index
  curl -s "https://api.pinecone.io/indexes" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
           "name": "docs-example",
           "dimension": 1536,
           "metric": "cosine",
           "spec": {
              "serverless": {
                 "cloud": "aws",
                 "region": "us-east-1"
              }
           },
          "tags": {
              "example": "tag",
              "environment": "development"
          },
           "deletion_protection": "disabled"
        }'
  ```
</CodeGroup>

<Tip>
  You can add tags during index creation using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/create-index/).
</Tip>

To add or update tags when configuring an existing index:

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.configure_index(
      name="docs-example", 
      tags={
          example: "tag", 
          environment: "development" 
      }
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const client = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await client.configureIndex('docs-example', { tags: { example: 'tag', environment: 'development' }});
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.*;
  import java.util.HashMap;

  public class ConfigureIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          HashMap<String, String> tags = new HashMap<>();
          tags.put("tag", "development");

          pc.configureServerlessIndex("docs-example", DeletionProtection.ENABLED, tags);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idx, err := pc.ConfigureIndex(ctx, 
          "docs-example", 
          pinecone.ConfigureIndexParams{
              Tags: pinecone.IndexTags{
  			    "example": "tag",
                  "environment": "development",
              },
          },
      )    
    	if err != nil {
          log.Fatalf("Failed to configure index \"%v\": %v", idx.Name, err)
      } else {
          fmt.Printf("Successfully configured index \"%v\"", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var configureIndexRequest = await pinecone.ConfigureIndexAsync("docs-example", new ConfigureIndexRequest
  {
      Tags = new Dictionary<string, string> 
      { 
          { "example", "tag" }, 
          { "environment", "development" }
      }
  });
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s -X PATCH "https://api.pinecone.io/indexes/docs-example-curl" \
      -H "Content-Type: application/json" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "tags": {
                  "example": "tag",
                  "environment": "development"
              }
          }'
  ```
</CodeGroup>

<Tip>
  You can add or update tags when configuring an existing index using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/indexes). Find the index to edit and click the **ellipsis (...) menu > Add tags**.
</Tip>

### View tags

To view the tags of an index, [list all indexes](/guides/manage-data/manage-indexes) in a project or [get information about a specific index](/guides/manage-data/manage-indexes).

### Remove tags

To remove a tag from an index, [configure the index](/reference/api/latest/control-plane/configure_index) and use the `tags` parameter to send the tag key with an empty value (`""`).

The following example removes the `example: tag` tag from `docs-example`:

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.configure_index(
      name="docs-example", 
      tags={"example": ""}
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const client = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await client.configureIndex('docs-example', { tags: { example: '' }});
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.*;
  import java.util.HashMap;

  public class ConfigureIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          HashMap<String, String> tags = new HashMap<>();
          tags.put("example", "");

          pc.configureServerlessIndex("docs-example", DeletionProtection.ENABLED, tags);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idx, err := pc.ConfigureIndex(ctx, 
          "docs-example", 
          pinecone.ConfigureIndexParams{
              Tags: pinecone.IndexTags{
  			    "example": "",
              },
          },
      )    
    	if err != nil {
          log.Fatalf("Failed to configure index \"%v\": %v", idx.Name, err)
      } else {
          fmt.Printf("Successfully configured index \"%v\"", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var configureIndexRequest = await pinecone.ConfigureIndexAsync("docs-example", new ConfigureIndexRequest
  {
      Tags = new Dictionary<string, string> 
      { 
          { "example", "" } 
      }
  });
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s -X PATCH "https://api.pinecone.io/indexes/docs-example-curl" \
      -H "Content-Type: application/json" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
              "tags": {
                  "example": ""
              }
          }'
  ```
</CodeGroup>

<Tip>
  You can remove tags from an index using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/indexes). Find the index to edit and click the **ellipsis (...) menu > \_\_ tags**.
</Tip>

## List backups for an index

Serverless indexes can be [backed up](/guides/manage-data/back-up-an-index). You can [list all backups for a specific index](/reference/api/latest/control-plane/list_index_backups), as in the following example:

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  index_backups = pc.list_backups(index_name="docs-example")

  print(index_backups)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  const indexBackups = await pc.listBackups({ indexName: 'docs-example' });

  console.log(indexBackups);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.ApiException;
  import org.openapitools.db_control.client.model.*;

  public class CreateBackup {
      public static void main(String[] args) throws ApiException {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();

          String indexName = "docs-example";
          BackupList indexBackupList = pc.listIndexBackups(indexName);

          System.out.println(indexBackupList);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
  	ctx := context.Background()

  	pc, err := pinecone.NewClient(pinecone.NewClientParams{
  		ApiKey: "YOUR_API_KEY",
  	})
  	if err != nil {
  		log.Fatalf("Failed to create Client: %v", err)
  	}

  	indexName := "docs-example"
  	limit := 2
  	indexBackups, err := pc.ListBackups(ctx, &pinecone.ListBackupsParams{
  		Limit: &limit,
  		IndexName: &indexName,
  	})
  	if err != nil {
  		log.Fatalf("Failed to list backups: %v", err)
  	}
  	fmt.Printf(prettifyStruct(indexBackups))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var indexBackups = await pinecone.Backups.ListByIndexAsync( "docs-example", new ListBackupsByIndexRequest());

  Console.WriteLine(indexBackups);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_NAME="docs-example"

  curl -X GET "https://api.pinecone.io/indexes/$INDEX_NAME/backups" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -H "accept: application/json"
  ```
</CodeGroup>

The example returns a response like the following:

<CodeGroup>
  ```python Python theme={null}
  [{
      "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
      "source_index_name": "docs-example",
      "source_index_id": "f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74",
      "status": "Ready",
      "cloud": "aws",
      "region": "us-east-1",
      "tags": {},
      "name": "example-backup",
      "description": "Monthly backup of production index",
      "dimension": 1024,
      "record_count": 98,
      "namespace_count": 3,
      "size_bytes": 1069169,
      "created_at": "2025-05-15T00:52:10.809305882Z"
  }]
  ```

  ```javascript JavaScript theme={null}
  {
    data: [
      {
        backupId: '8c85e612-ed1c-4f97-9f8c-8194e07bcf71',
        sourceIndexName: 'docs-example',
        sourceIndexId: 'f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74',
        name: 'example-backup',
        description: 'Monthly backup of production index',
        status: 'Ready',
        cloud: 'aws',
        region: 'us-east-1',
        dimension: 1024,
        metric: undefined,
        recordCount: 98,
        namespaceCount: 3,
        sizeBytes: 1069169,
        tags: {},
        createdAt: '2025-05-14T16:37:25.625540Z'
      }
    ],
    pagination: undefined
  }
  ```

  ```java Java theme={null}
  class BackupList {
      data: [class BackupModel {
          backupId: 8c85e612-ed1c-4f97-9f8c-8194e07bcf71
          sourceIndexName: docs-example
          sourceIndexId: f73b36c9-faf5-4a2c-b1d6-4013d8b1cc74
          name: example-backup
          description: Monthly backup of production index
          status: Initializing
          cloud: aws
          region: us-east-1
          dimension: null
          metric: null
          recordCount: null
          namespaceCount: null
          sizeBytes: null
          tags: {}
          createdAt: 2025-05-16T19:46:26.248428Z
          additionalProperties: null
      }]
      pagination: null
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  {
    "data": [
      {
        "backup_id": "bf2cda5d-b233-4a0a-aae9-b592780ad3ff",
        "cloud": "aws",
        "created_at": "2025-05-16T18:01:51.531129Z",
        "description": "Monthly backup of production index",
        "dimension": 0,
        "name": "example-backup",
        "namespace_count": 1,
        "record_count": 96,
        "region": "us-east-1",
        "size_bytes": 86393,
        "source_index_id": "bcb5b3c9-903e-4cb6-8b37-a6072aeb874f",
        "source_index_name": "docs-example",
        "status": "Ready",
        "tags": {}
      },
      {
        "backup_id": "e12269b0-a29b-4af0-9729-c7771dec03e3",
        "cloud": "aws",
        "created_at": "2025-05-14T17:00:45.803146Z",
        "dimension": 0,
        "name": "example-backup2",
        "namespace_count": 1,
        "record_count": 96,
        "region": "us-east-1",
        "size_bytes": 86393,
        "source_index_id": "bcb5b3c9-903e-4cb6-8b37-a6072aeb874f",
        "source_index_name": "docs-example",
        "status": "Ready"
      }
    ],
    "pagination": {
      "next": "eyJsaW1pdCI6Miwib2Zmc2V0IjoyfQ=="
    }
  }
  ```

  ```csharp C# theme={null}
  {
    "data":
    [
      {
        "backup_id":"9947520e-d5a1-4418-a78d-9f464c9969da",
        "source_index_id":"8433941a-dae7-43b5-ac2c-d3dab4a56b2b",
        "source_index_name":"docs-example",
        "tags":{},
        "name":"example-backup",
        "description":"Monthly backup of production index",
        "status":"Pending",
        "cloud":"aws",
        "region":"us-east-1",
        "dimension":1024,
        "record_count":98,
        "namespace_count":3,
        "size_bytes":1069169,
        "created_at":"2025-03-11T18:29:50.549505Z"
      }
    ]
  }
  ```

  ```json curl theme={null}
  {
    "data":
    [
      {
        "backup_id":"9947520e-d5a1-4418-a78d-9f464c9969da",
        "source_index_id":"8433941a-dae7-43b5-ac2c-d3dab4a56b2b",
        "source_index_name":"docs-example",
        "tags":{},
        "name":"example-backup",
        "description":"Monthly backup of production index",
        "status":"Pending",
        "cloud":"aws",
        "region":"us-east-1",
        "dimension":1024,
        "record_count":98,
        "namespace_count":3,
        "size_bytes":1069169,
        "created_at":"2025-03-11T18:29:50.549505Z"
        }
      ],
    "pagination":null
  }
  ```
</CodeGroup>

<Tip>
  You can view the backups for a specific index from either the [Backups](https://app.pinecone.io/organizations/-/projects/-/backups) tab or the [Indexes](https://app.pinecone.io/organizations/-/projects/-/indexes) tab in the Pinecone console.
</Tip>

---

# Source: https://docs.pinecone.io/guides/manage-data/manage-namespaces.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage namespaces

> Create and manage namespaces in serverless indexes.

## Create a namespace

<Warning>
  This feature is in [early access](/release-notes/feature-availability) and available only on the `2025-10` version of the API.
</Warning>

Namespaces are created automatically during [upsert](/guides/index-data/upsert-data). However, you can also create namespaces ahead of time using the [`create_namespace`](/reference/api/2025-10/data-plane/createnamespace) operation. Specify a name for the namespace and, optionally, the [metadata fields to index](/guides/index-data/create-an-index#metadata-indexing).

```shell curl theme={null}
# To get the unique host for an index,
# see https://docs.pinecone.io/guides/manage-data/target-an-index
PINECONE_API_KEY="YOUR_API_KEY"
INDEX_HOST="INDEX_HOST"

curl "https://$INDEX_HOST/namespaces" \
  -H "Accept: application/json" \
  -H "Content-Type: application/json" \
  -H "Api-Key: $PINECONE_API_KEY" \
  -H "X-Pinecone-Api-Version: 2025-10" \
  -d '{
        "name": "example-namespace",
        "schema": {
		  "fields": { 
            "document_id": {"filterable": true},
            "document_title": {"filterable": true},
            "chunk_number": {"filterable": true},
            "document_url": {"filterable": true},
            "created_at": {"filterable": true}
          }
        }
      }'
```

The response will look like the following:

```json curl theme={null}
{
    "name": "example-namespace",
    "record_count": "0",
    "schema": {
        "fields": {
            "document_title": {
                "filterable": true
            },
            "document_url": {
                "filterable": true
            },
            "chunk_number": {
                "filterable": true
            },
            "document_id": {
                "filterable": true
            },
            "created_at": {
                "filterable": true
            }
        }
    }
}
```

## List all namespaces in an index

Use the [`list_namespaces`](/reference/api/latest/data-plane/listnamespaces) operation to list all namespaces in a serverless index.

Up to 100 namespaces are returned at a time by default, in sorted order (bitwise “C” collation). If the `limit` parameter is set, up to that number of namespaces are returned instead. Whenever there are additional namespaces to return, the response also includes a `pagination_token` that you can use to get the next batch of namespaces. When the response does not include a `pagination_token`, there are no more namespaces to return.

<CodeGroup>
  ```python Python theme={null}
  # Not supported with pinecone["grpc"] extras installed
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  index = pc.Index(host="INDEX_HOST")

  # Implicit pagination using a generator function
  for namespace in index.list_namespaces():
      print(namespace.name, ":", namespace.record_count)

  # Manual pagination
  namespaces = index.list_namespaces_paginated(
      limit=2,
      pagination_token="eyJza2lwX3Bhc3QiOiIxMDEwMy0="
  )

  print(namespaces)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  const namespaceList = await index.listNamespaces();

  console.log(namespaceList);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.AsyncIndex;
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import io.pinecone.proto.ListNamespacesResponse;
  import org.openapitools.db_data.client.ApiException;

  public class Namespaces {
      public static void main(String[] args) throws ApiException {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);

          Index index = new Index(config, connection, "docs-example");

          // List all namespaces with default pagination limit (100)
          ListNamespacesResponse listNamespacesResponse = index.listNamespaces(null, null);

          // List all namespaces with pagination limit of 2
          ListNamespacesResponse listNamespacesResponseWithLimit = index.listNamespaces(2);

          // List all namespaces with pagination limit and token
          ListNamespacesResponse listNamespacesResponsePaginated = index.listNamespaces(5, "eyJza2lwX3Bhc3QiOiIxMDEwMy0=");

          System.out.println(listNamespacesResponseWithLimit);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
    	bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
      } 

      limit := uint32(10)
      namespaces, err := idxConnection.ListNamespaces(ctx, &pinecone.ListNamespacesParams{
          Limit: &limit,
      })
      if err != nil {
          log.Fatalf("Failed to list namespaces: %v", err)
      }
      fmt.Printf(prettifyStruct(namespaces))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("PINECONE_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var namespaces = await index.ListNamespacesAsync(new ListNamespacesRequest());

  Console.WriteLine(namespaces);
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -X GET "https://$INDEX_HOST/namespaces" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

The response will look like the following:

<CodeGroup>
  ```python Python theme={null}
  # Implicit pagination
  example-namespace : 20000
  example-namespace2 : 10500
  example-namespace3 : 10000
  ...

  # Manual pagination
  {
      "namespaces": [
          {
              "name": "example-namespace",
              "record_count": "20000"
          },
          {
              "name": "example-namespace2",
              "record_count": "10500"
          }
      ],
      "pagination": {
          "next": "Tm90aGluZyB0byBzZWUgaGVyZQo="
      }
  }
  ```

  ```javascript JavaScript theme={null}
  {
    namespaces: [
      { name: 'example-namespace', recordCount: '20000' },
      { name: 'example-namespace2', recordCount: '10500' },
      ...
    ],
    pagination: "Tm90aGluZyB0byBzZWUgaGVyZQo="
  }
  ```

  ```java Java theme={null}
  namespaces {
    name: "example-namespace"
    record_count: 20000
  }
  namespaces {
    name: "example-namespace2"
    record_count: 10500
  }
  pagination {
    next: "eyJza2lwX3Bhc3QiOiJlZDVhYzFiNi1kMDFiLTQ2NTgtYWVhZS1hYjJkMGI2YzBiZjQiLCJwcmVmaXgiOm51bGx9"
  }
  ```

  ```go Go theme={null}
  {
    "Namespaces": [
      {
        "name": "example-namespace",
        "record_count": 20000
      },
      {
        "name": "example-namespace2",
        "record_count": 10500
      },
      ...
    ],
    "Pagination": {
      "next": "eyJza2lwX3Bhc3QiOiIyNzQ5YTU1YS0zZTQ2LTQ4MDItOGFlNi1hZTJjZGNkMTE5N2IiLCJwcmVmaXgiOm51bGx9"
    }
  }
  ```

  ```csharp C# theme={null}
  {
    "namespaces":[
      {"name":"example-namespace","recordCount":20000},
      {"name":"example-namespace2","recordCount":10500},
      ...
    ],
    "pagination":"Tm90aGluZyB0byBzZWUgaGVyZQo="
  }
  ```

  ```json curl theme={null}
  {
    "namespaces": [
      {
        "name": "example-namespace",
        "record_count": 20000
      },
      {
        "name": "example-namespace2",
        "record_count": 10500
      },
      ...
    ],
    "pagination": {
      "next": "Tm90aGluZyB0byBzZWUgaGVyZQo="
    }
  }
  ```
</CodeGroup>

## Describe a namespace

Use the [`describe_namespace`](/reference/api/latest/data-plane/describenamespace) operation to get details about a namespace in a serverless index, including the total number of vectors in the namespace.

<CodeGroup>
  ```python Python theme={null}
  # Not supported with pinecone["grpc"] extras installed
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  index = pc.Index(host="INDEX_HOST")

  namespace = index.describe_namespace(namespace="example-namespace")

  print(namespace)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  const index = pc.index('docs-example');

  const namespace = await index.describeNamespace('example-namespace');

  console.log(namespace);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import io.pinecone.proto.NamespaceDescription;

  import org.openapitools.db_data.client.ApiException;

  public class Namespaces {
      public static void main(String[] args) throws ApiException {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);

          Index index = new Index(config, connection, "docs-example");

          NamespaceDescription namespaceDescription = index.describeNamespace("example-namespace");

          System.out.println(namespaceDescription);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
    	bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
      } 

      namespace, err := idxConnection.DescribeNamespace(ctx, "example-namespace")
      if err != nil {
          log.Fatalf("Failed to describe namespace: %v", err)
      }
      fmt.Printf(prettifyStruct(namespace))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("PINECONE_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var @namespace = await index.DescribeNamespaceAsync("example-namespace");

  Console.WriteLine(@namespace);
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"
  NAMESPACE="NAMESPACE_NAME"  # To target the default namespace, use "__default__".

  curl -X GET "https://$INDEX_HOST/namespaces/$NAMESPACE" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

The response will look like the following:

<CodeGroup>
  ```python Python theme={null}
  {
      "name": "example-namespace",
      "record_count": "20000"
  }
  ```

  ```javascript JavaScript theme={null}
  { name: 'example-namespace', recordCount: '20000' }
  ```

  ```java Java theme={null}
  name: "example-namespace"
  record_count: 20000
  ```

  ```go Go theme={null}
  {
    "name": "example-namespace",
    "record_count": 20000
  }
  ```

  ```csharp C# theme={null}
  {"name":"example-namespace","recordCount":20000}
  ```

  ```json curl theme={null}
  {
    "name": "example-namespace",
    "record_count": 20000
  }
  ```
</CodeGroup>

## Delete a namespace

Use the [`delete_namespace`](/reference/api/latest/data-plane/deletenamespace) operation to delete a namespace in a serverless index.

<Warning>
  Deleting a namespace is irreversible. All data in the namespace is permanently deleted.
</Warning>

<CodeGroup>
  ```python Python theme={null}
  # Not supported with pinecone["grpc"] extras installed
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  index = pc.Index(host="INDEX_HOST")

  index.delete_namespace(namespace="example-namespace")
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  const index = pc.index('INDEX_NAME', 'INDEX_HOST');

  const namespace = await index.deleteNamespace('example-namespace');

  console.log(namespace);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;

  import java.util.concurrent.ExecutionException;

  public class DeleteNamespace {
      public static void main(String[] args) throws ExecutionException, InterruptedException {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
        // To get the unique host for an index, 
        // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);

          Index index = new Index(config, connection, "docs-example");

          index.deleteNamespace("example-namespace");
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
    	bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
      } 

      err := idxConnection.DeleteNamespace(ctx, "example-namespace")
      if err != nil {
          log.Fatalf("Failed to delete namespace: %v", err)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  const pinecone = new PineconeClient("PINECONE_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pinecone.Index(host: "INDEX_HOST");

  await index.DeleteNamespaceAsync("example-namespace");
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"
  NAMESPACE="NAMESPACE_NAME" # To target the default namespace, use "__default__".

  curl -X DELETE "https://$INDEX_HOST/namespaces/$NAMESPACE" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

## Rename a namespace

Pinecone does not support renaming namespaces directly. Instead, you must [delete the records](/guides/manage-data/delete-data) in the namespace and [upsert the records](/guides/index-data/upsert-data) to a new namespace.

## Move records to a new namespace

Pinecone does not support moving records between namespaces directly. Instead, you must [delete the records](/guides/manage-data/delete-data) in the old namespace and [upsert the records](/guides/index-data/upsert-data) to the new namespace.

## Use the default namespace

To use the default namespace for upserts, queries, or other data operations, set the `namespace` parameter to `__default__`, for example:

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  results = index.search(
      namespace="example-namespace", 
      query={
          "inputs": {"text": "Disease prevention"}, 
          "top_k": 2
      },
      fields=["category", "chunk_text"]
  )

  print(results)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const namespace = pc.index("INDEX_NAME", "INDEX_HOST").namespace("example-namespace");

  const response = await namespace.searchRecords({
    query: {
      topK: 2,
      inputs: { text: 'Disease prevention' },
    },
    fields: ['chunk_text', 'category'],
  });

  console.log(response);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import org.openapitools.db_data.client.ApiException;
  import org.openapitools.db_data.client.model.SearchRecordsResponse;

  import java.util.*;

  public class SearchText {
      public static void main(String[] args) throws ApiException {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);

          Index index = new Index(config, connection, "integrated-dense-java");

          String query = "Disease prevention";
          List<String> fields = new ArrayList<>();
          fields.add("category");
          fields.add("chunk_text");

          // Search the dense index
          SearchRecordsResponse recordsResponse = index.searchRecordsByText(query,  "example-namespace", fields, 2, null, null);

          // Print the results
          System.out.println(recordsResponse);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
    	bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
      } 

      res, err := idxConnection.SearchRecords(ctx, &pinecone.SearchRecordsRequest{
          Query: pinecone.SearchRecordsQuery{
              TopK: 2,
              Inputs: &map[string]interface{}{
                  "text": "Disease prevention",
              },
          },
          Fields: &[]string{"chunk_text", "category"},
      })
      if err != nil {
          log.Fatalf("Failed to search records: %v", err)
      }
      fmt.Printf(prettifyStruct(res))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var response = await index.SearchRecordsAsync(
      "example-namespace",
      new SearchRecordsRequest
      {
          Query = new SearchRecordsRequestQuery
          {
              TopK = 4,
              Inputs = new Dictionary<string, object?> { { "text", "Disease prevention" } },
          },
          Fields = ["category", "chunk_text"],
      }
  );

  Console.WriteLine(response);
  ```

  ```shell curl theme={null}
  INDEX_HOST="INDEX_HOST"
  NAMESPACE="NAMESPACE_NAME"
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://$INDEX_HOST/records/namespaces/$NAMESPACE/search" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: unstable" \
    -d '{
          "query": {
              "inputs": {"text": "Disease prevention"},
              "top_k": 2
          },
          "fields": ["category", "chunk_text"]
       }'
  ```
</CodeGroup>

---

# Source: https://docs.pinecone.io/guides/organizations/manage-organization-members.md

# Source: https://docs.pinecone.io/guides/assistant/admin/manage-organization-members.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage organization members

> Invite and control organization member access levels.

This page shows how [organization owners](guides/assistant/admin/organizations-overview#organization-roles) can add and manage organization members.

<Tip>
  For information about managing members at the **project-level**, see [Manage project members](/guides/assistant/admin/manage-project-members).
</Tip>

## Add a member to an organization

You can add members to your organization in the [Pinecone console](https://app.pinecone.io):

1. Go to [**Settings > Access > Members**](https://app.pinecone.io/organizations/-/settings/access/members).
2. In the **Invite by email** field, enter the member's email address.
3. Choose an [**Organization role**](/guides/assistant/admin/organizations-overview#organization-roles) for the member. The role determines the member's permissions within Pinecone.
4. Click **Invite**.

When you invite a member to join your organization, Pinecone sends them an email containing a link that enables them to gain access to the organization or project. If they already have a Pinecone account, they still receive an email, but they can also immediately view the project.

## Change a member's role

You can change a member's role in the [Pinecone console](https://app.pinecone.io):

1. Go to [**Settings > Access > Members**](https://app.pinecone.io/organizations/-/settings/access/members).
2. In the row of the member whose role you want to change, click **ellipsis (...) menu > Edit role**.
3. Select a [**Project role**](/guides/assistant/admin/projects-overview#project-roles) for the member.
4. Click **Edit role**.

## Remove a member

You can remove a member from your organization in the [Pinecone console](https://app.pinecone.io):

1. In the Pinecone console, go to [**Settings > Access > Members**](https://app.pinecone.io/organizations/-/settings/access/members).
2. In the row of the member you want to remove, click **ellipsis (...) menu > Remove member**.
3. Click **Remove Member**.

<Note>
  To remove yourself from an organization, click the **Leave organization** button in your user's row and confirm.
</Note>

---

# Source: https://docs.pinecone.io/guides/assistant/admin/manage-organization-service-accounts.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage service accounts at the organization-level

> Create service accounts for organization-level API access.

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

This page shows how [organization owners](/guides/assistant/admin/organizations-overview#organization-roles) can add and manage service accounts at the organization-level. Service accounts enable programmatic access to Pinecone's Admin API, which can be used to create and manage projects and API keys.

<Tip>
  Once a service account is added at the organization-level, it can be added to a project. For more information, see [Manage service accounts at the project-level](/guides/assistant/admin/manage-project-service-accounts).
</Tip>

## Create a service account

You can create a service account in the [Pinecone console](https://app.pinecone.io):

1. Go to [**Settings > Access > Service accounts**](https://app.pinecone.io/organizations/-/settings/access/service-accounts).

2. Enter a **Name** for the service account.

3. Choose an [**Organization Role**](/guides/assistant/admin/organizations-overview#organization-roles) for the service account. The role determines the service account's permissions within Pinecone.

4. Click **Create**.

5. Copy and save the **Client secret** in a secure place for future use.

   <Warning>
     You will not be able to see the client secret again after you close the dialog.
   </Warning>

6. Click **Close**.

Once you have created a service account, [add it to a project](/guides/assistant/admin/manage-project-service-accounts#add-a-service-account-to-a-project) to allow it access to the project's resources.

## Retrieve an access token

To access the Admin API, you must provide an access token to authenticate. Retrieve the access token using the client secret of a service account, which was [provided at time of creation](#create-a-service-account).

You can retrieve an access token for a service account from the `https://login.pinecone.io/oauth/token` endpoint, as shown in the following example:

```bash curl theme={null}
curl "https://login.pinecone.io/oauth/token" \ # Note: Base URL is login.pinecone.io
	-H "X-Pinecone-Api-Version: 2025-10" \
	-H "Content-Type: application/json" \
	-d '{
		"grant_type": "client_credentials",
		"client_id": "YOUR_CLIENT_ID",
		"client_secret": "YOUR_CLIENT_SECRET",
		"audience": "https://api.pinecone.io/"
	}'
```

The response will include an `access_token` field, which you can use to authenticate with the Admin API.

```
{
    "access_token":"YOUR_ACCESS_TOKEN",
    "expires_in":86400,
    "token_type":"Bearer"
}
```

## Change a service account's role

You can change a service account's role in the [Pinecone console](https://app.pinecone.io):

1. Go to [**Settings > Access > Service accounts**](https://app.pinecone.io/organizations/-/settings/service-accounts).
2. In the row of the service account you want to update, click **ellipsis (...) menu > Manage**.
3. Select an [**Organization role**](/guides/assistant/admin/organizations-overview#organization-roles) for the service account.
4. Click **Update**.

## Update service account name

You can change a service account's name in the [Pinecone console](https://app.pinecone.io):

1. Go to [**Settings > Access > Service accounts**](https://app.pinecone.io/organizations/-/settings/service-accounts).
2. In the row of the service account you want to update, click **ellipsis (...) menu > Manage**.
3. Enter a new **Service account name**.
4. Click **Update**.

## Rotate a service account's secret

You can rotate a service account's client secret in the [Pinecone console](https://app.pinecone.io):

1. Go to [**Settings > Access > Service accounts**](https://app.pinecone.io/organizations/-/settings/service-accounts).

2. In the row of the service account you want to update, click **ellipsis (...) menu > Rotate secret**.

3. **Enter the service account name** to confirm.

4. Click **Rotate client secret**.

5. Copy and save the **Client secret** in a secure place for future use.

   <Warning>
     You will not be able to see the client secret again after you close the dialog.
   </Warning>

6. Click **Close**.

## Delete a service account

Deleting a service account will remove it from all projects and will disrupt any applications using it to access Pinecone. You delete a service account in the [Pinecone console](https://app.pinecone.io):

1. Go to [**Settings > Access > Service accounts**](https://app.pinecone.io/organizations/-/settings/service-accounts).
2. In the row of the service account you want to update, click **ellipsis (...) menu > Delete**.
3. **Enter the service account name** to confirm.
4. Click **Delete service account**.

---

# Source: https://docs.pinecone.io/guides/indexes/pods/manage-pod-based-indexes.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage pod-based indexes

> List, describe, and configure pod-based indexes

<Warning>
  Customers who sign up for a Standard or Enterprise plan on or after August 18, 2025 cannot create pod-based indexes. Instead, create [serverless indexes](/guides/index-data/create-an-index), and consider using [dedicated read nodes](/guides/index-data/dedicated-read-nodes) for large workloads (millions of records or more, and moderate or high query rates).
</Warning>

This page shows you how to manage pod-based indexes.

For guidance on serverless indexes, see [Manage serverless indexes](/guides/manage-data/manage-indexes).

## Describe a pod-based index

Use the [`describe_index`](/reference/api/latest/control-plane/describe_index/) endpoint to get a complete description of a specific index:

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.describe_index(name="docs-example")

  # Response:
  # {'dimension': 1536,
  #  'host': 'docs-example-4mkljsz.svc.aped-4627-b74a.pinecone.io',
  #  'metric': 'cosine',
  #  'name': 'docs-example',
  #  'spec': {'pod': {'environment': 'us-east-1-aws',
  #                   'pod_type': 's1.x1',
  #                   'pods': 1,
  #                   'replicas': 1,
  #                   'shards': 1}},
  #  'status': {'ready': True, 'state': 'Ready'}}
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await pc.describeIndex('docs-example');

  // Response:
  // {
  //    "name": "docs-example",
  //    "dimension": 1536,
  //    "metric": "cosine",
  //    "host": "docs-example-4mkljsz.svc.aped-4627-b74a.pinecone.io",
  //    "deletionProtection": "disabled",
  //    "spec": {
  //       "pod": {
  //          "environment": "us-east-1-aws",
  //          "pod_type": "s1.x1",
  //          "pods": 1,
  //          "replicas": 1,
  //          "shards": 1
  //       }
  //    },
  //    "status": {
  //       "ready": true,
  //       "state": "Ready"
  //    }
  // }
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.*;

  public class DescribeIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOURE_API_KEY").build();
          IndexModel indexModel = pc.describeIndex("docs-example");
          System.out.println(indexModel);
      }
  }

  // Response:
  // class IndexModel {
  //     name: docs-example
  //     dimension: 1536
  //     metric: cosine
  //     host: docs-example-4mkljsz.svc.aped-4627-b74a.pinecone.io
  //     deletionProtection: disabled
  //     spec: class IndexModelSpec {
  //         serverless: null
  //         pod: class PodSpec {
  //             cloud: aws
  //             region: us-east-1
  //             environment: us-east-1-aws,
  //             podType: s1.x1,
  //             pods: 1,
  //             replicas: 1,
  //             shards: 1
  //         }
  //     }
  //     status: class IndexModelStatus {
  //         ready: true
  //         state: Ready
  //     }
  // }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
      bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idx, err := pc.DescribeIndex(ctx, "docs-example")
      if err != nil {
          log.Fatalf("Failed to describe index \"%v\": %v", idx.Name, err)
      } else {
          fmt.Printf("index: %v\n", prettifyStruct(idx))
      }
  }

  // Response:
  // index: {
  // 	"name": "docs-example",
  // 	"dimension": 1536,
  // 	"host": "docs-example-4mkljsz.svc.aped-4627-b74a.pinecone.io",
  // 	"metric": "cosine",
  //  "deletion_protection": "disabled",
  // 	"spec": {
  // 	  "pod": {
  //       "environment": "us-east-1-aws",
  //       "pod_type": "s1.x1",
  //       "pods": 1,
  //       "replicas": 1,
  //       "shards": 1
  // 	  }
  // 	},
  // 	"status": {
  // 	  "ready": true,
  // 	  "state": "Ready"
  // 	}
  // }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var indexModel = await pinecone.DescribeIndexAsync("docs-example");

  Console.WriteLine(indexModel);

  // Response:
  // {
  //   "name": "docs-example",
  //   "dimension": 1536,
  //   "metric": "cosine",
  //   "host": "docs-example-4mkljsz.svc.aped-4627-b74a.pinecone.io",
  //   "deletion_protection": "disabled",
  //   "spec": {
  //     "serverless": null,
  //     "pod": {
  //        "environment": "us-east-1-aws",
  //        "pod_type": "s1.x1",
  //        "pods": 1,
  //        "replicas": 1,
  //        "shards": 1
  //     }
  //   },
  //   "status": {
  //     "ready": true,
  //     "state": "Ready"
  //   }
  // }
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X GET "https://api.pinecone.io/indexes/docs-example" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"
  # Response:
  # {
  #    "name": "docs-example",
  #    "metric": "cosine",
  #    "dimension": 1536,
  #    "status": {
  #       "ready": true,
  #       "state": "Ready"
  #    },
  #    "host": "docs-example-4mkljsz.svc.aped-4627-b74a.pinecone.io",
  #    "spec": {
  #       "pod": {
  #          "environment": "us-east-1-aws",
  #          "pod_type": "s1.x1",
  #          "pods": 1,
  #          "replicas": 1,
  #          "shards": 1
  #       }
  #    }
  # }
  ```
</CodeGroup>

<Warning>
  **Do not target an index by name in production.**

  When you target an index by name for data operations such as `upsert` and `query`, the SDK gets the unique DNS host for the index using the `describe_index` operation. This is convenient for testing but should be avoided in production because `describe_index` uses a different API than data operations and therefore adds an additional network call and point of failure. Instead, you should get an index host once and cache it for reuse or specify the host directly.
</Warning>

## Delete a pod-based index

Use the [`delete_index`](/reference/api/latest/control-plane/delete_index) operation to delete a pod-based index and all of its associated resources.

<Note>
  You are billed for a pod-based index even when it is not in use.
</Note>

<CodeGroup>
  ```python Python theme={null}
  # pip install "pinecone[grpc]"
  from pinecone.grpc import PineconeGRPC as Pinecone, PodSpec

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.delete_index(name="docs-example")
  ```

  ```javascript JavaScript theme={null}
  // npm install @pinecone-database/pinecone
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({
    apiKey: 'YOUR_API_KEY'
  });

  await pc.deleteIndex('docs-example');
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;

  public class DeleteIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          pc.deleteIndex("docs-example");
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      indexName := "docs-example"

      err = pc.DeleteIndex(ctx, indexName)
      if err != nil {
          log.Fatalf("Failed to delete index: %v", err)
      } else {
          fmt.Println("Index \"%v\" deleted successfully", indexName)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  await pinecone.DeleteIndexAsync("docs-example");
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X DELETE "https://api.pinecone.io/indexes/docs-example" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

If deletion protection is enabled on an index, requests to delete it will fail and return a `403 - FORBIDDEN` status with the following error:

```
Deletion protection is enabled for this index. Disable deletion protection before retrying.
```

Before you can delete such an index, you must first [disable deletion protection](/guides/manage-data/manage-indexes#configure-deletion-protection).

<Tip>
  You can delete an index using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/indexes). For the index you want to delete, click the three dots to the right of the index name, then click **Delete**.
</Tip>

## Selective metadata indexing

For pod-based indexes, Pinecone indexes all metadata fields by default. When metadata fields contains many unique values, pod-based indexes will consume significantly more memory, which can lead to performance issues, pod fullness, and a reduction in the number of possible vectors that fit per pod.

To avoid indexing high-cardinality metadata that is not needed for [filtering your queries](/guides/index-data/indexing-overview#metadata) and keep memory utilization low, specify which metadata fields to index using the `metadata_config` parameter.

<Note>
  Since high-cardinality metadata does not cause high memory utilization in serverless indexes, selective metadata indexing is not supported.
</Note>

The value for the `metadata_config` parameter is a JSON object containing the names of the metadata fields to index.

```JSON JSON theme={null}
{
    "indexed": [
        "metadata-field-1",
        "metadata-field-2",
        "metadata-field-n"
    ]
}
```

**Example**

The following example creates a pod-based index that only indexes the `genre` metadata field. Queries against this index that filter for the `genre` metadata field may return results; queries that filter for other metadata fields behave as though those fields do not exist.

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone, PodSpec

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.create_index(
    name="docs-example",
    dimension=1536,
    metric="cosine",
    spec=PodSpec(
      environment="us-west1-gcp",
      pod_type="p1.x1",
      pods=1,
      metadata_config = {
        "indexed": ["genre"]
      }
    ),
    deletion_protection="disabled"

  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({
    apiKey: 'YOUR_API_KEY'
  });

  await pc.createIndex({
    name: 'docs-example',
    dimension: 1536,
    metric: 'cosine',
    spec: {
      pod: {
        environment: 'us-west1-gcp',
        podType: 'p1.x1',
        pods: 1,
        metadata_config: {
          indexed: ["genre"]
        }
      }
    },
    deletionProtection: 'disabled',
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.IndexModel;
  import org.openapitools.db_control.client.model.DeletionProtection;

  public class CreateIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          CreateIndexRequestSpecPodMetadataConfig podSpecMetadataConfig = new CreateIndexRequestSpecPodMetadataConfig();
          List<String> indexedItems = Arrays.asList("genre", "year");
          podSpecMetadataConfig.setIndexed(indexedItems);
          pc.createPodsIndex("docs-example", 1536, "us-west1-gcp",
                  "p1.x1", "cosine", podSpecMetadataConfig, DeletionProtection.DISABLED);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      podIndexMetadata := &pinecone.PodSpecMetadataConfig{
          Indexed: &[]string{"genre"},
      }

      indexName := "docs-example"
    	metric := pinecone.Dotproduct
  	deletionProtection := pinecone.DeletionProtectionDisabled

  	idx, err := pc.CreatePodIndex(ctx, &pinecone.CreatePodIndexRequest{
          Name:               indexName,
          Metric:             &metric,
          Dimension:          1536,
          Environment:        "us-east1-gcp",
          PodType:            "p1.x1",
          DeletionProtection: &deletionProtection,
    	})
    	if err != nil {
          log.Fatalf("Failed to create pod-based index: %v", idx.Name)
      } else {
          fmt.Printf("Successfully created pod-based index: %v", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var createIndexRequest = await pinecone.CreateIndexAsync(new CreateIndexRequest
  {
      Name = "docs-example",
      Dimension = 1536,
      Metric = MetricType.Cosine,
      Spec = new PodIndexSpec
      {
          Pod = new PodSpec
          {
              Environment = "us-east1-gcp",
              PodType = "p1.x1",
              Pods = 1,
              MetadataConfig = new PodSpecMetadataConfig
              {
                  Indexed = new List<string> { "genre" },
              },
          }
      },
      DeletionProtection = DeletionProtection.Disabled
  });
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s https://api.pinecone.io/indexes \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
           "name": "docs-example",
           "dimension": 1536,
           "metric": "cosine",
           "spec": {
              "pod": {
                 "environment": "us-west1-gcp",
                 "pod_type": "p1.x1",
                 "pods": 1,
                 "metadata_config": {
                    "indexed": ["genre"]
                 }
              }
           },
           "deletion_protection": "disabled"
        }'
  ```
</CodeGroup>

## Prevent index deletion

<Note>
  This feature requires [Pinecone API version](/reference/api/versioning) `2024-07`, [Python SDK](/reference/sdks/python/overview) v5.0.0, [Node.js SDK](/reference/sdks/node/overview) v3.0.0, [Java SDK](/reference/sdks/java/overview) v2.0.0, or [Go SDK](/reference/sdks/go/overview) v1.0.0 or later.
</Note>

You can prevent an index and its data from accidental deleting when [creating a new index](/guides/index-data/create-an-index) or when [configuring an existing index](/guides/indexes/pods/manage-pod-based-indexes). In both cases, you set the `deletion_protection` parameter to `enabled`.

To enable deletion protection when creating a new index:

<CodeGroup>
  ```python Python theme={null}
  # pip install "pinecone[grpc]"
  from pinecone.grpc import PineconeGRPC as Pinecone, PodSpec

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.create_index(
    name="docs-example",
    dimension=1536,
    metric="cosine",
    spec=PodSpec(
      environment="us-west1-gcp",
      pod_type="p1.x1",
      pods=1
    ),
      deletion_protection="enabled"
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({
    apiKey: 'YOUR_API_KEY'
  });

  await pc.createIndex({
    name: 'docs-example',
    dimension: 1536,
    metric: 'cosine',
    spec: {
      pod: {
        environment: 'us-west1-gcp',
        podType: 'p1.x1',
        pods: 1
      }
    },
    deletionProtection: 'enabled',
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.IndexModel;
  import org.openapitools.db_control.client.model.DeletionProtection;

  public class CreateIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          pc.createPodsIndex("docs-example", 1536, "us-west1-gcp",
                  "p1.x1", "cosine", DeletionProtection.ENABLED);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      indexName := "docs-example"
    	metric := pinecone.Dotproduct
  	deletionProtection := pinecone.DeletionProtectionDisabled

  	idx, err := pc.CreatePodIndex(ctx, &pinecone.CreatePodIndexRequest{
          Name:               indexName,
          Metric:             &metric,
          Dimension:          1536,
          Environment:        "us-east1-gcp",
          PodType:            "p1.x1",
          DeletionProtection: &deletionProtection,
    	})
    	if err != nil {
          log.Fatalf("Failed to create pod-based index: %v", idx.Name)
      } else {
          fmt.Printf("Successfully created pod-based index: %v", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var createIndexRequest = await pinecone.CreateIndexAsync(new CreateIndexRequest
  {
      Name = "docs-example",
      Dimension = 1536,
      Metric = MetricType.Cosine,
      Spec = new PodIndexSpec
      {
          Pod = new PodSpec
          {
              Environment = "us-east1-gcp",
              PodType = "p1.x1",
              Pods = 1,
          }
      },
      DeletionProtection = DeletionProtection.Enabled
  });
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s "https://api.pinecone.io/indexes" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
           "name": "docs-example",
           "dimension": 1536,
           "metric": "cosine",
           "spec": {
              "pod": {
                 "environment": "us-west1-gcp",
                 "pod_type": "p1.x1",
                 "pods": 1
              }
           },
           "deletion_protection": "enabled"
        }'
  ```
</CodeGroup>

To enable deletion protection when configuring an existing index:

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.configure_index(
     name="docs-example", 
     deletion_protection="enabled"
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const client = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await client.configureIndex('docs-example', { deletionProtection: 'enabled' });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.*;

  public class ConfigureIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          pc.configurePodsIndex("docs-example", DeletionProtection.ENABLED);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idx, err := pc.ConfigureIndex(ctx, "docs-example", pinecone.ConfigureIndexParams{DeletionProtection: "enabled"})
    	if err != nil {
          log.Fatalf("Failed to configure index \"%v\": %v", idx.Name, err)
      } else {
          fmt.Printf("Successfully configured index \"%v\"", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var indexMetadata = await pinecone.ConfigureIndexAsync("docs-example", new ConfigureIndexRequest
  {
      DeletionProtection = DeletionProtection.Enabled,
  });
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s -X PATCH "https://api.pinecone.io/indexes/docs-example" \
      -H "Content-Type: application/json" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
          "deletion_protection": "enabled"
          }'
  ```
</CodeGroup>

When deletion protection is enabled on an index, requests to delete the index fail and return a `403 - FORBIDDEN` status with the following error:

```
Deletion protection is enabled for this index. Disable deletion protection before retrying.
```

## Disable deletion protection

Before you can [delete an index](#delete-a-pod-based-index) with deletion protection enabled, you must first disable deletion protection as follows:

<CodeGroup>
  ```python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.configure_index(
     name="docs-example", 
     deletion_protection="disabled"
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const client = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await client.configureIndex('docs-example', { deletionProtection: 'disabled' });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.*;

  public class ConfigureIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          pc.configurePodsIndex("docs-example", DeletionProtection.DISABLED);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idx, err := pc.ConfigureIndex(ctx, "docs-example", pinecone.ConfigureIndexParams{DeletionProtection: "disabled"})
    	if err != nil {
          log.Fatalf("Failed to configure index \"%v\": %v", idx.Name, err)
      } else {
          fmt.Printf("Successfully configured index \"%v\"", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var configureIndexRequest = await pinecone.ConfigureIndexAsync("docs-example", new ConfigureIndexRequest
  {
      DeletionProtection = DeletionProtection.Disabled,
  });
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s -X PATCH "https://api.pinecone.io/indexes/docs-example" \
      -H "Content-Type: application/json" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
          "deletion_protection": "disabled"
          }'
  ```
</CodeGroup>

## Delete an entire namespace

In pod-based indexes, reads and writes share compute resources, so deleting an entire namespace with many records can increase the latency of read operations. In such cases, consider [deleting records in batches](#delete-records-in-batches).

## Delete records in batches

In pod-based indexes, reads and writes share compute resources, so deleting an entire namespace or a large number of records can increase the latency of read operations. To avoid this, delete records in batches of up to 1000, with a brief sleep between requests. Consider using smaller batches if the index has active read traffic.

<CodeGroup>
  ```python Batch delete a namespace theme={null}
  from pinecone import Pinecone
  import numpy as np
  import time

  pc = Pinecone(api_key='API_KEY')

  INDEX_NAME = 'INDEX_NAME'
  NAMESPACE = 'NAMESPACE_NAME'
  # Consider using smaller batches if you have a high RPS for read operations
  BATCH = 1000

  index = pc.Index(name=INDEX_NAME)
  dimensions = index.describe_index_stats()['dimension']

  # Create the query vector
  query_vector = np.random.uniform(-1, 1, size=dimensions).tolist()
  results = index.query(vector=query_vector, namespace=NAMESPACE, top_k=BATCH)

  # Delete in batches until the query returns no results
  while len(results['matches']) > 0:
      ids = [i['id'] for i in results['matches']]
      index.delete(ids=ids, namespace=NAMESPACE)
      time.sleep(0.01)
      results = index.query(vector=query_vector, namespace=NAMESPACE, top_k=BATCH)
  ```

  ```python Batch delete by metadata theme={null}
  from pinecone import Pinecone
  import numpy as np
  import time

  pc = Pinecone(api_key='API_KEY')

  INDEX_NAME = 'INDEX_NAME'
  NAMESPACE = 'NAMESPACE_NAME'
  # Consider using smaller batches if you have a high RPS for read operations
  BATCH = 1000

  index = pc.Index(name=INDEX_NAME)
  dimensions = index.describe_index_stats()['dimension']

  METADATA_FILTER = {}

  # Create the query vector with a filter
  query_vector = np.random.uniform(-1, 1, size=dimensions).tolist()
  results = index.query(vector=query_vector, namespace=NAMESPACE, filter=METADATA_FILTER, top_k=BATCH)

  # Delete in batches until the query returns no results
  while len(results['matches']) > 0:
      ids = [i['id'] for i in results['matches']]
      index.delete(ids=ids, namespace=NAMESPACE)
      time.sleep(0.01)
      results = index.query(vector=query_vector, namespace=NAMESPACE, filter=METADATA_FILTER, top_k=BATCH)
  ```
</CodeGroup>

## Delete records by metadata

<Note>
  In pod-based indexes, if you are targeting a large number of records for deletion and the index has active read traffic, consider [deleting records in batches](#delete-records-in-batches).
</Note>

To delete records from a namespace based on their metadata values, pass a [metadata filter expression](/guides/index-data/indexing-overview#metadata-filter-expressions) to the `delete` operation. This deletes all records in the namespace that match the filter expression.

For example, the following code deletes all records with a `genre` field set to `documentary` from namespace `example-namespace`:

<CodeGroup>
  ```Python Python theme={null}
  # pip install "pinecone[grpc]"
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  index.delete(
      filter={
          "genre": {"$eq": "documentary"}
      },
      namespace="example-namespace" 
  )
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  const ns = index.namespace('example-namespace')

  await ns.deleteMany({
    genre: { $eq: "documentary" },
  });
  ```

  ```java Java theme={null}
  import com.google.protobuf.Struct;
  import com.google.protobuf.Value;
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;

  import java.util.Arrays;
  import java.util.List;

  public class DeleteExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");
          Struct filter = Struct.newBuilder()
                  .putFields("genre", Value.newBuilder()
                          .setStructValue(Struct.newBuilder()
                                  .putFields("$eq", Value.newBuilder()
                                          .setStringValue("documentary")
                                          .build()))
                          .build())
                  .build();
          index.deleteByFilter(filter, "example-namespace");
          
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      metadataFilter := map[string]interface{}{
  		"genre": map[string]interface{}{
  			"$eq": "documentary",
  		},
      }

      filter, err := structpb.NewStruct(metadataFilter)
      if err != nil {
          log.Fatalf("Failed to create metadata filter: %v", err)
      }

      err = idxConnection.DeleteVectorsByFilter(ctx, filter)
      if err != nil {
          log.Fatalf("Failed to delete vector(s) with filter %+v: %v", filter, err)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var deleteResponse = await index.DeleteAsync(new DeleteRequest {
      Namespace = "example-namespace",
      Filter = new Metadata
      {
          ["genre"] =
              new Metadata
              {
                  ["$eq"] = "documentary"
              }
      }
  });
  ```

  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl -i "https://$INDEX_HOST/vectors/delete" \
    -H 'Api-Key: $PINECONE_API_KEY' \
    -H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "filter": {"genre": {"$eq": "documentary"}},
      "namespace": "example-namespace"
    }'
  ```
</CodeGroup>

## Tag an index

When configuring an index, you can tag the index to help with index organization and management. For more details, see [Tag an index](/guides/manage-data/manage-indexes#configure-index-tags).

## Manage costs

### Set a project pod limit

To control costs, [project owners](/guides/projects/understanding-projects#project-roles) can [set the maximum total number of pods](/reference/api/database-limits#pods-per-project) allowed across all pod-based indexes in a project. The default pod limit is 5.

<Tabs>
  <Tab title="Pinecone console">
    1. Go to [Settings > Projects](https://app.pinecone.io/organizations/-/settings/projects).
    2. For the project you want to update, click the **ellipsis (...) menu > Configure**.
    3. In the **Pod Limit** section, update the number of pods.
    4. Click **Save Changes**.
  </Tab>

  <Tab title="API">
    <token />

    ```bash curl theme={null}
    PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"
    PROJECT_ID="YOUR_PROJECT_ID"

    curl -X PATCH "https://api.pinecone.io/admin/projects/$PROJECT_ID" \
      -H "accept: application/json" \
      -H "Content-Type: application/json" \
    	-H "X-Pinecone-Api-Version: 2025-10" \
      -d '{
        "max_pods": 5
        }'
    ```

    The example returns a response like the following:

    ```json  theme={null}
    {
      "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
      "name": "example-project",
      "max_pods": 5,
      "force_encryption_with_cmek": false,
      "organization_id": "string",
      "created_at": "2025-03-17T00:42:31.912Z"
    }
    ```
  </Tab>
</Tabs>

### Back up inactive pod-based indexes

For each pod-based index, billing is determined by the per-minute price per pod and the number of pods the index uses, regardless of index activity. When a pod-based index is not in use, [back it up using collections](/guides/indexes/pods/back-up-a-pod-based-index) and delete the inactive index. When you're ready to use the vectors again, you can [create a new index from the collection](/guides/indexes/pods/create-a-pod-based-index#create-a-pod-index-from-a-collection). This new index can also use a different index type or size. Because it's relatively cheap to store collections, you can reduce costs by only running an index when it's in use.

### Choose the right index type and size

Pod sizes are designed for different applications, and some are more expensive than others. [Choose the appropriate pod type and size](/guides/indexes/pods/choose-a-pod-type-and-size), so you pay for the resources you need. For example, the `s1` pod type provides large storage capacity and lower overall costs with slightly higher query latencies than `p1` pods. By switching to a different pod type, you may be able to reduce costs while still getting the performance your application needs.

<Note>
  For pod-based indexes, project owners can [set limits for the total number of pods](/reference/api/database-limits#pods-per-project) across all indexes in the project. The default pod limit is 5.
</Note>

## Monitor performance

Pinecone generates time-series performance metrics for each Pinecone index. You can monitor these metrics directly in the Pinecone console or with tools like Prometheus or Datadog.

### Use the Pinecone Console

To view performance metrics in the Pinecone console:

1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).
2. Select the project containing the index you want to monitor.
3. Go to **Database > Indexes**.
4. Select the index.
5. Go to the **Metrics** tab.

### Use Datadog

To monitor Pinecone with Datadog, use Datadog's [Pinecone integration](/integrations/datadog).

<Note>
  This feature is available on [Standard and Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

### Use Prometheus

<Note>
  This feature is available on [Standard and Enterprise plans](https://www.pinecone.io/pricing/). When using [Bring Your Own Cloud](/guides/production/bring-your-own-cloud), you must configure Prometheus monitoring within your VPC.
</Note>

To monitor all pod-based indexes in a specific region of a project, insert the following snippet into the [`scrape_configs`](https://prometheus.io/docs/prometheus/latest/configuration/configuration/#scrape_config) section of your `prometheus.yml` file and update it with values for your Prometheus integration:

```YAML  theme={null}
scrape_configs:
  - job_name: "pinecone-pod-metrics"
    scheme: https
    metrics_path: '/metrics'
    authorization:
      credentials: API_KEY
    static_configs:
      - targets: ["metrics.ENVIRONMENT.pinecone.io" ]
```

* Replace `API_KEY` with an API key for the project you want to monitor. If necessary, you can [create an new API key](/reference/api/authentication) in the Pinecone console.

* Replace `ENVIRONMENT` with the [environment](/guides/indexes/pods/understanding-pod-based-indexes#pod-environments) of the pod-based indexes you want to monitor.

For more configuration details, see the [Prometheus docs](https://prometheus.io/docs/prometheus/latest/configuration/configuration/).

#### Available metrics

The following metrics are available when you integrate Pinecone with Prometheus:

| Name                                 | Type      | Description                                                                       |
| :----------------------------------- | :-------- | :-------------------------------------------------------------------------------- |
| `pinecone_vector_count`              | gauge     | The number of records per pod in the index.                                       |
| `pinecone_request_count_total`       | counter   | The number of data plane calls made by clients.                                   |
| `pinecone_request_error_count_total` | counter   | The number of data plane calls made by clients that resulted in errors.           |
| `pinecone_request_latency_seconds`   | histogram | The distribution of server-side processing latency for pinecone data plane calls. |
| `pinecone_index_fullness`            | gauge     | The fullness of the index on a scale of 0 to 1.                                   |

#### Metric labels

Each metric contains the following labels:

| Label          | Description                                                                                                                                    |
| :------------- | :--------------------------------------------------------------------------------------------------------------------------------------------- |
| `pid`          | Process identifier.                                                                                                                            |
| `index_name`   | Name of the index to which the metric applies.                                                                                                 |
| `project_name` | Name of the project containing the index.                                                                                                      |
| `request_type` | Type of request: `upsert`, `delete`, `fetch`, `query`, or `describe_index_stats`. This label is included only in `pinecone_request_*` metrics. |

#### Example queries

Return the average latency in seconds for all requests against the Pinecone index `docs-example`:

```shell  theme={null}
avg by (request_type) (pinecone_request_latency_seconds{index_name="docs-example"})
```

Return the vector count for the Pinecone index `docs-example`:

```shell  theme={null}
sum ((avg by (app) (pinecone_vector_count{index_name="docs-example"})))
```

Return the total number of requests against the Pinecone index `docs-example` over one minute:

```shell  theme={null}
sum by (request_type)(increase(pinecone_request_count_total{index_name="docs-example"}[60s]))
```

Return the total number of upsert requests against the Pinecone index `docs-example` over one minute:

```shell  theme={null}
sum by (request_type)(increase(pinecone_request_count_total{index_name="docs-example", request_type="upsert"}[60s]))
```

Return the total errors returned by the Pinecone index `docs-example` over one minute:

```shell  theme={null}
sum by (request_type) (increase(pinecone_request_error_count{
      index_name="docs-example"}[60s]))
```

Return the index fullness metric for the Pinecone index `docs-example`:

```
round(max (pinecone_index_fullness{index_name="docs-example"} * 100))
```

## Troubleshooting

### Index fullness errors

Serverless indexes automatically scale as needed.

However, pod-based indexes can run out of capacity. When that happens, upserting new records will fail with the following error:

```console console theme={null}
Index is full, cannot accept data.
```

### High-cardinality metadata and over-provisioning

This [Loom video walkthrough](https://www.loom.com/share/ce6f5dd0c3e14ba0b988fe32d96b703a?sid=48646dfe-c10c-4143-82c6-031fefe05a68) shows you how to manage two scenarios:

* The first scenario involves customers loading an index replete with high cardinality metadata. This can trigger a series of unforeseen challenges, and hence, it's vital to comprehend how to manage this situation effectively. This methodology can be applied whenever you need to change your metadata configuration.

* The second scenario that we will address involves customers who have over-provisioned the number of pods they need. More specifically, we will discuss the process of re-scaling an index in instances where the customer has previously scaled vertically and now desires to scale the index back down.

---

# Source: https://docs.pinecone.io/guides/projects/manage-project-members.md

# Source: https://docs.pinecone.io/guides/assistant/admin/manage-project-members.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage project members

> Add and manage team members in your project.

[Organization owners](/guides/assistant/admin/organizations-overview#organization-roles) or project owners can manage members in a project. Members can be added to a project with different [roles](/guides/assistant/admin/projects-overview#project-roles), which determine their permissions within the project.

<Tip>
  For information about managing members at the **organization-level**, see [Manage organization members](/guides/assistant/admin/manage-organization-members).
</Tip>

## Add members to a project

You can add members to a project in the [Pinecone console](https://app.pinecone.io/organizations/-/projects):

1. Select your project.
2. Go to the [**Manage > Access > Members** tab](https://app.pinecone.io/organizations/-/projects/-/access/members).
3. Enter the member's email address or name.
4. Select a [Project role](/guides/assistant/admin/projects-overview#project-roles) for the member. The role determines the member's permissions within Pinecone.
5. Click **Invite**.

When you invite a member to join your project, Pinecone sends them an email containing a link that enables them to gain access to the project. If they already have a Pinecone account, they still receive an email, but they can also immediately view the project.

## Change a member's role

You can change a member's role in the [Pinecone console](https://app.pinecone.io/organizations/-/projects):

1. Select your project.
2. Go to the [**Manage > Access > Members** tab](https://app.pinecone.io/organizations/-/projects/-/access/members).
3. In the row of the member you want to edit, click **ellipsis (...) menu > Edit role**.
4. Select a [Project role](/guides/assistant/admin/projects-overview#project-roles) for the member.
5. Click **Edit role**.

## Remove a member

You can remove a member from a project in the [Pinecone console](https://app.pinecone.io/organizations/-/projects):

1. Select your project.
2. Go to the [**Manage > Access > Members** tab](https://app.pinecone.io/organizations/-/projects/-/access/members).
3. In the row of the member you want to delete, click **ellipsis (...) menu > Remove member**.
4. Click **Remove Member**.

<Note>
  To remove yourself from a project, click the **Leave organization** button in your user's row and confirm.
</Note>

---

# Source: https://docs.pinecone.io/guides/assistant/admin/manage-project-service-accounts.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage service accounts at the project-level

> Enable programmatic access with project-level service accounts.

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

This page shows how [organization owners](/guides/assistant/admin/organizations-overview#organization-roles) and [project owners](/guides/assistant/admin/projects-overview#project-roles) can add and manage service accounts at the project-level. Service accounts enable programmatic access to Pinecone's Admin API, which can be used to create and manage projects and API keys.

## Add a service account to a project

After a service account has been [added to an organization](/guides/assistant/admin/manage-organization-service-accounts#create-a-service-account), it can be added to a project in the [Pinecone console](https://app.pinecone.io/organizations/-/projects):

1. Select your project.
2. Go to the [**Manage > Access > Service accounts** tab](https://app.pinecone.io/organizations/-/projects/-/access/service-accounts).
3. Select the service account to add.
4. Select a [**Project role**](/guides/assistant/admin/projects-overview#project-roles) for the service account. The role determines its permissions within Pinecone.
5. Click **Connect**.

## Change project role

To change a service account's role in the [Pinecone console](https://app.pinecone.io/organizations/-/projects):

1. Select your project.
2. Go to the [**Manage > Access > Service accounts** tab](https://app.pinecone.io/organizations/-/projects/-/access/service-accounts).
3. In the row of the service account you want to edit, click **ellipsis (...) menu > Edit role**.
4. Select a [**Project role**](/guides/projects/understanding-projects#project-roles) for the service account.
5. Click **Edit role**.

## Remove a service account from a project

To remove a service account from a project in the [Pinecone console](https://app.pinecone.io/organizations/-/projects):

1. Select your project.
2. Go to the [**Manage > Access > Service accounts** tab](https://app.pinecone.io/organizations/-/projects/-/access/service-accounts).
3. In the row of the service account you want to remove, click **ellipsis (...) menu > Disconnect**.
4. Enter the service account name to confirm.
5. Click **Disconnect**.

---

# Source: https://docs.pinecone.io/guides/projects/manage-projects.md

# Source: https://docs.pinecone.io/guides/assistant/admin/manage-projects.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage projects

> View, rename, and delete projects in your organization.

This page shows you how to view project details, rename a project, and delete a project.

<Note>
  You must be an [organization owner](/guides/assistant/admin/organizations-overview#organization-roles) or [project owner](/guides/assistant/admin/projects-overview#project-roles) to edit project details or delete a project.
</Note>

## View project details

You can view the details of a project, as in the following example:

<Note>
  An [access token](/guides/assistant/admin/manage-organization-service-accounts#retrieve-an-access-token) must be provided to complete this action through the Admin API. The Admin API is in [public preview](/assistant-release-notes/feature-availability).
</Note>

<CodeGroup>
  ```bash curl theme={null}
  PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"
  PROJECT_ID="32c8235a-5220-4a80-a9f1-69c24109e6f2"

  curl -X GET "https://api.pinecone.io/admin/projects/$PROJECT_ID" \
       -H "Authorization: Bearer $PINECONE_ACCESS_TOKEN" \
       -H "X-Pinecone-Api-Version: 2025-10" \
       -H "accept: application/json"
  ```

  ```bash CLI theme={null}
  PROJECT_ID="32c8235a-5220-4a80-a9f1-69c24109e6f2"

  # Target the organization that contains the project.
  pc target -o "example-org" 
  # Fetch the project details.
  pc project describe -i $PROJECT_ID
  ```
</CodeGroup>

The example returns a response like the following:

<CodeGroup>
  ```json curl theme={null}
  {
    "id": "32c8235a-5220-4a80-a9f1-69c24109e6f2",
    "name": "example-project",
    "max_pods": 5,
    "force_encryption_with_cmek": false,
    "organization_id": "-NM7af6f234168c4e44a",
    "created_at": "2025-10-27T23:27:46.370088Z"
  }
  ```

  ```text CLI theme={null}
  ATTRIBUTE           VALUE
  Name                example-project
  ID                  32c8235a-5220-4a80-a9f1-69c24109e6f2
  Organization ID     -NM7af6f234168c4e44a
  Created At          2025-10-27 23:27:46.370088 +0000 UTC
  Force Encryption    false
  Max Pods            5
  ```
</CodeGroup>

<Tip>
  You can view project details using the [Pinecone console](https://app.pinecone.io/organizations/-/settings/projects/-/indexes).
</Tip>

## Rename a project

You can change the name of your project:

<Tabs>
  <Tab title="Pinecone console">
    1. In the Pinecone console, go to [**Settings > Projects**](https://app.pinecone.io/organizations/-/settings/projects).

    2. Click the **ellipsis (...) menu > Configure** icon next to the project you want to update.

    3. Enter a new **Project Name**.

       <Note>
         A project name can contain up to 512 characters.
       </Note>

    4. Click **Save Changes**.
  </Tab>

  <Tab title="Code">
    <Note>
      An [access token](/guides/assistant/admin/manage-organization-service-accounts#retrieve-an-access-token) must be provided to complete this action through the Admin API. The Admin API is in [public preview](/assistant-release-notes/feature-availability).
    </Note>

    <CodeGroup>
      ```bash curl theme={null}
      PROJECT_ID="YOUR_PROJECT_ID"
      PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"

      curl -X PATCH "https://api.pinecone.io/admin/projects/$PROJECT_ID" \
           -H "accept: application/json" \
           -H "Content-Type: application/json" \
           -H "X-Pinecone-Api-Version: 2025-10" \
           -d '{
                 "name": "updated-example-project"
               }'
      ```

      ```bash CLI theme={null}
      PROJECT_ID="YOUR_PROJECT_ID"

      # Target the project to update.
      pc target -o "example-org" "example-project"
      # Update the project name.
      pc project update -i $PROJECT_ID -n "updated-example-project"
      ```
    </CodeGroup>

    The example returns a response like the following:

    <CodeGroup>
      ```json curl theme={null}
      {
        "id": "32c8235a-5220-4a80-a9f1-69c24109e6f2",
        "name": "updated-example-project",
        "max_pods": 5,
        "force_encryption_with_cmek": false,
        "organization_id": "-NM7af6f234168c4e44a",
        "created_at": "2025-10-27T23:27:46.370088Z"
      }
      ```

      ```text CLI theme={null}
      [SUCCESS] Project 32c8235a-5220-4a80-a9f1-69c24109e6f2 updated successfully.
      ATTRIBUTE           VALUE
      Name                updated-example-project
      ID                  32c8235a-5220-4a80-a9f1-69c24109e6f2
      Organization ID     -NM7af6f234168c4e44a
      Created At          2025-10-27 23:27:46.370088 +0000 UTC
      Force Encryption    false
      Max Pods            5
      ```
    </CodeGroup>
  </Tab>
</Tabs>

## Add project tags

Project tags are key-value pairs that you can use to categorize and identify a project.

To add project tags, use the Pinecone console.

1. In the Pinecone console, go to [**Settings > Projects**](https://app.pinecone.io/organizations/-/settings/projects).
2. Click the **ellipsis (...) menu > Configure** icon next to the project you want to update.
3. Click **+ Add tag** and enter a tag key and value. Repeat for each tag you want to add.
4. Click **Save Changes**.

<Tip>
  You can also [add tags to indexes](/guides/manage-data/manage-indexes#configure-index-tags).
</Tip>

## Delete a project

To delete a project, you must first [delete all data](/guides/manage-data/delete-data), [indexes](/guides/manage-data/manage-indexes#delete-an-index), [collections](/guides/indexes/pods/back-up-a-pod-based-index#delete-a-collection), [backups](/guides/manage-data/back-up-an-index#delete-a-backup) and [assistants](/guides/assistant/manage-assistants#delete-an-assistant) associated with the project. Then, you can delete the project itself:

<Tabs>
  <Tab title="Pinecone console">
    1. In the Pinecone console, go to [**Settings > Projects**](https://app.pinecone.io/organizations/-/settings/projects).
    2. For the project you want to delete, click the **ellipsis (...) menu > Delete**.
    3. Enter the project name to confirm the deletion.
    4. Click **Delete Project**.
  </Tab>

  <Tab title="Code">
    <Note>
      An [access token](/guides/assistant/admin/manage-organization-service-accounts#retrieve-an-access-token) must be provided to complete this action through the Admin API. The Admin API is in [public preview](/assistant-release-notes/feature-availability).
    </Note>

    <CodeGroup>
      ```bash curl theme={null}
      PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"
      PROJECT_ID="32c8235a-5220-4a80-a9f1-69c24109e6f2"

      curl -X DELETE "https://api.pinecone.io/admin/projects/$PROJECT_ID" \
           -H "X-Pinecone-Api-Version: 2025-10" \
           -H "Authorization: Bearer $PINECONE_ACCESS_TOKEN"
      ```

      ```bash CLI theme={null}
      PINECONE_PROJECT_ID="32c8235a-5220-4a80-a9f1-69c24109e6f2"

      # Target the organization that contains the project.
      pc target -o "example-org" 
      # Delete the project. Use --skip-confirmation to skip 
      # the confirmation prompt.
      pc project delete -i $PINECONE_PROJECT_ID
      ```
    </CodeGroup>
  </Tab>
</Tabs>

---

# Source: https://docs.pinecone.io/guides/projects/manage-service-accounts.md

# Source: https://docs.pinecone.io/guides/organizations/manage-service-accounts.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage service accounts at the organization-level

> Create service accounts for organization-level API access.

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

This page shows how [organization owners](/guides/organizations/understanding-organizations#organization-roles) can add and manage service accounts at the organization-level. Service accounts enable programmatic access to Pinecone's Admin API, which can be used to create and manage projects and API keys.

<Tip>
  Once a service account is added at the organization-level, it can be added to a project. For more information, see [Manage service accounts at the project-level](/guides/projects/manage-service-accounts).
</Tip>

## Create a service account

You can create a service account in the [Pinecone console](https://app.pinecone.io):

1. Go to [**Settings > Access > Service accounts**](https://app.pinecone.io/organizations/-/settings/access/service-accounts).

2. Enter a **Name** for the service account.

3. Choose an [**Organization Role**](/guides/organizations/understanding-organizations#organization-roles) for the service account. The role determines the service account's permissions within Pinecone.

4. Click **Create**.

5. Copy and save the **Client secret** in a secure place for future use. You will need the client secret to retrieve an access token.

   <Warning>
     You will not be able to see the client secret again after you close the dialog.
   </Warning>

6. Click **Close**.

Once you have created a service account, [add it to a project](/guides/projects/manage-service-accounts#add-a-service-account-to-a-project) to allow it access to the project's resources.

## Retrieve an access token

To access the Admin API, you must provide an access token to authenticate. Retrieve the access token using the client secret of a service account, which was [provided at time of creation](#create-a-service-account).

You can retrieve an access token for a service account from the `https://login.pinecone.io/oauth/token` endpoint, as shown in the following example:

```bash curl theme={null}
curl "https://login.pinecone.io/oauth/token" \ # Note: Base URL is login.pinecone.io
	-H "X-Pinecone-Api-Version: 2025-10" \
	-H "Content-Type: application/json" \
	-d '{
		"grant_type": "client_credentials",
		"client_id": "YOUR_CLIENT_ID",
		"client_secret": "YOUR_CLIENT_SECRET",
		"audience": "https://api.pinecone.io/"
	}'
```

The response will include an `access_token` field, which you can use to authenticate with the Admin API.

```
{
    "access_token":"YOUR_ACCESS_TOKEN",
    "expires_in":86400,
    "token_type":"Bearer"
}
```

## Change a service account's role

You can change a service account's role in the [Pinecone console](https://app.pinecone.io):

1. Go to [**Settings > Access > Service accounts**](https://app.pinecone.io/organizations/-/settings/service-accounts).
2. In the row of the service account you want to update, click **ellipsis (...) menu > Manage**.
3. Select an [**Organization role**](/guides/organizations/understanding-organizations#organization-roles) for the service account.
4. Click **Update**.

## Update service account name

You can change a service account's name in the [Pinecone console](https://app.pinecone.io):

1. Go to [**Settings > Access > Service accounts**](https://app.pinecone.io/organizations/-/settings/service-accounts).
2. In the row of the service account you want to update, click **ellipsis (...) menu > Manage**.
3. Enter a new **Service account name**.
4. Click **Update**.

## Rotate a service account's secret

You can rotate a service account's client secret in the [Pinecone console](https://app.pinecone.io):

1. Go to [**Settings > Access > Service accounts**](https://app.pinecone.io/organizations/-/settings/service-accounts).

2. In the row of the service account you want to update, click **ellipsis (...) menu > Rotate secret**.

3. **Enter the service account name** to confirm.

4. Click **Rotate client secret**.

5. Copy and save the **Client secret** in a secure place for future use.

   <Warning>
     You will not be able to see the client secret again after you close the dialog.
   </Warning>

6. Click **Close**.

## Delete a service account

Deleting a service account will remove it from all projects and will disrupt any applications using it to access Pinecone. You delete a service account in the [Pinecone console](https://app.pinecone.io):

1. Go to [**Settings > Access > Service accounts**](https://app.pinecone.io/organizations/-/settings/service-accounts).
2. In the row of the service account you want to update, click **ellipsis (...) menu > Delete**.
3. **Enter the service account name** to confirm.
4. Click **Delete service account**.

---

# Source: https://docs.pinecone.io/guides/operations/integrations/manage-storage-integrations.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage storage integrations

> Update and manage cloud storage integrations.

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Standard and Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

This page shows you how to manage storage integrations for your Pinecone project.

To set up cloud storage for integration with Pinecone, see the following guides:

* [Integrate with Amazon S3](/guides/operations/integrations/integrate-with-amazon-s3)
* [Integrate with Google Cloud Storage](/guides/operations/integrations/integrate-with-google-cloud-storage)
* [Integrate with Azure Blob Storage](/guides/operations/integrations/integrate-with-azure-blob-storage)

## Update an integration

To update information for a storage integration through the [Pinecone console](https://app.pinecone.io/organizations/-/projects), take the following steps:

1. Select your project.
2. Go to [**Manage > Storage integrations**](https://app.pinecone.io/organizations/-/projects/-/storage).
3. For the integration you want to update, click the *...* (Actions) icon.
4. Click **Manage**.
5. Update the integration details as needed.
6. Click **Add integration**.

## Delete an integration

To delete a storage integration through the [Pinecone console](https://app.pinecone.io/organizations/-/projects), take the following steps:

1. Select your project.
2. Go to [**Manage > Storage integrations**](https://app.pinecone.io/organizations/-/projects/-/storage).
3. For the integration you want to update, click the *...* (Actions) icon.
4. Click **Delete**.
5. Enter the integration name.
6. Click **Confirm deletion**.

---

# Source: https://docs.pinecone.io/integrations/matillion.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Matillion

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

[Matillion Data Productivity Cloud](https://www.matillion.com/) is a unified platform that helps your team move faster with one central place to build and manage graphical, low-code data pipelines. It allows data teams to use structured, semi-structured, and unstructured data in analytics; build AI pipelines for new use cases; and be more productive.

Matillion Data Productivity Cloud and Pinecone can be used together for retrieval augmented generation (RAG) use cases, helping to contextualize business insights without code.

Matillion supports 150+ pre-built data source connectors, as well as the ability to build custom connectors to any REST API source system, making it easy to chunk unstructured datasets, create embeddings, and upsert to Pinecone.
Matillion's graphical AI Prompt Components integrate with large language models (LLM) running in OpenAI, Amazon Bedrock, Azure OpenAI, and Snowpark Container Services. They enable no-code lookup of external knowledge stored in Pinecone, enabling data engineers to enrich GenAI answers with contextualized and proprietary data.

<PrimarySecondaryCTA primaryLabel={"Get started"} primaryHref={"https://docs.matillion.com/data-productivity-cloud/designer/docs/pinecone-vector-upsert/"} />

## Additional resources

* Video: [Use RAG with a Pinecone Vector database on the Data Productivity Cloud](https://www.youtube.com/watch?v=BsH7WlJdoFs)
* Video: [How to upsert to your Pinecone Vector database](https://www.youtube.com/watch?v=l9qt-EzLkgY)
* [Unlock the power of AI in Data Engineering](https://www.matillion.com/blog/matillion-new-ai-capabilities-for-data-engineering)

---

# Source: https://docs.pinecone.io/guides/operations/mcp-server.md

# Source: https://docs.pinecone.io/guides/assistant/mcp-server.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Use an Assistant MCP server

> Connect AI agents to Pinecone Assistant via Model Context Protocol.

<Note>
  This feature is in [early access](/release-notes/feature-availability) and is not intended for production usage.
</Note>

Every Pinecone Assistant has a dedicated MCP server that gives AI agents direct access to the assistant's knowledge through the standardized [Model Context Protocol (MCP)](https://modelcontextprotocol.io/). This page shows you how to connect an assistant's MCP server with Cursor, Claude Desktop, and LangChain.

There are two ways to connect to an assistant MCP server:

* [Remote MCP server](#remote-mcp-server) - Use a dedicated MCP endpoint to connect directly to an assistant.
* [Local MCP server](#local-mcp-server) - Run a Docker container locally that connects to an assistant

Both options support a context tool that allows agents to retrieve relevant context snippets from your assistant's knowledge. This is similar to the [context API](/guides/assistant/retrieve-context-snippets) but fine-tuned for MCP clients. Additional capabilities, such as file access, will be added in future releases.

## Remote MCP server

Every Pinecone Assistant has a dedicated MCP endpoint that you can connect directly to your AI applications. This option doesn't require running any infrastructure and is managed by Pinecone.

The MCP endpoint for an assistant is:

```
https://<YOUR_PINECONE_ASSISTANT_HOST>/mcp/assistants/<YOUR_ASSISTANT_NAME>
```

<Warning>
  The previous SSE-based endpoint (with `/sse` suffix) is deprecated and will stop working on August 31, 2025 at 11:59:59 PM UTC. Before then, update to the [streamable HTTP transport](https://modelcontextprotocol.io/specification/2025-03-26/basic/transports#streamable-http) MCP endpoint shown above, which implements the current MCP specification and provides improved flexibility and compatibility.
</Warning>

### Prerequisites

Before you begin, make sure you have the following values, which you'll use in the commands below:

* `<YOUR_PINECONE_API_KEY>`: [A Pinecone API key](/guides/projects/manage-api-keys).
* `<YOUR_PINECONE_ASSISTANT_HOST>`: In the Pinecone console, this is your assistant's **Host** value.
* `<YOUR_ASSISTANT_NAME>`: Your assistant's name, as displayed in the Pinecone console. For example, `example-assistant`.

### Use with Claude Code

You can use the Claude CLI to configure Claude Code to use your assistant's remote MCP server. For more information, see [Claude Code's MCP documentation](https://docs.anthropic.com/en/docs/claude-code/mcp).

1. Add the MCP server using the Claude CLI:

   ```bash  theme={null}
   claude mcp add --transport http my-assistant https://<YOUR_PINECONE_ASSISTANT_HOST>/mcp/assistants/<YOUR_ASSISTANT_NAME> --header "Authorization: Bearer <YOUR_PINECONE_API_KEY>"
   ```

   Replace `<YOUR_PINECONE_API_KEY>` with your Pinecone API key, `<YOUR_PINECONE_ASSISTANT_HOST>` with your Pinecone Assistant host, and `<YOUR_ASSISTANT_NAME>` with your assistant's name.

2. Verify the server was added successfully:

   ```bash  theme={null}
   claude mcp get my-assistant
   ```

3. The MCP server tools should now be available in Claude Code's chat interface.

### Use with Claude Desktop

You can configure Claude Desktop to use your assistant's remote MCP server. However, at this early stage of **remote** MCP server adoption, the Claude Desktop application does not support remote server URLs. In the example below, we work around this by using a local proxy server, [supergateway](https://github.com/supercorp-ai/supergateway), to forward requests to the remote MCP server with your API key.

<Warning>
  [supergateway](https://github.com/supercorp-ai/supergateway) is an open-source third-party tool. Use at your own risk.
</Warning>

1. Open [Claude Desktop](https://claude.ai/download) and go to **Settings**.

2. On the **Developer** tab, click **Edit Config** to open the configuration file.

3. Add the following configuration:

   ```json  theme={null}
   {
     "mcpServers": {
       "Assistant over supergateway": {
         "command": "npx",
         "args": [
           "-y",
           "supergateway",
           "--streamableHttp",
           "https://<YOUR_PINECONE_ASSISTANT_HOST>/mcp/assistants/<YOUR_ASSISTANT_NAME>",
           "--header",
           "Authorization: Bearer <YOUR_PINECONE_API_KEY>"
         ]
       }
     }
   }
   ```

   Replace `<YOUR_PINECONE_API_KEY>` with your Pinecone API key and `<YOUR_PINECONE_ASSISTANT_HOST>` with your Pinecone Assistant host.

4. Save the configuration file and restart Claude Desktop.

5. From the new chat screen, you should see a hammer (MCP) icon appear with the new MCP server available.

### Use with Cursor

You can configure Cursor to use your assistant's remote MCP server directly through the `.cursor/mcp.json` configuration file.

1. Open [Cursor](https://www.cursor.com/) and create a `.cursor` directory in your project root if it doesn't exist.

2. Open `.cursor/mcp.json` (create it if necessary). To learn more, refer to [Cursor's MCP documentation](https://docs.cursor.com/context/mcp).

3. Add the following configuration:

   ```json  theme={null}
   {
     "mcpServers": {
       "pinecone-assistant": {
         "url": "https://<YOUR_PINECONE_ASSISTANT_HOST>/mcp/assistants/<YOUR_ASSISTANT_NAME>",
         "headers": {
           "Authorization": "Bearer <YOUR_PINECONE_API_KEY>"
         }
       }
     }
   }
   ```

   Replace `<YOUR_PINECONE_API_KEY>` with your Pinecone API key, `<YOUR_PINECONE_ASSISTANT_HOST>` with your Pinecone Assistant host, and `<YOUR_ASSISTANT_NAME>` with your assistant's name.

4. Save the configuration file.

5. The MCP server tools should now be available in Cursor's chat interface.

### Use with LangChain

You can use the [LangChain MCP client](https://github.com/langchain-ai/langchain-mcp-adapters) to integrate with LangChain to create a powerful multi-agent workflow.

For example, the following code integrates Langchain with two assistants, one called `ai-news` and the other called `industry-reports`:

```python Python theme={null}
# Example code for integrating with LangChain
from langchain_mcp_adapters.client import MultiServerMCPClient
from langgraph.prebuilt import create_react_agent

from langchain_anthropic import ChatAnthropic

model = ChatAnthropic(model_name="claude-3-7-sonnet-latest", api_key=<YOUR_ANTHROPIC_API_KEY_HERE>)
pinecone_api_key = "<YOUR_PINECONE_API_KEY>"

async with MultiServerMCPClient(
    {
        "assistant_ai_news": {
            "url": "https://prod-1-data.ke.pinecone.io/mcp/assistants/ai-news",
            "transport": "streamable_http",
            "headers": {
                "Authorization": f"Bearer {pinecone_api_key}"
            }
        },
        "assistant_industry_reports": {
            "url": "https://prod-1-data.ke.pinecone.io/mcp/assistants/industry-reports",
            "transport": "streamable_http",
            "headers": {
                "Authorization": f"Bearer {pinecone_api_key}"
            }
        }
    }
) as client:
    agent = create_react_agent(model, client.get_tools())

    response = await agent.ainvoke({"messages": "Your task is research the next trends in AI, and form a report with the most undervalued companies in the space. You have access to two assistants, one that can help you find the latest trends in AI, and one that can help you find reports on companies."})
    print(response["messages"][-1].content)
```

## Local MCP server

Pinecone provides an open-source Pinecone Assistant MCP server that you can run locally with Docker. This option is useful for development, testing, or when you want to run the MCP server within your own infrastructure or expand the MCP server to include additional capabilities.

For the most up-to-date information on the local MCP server, see the [Pinecone Assistant MCP server repository](https://github.com/pinecone-io/assistant-mcp).

### Prerequisites

* Docker is installed and running on your system.
* A Pinecone API key. You can create a new key in the [Pinecone console](https://app.pinecone.io/organizations/-/keys).
* Your Pinecone Assistant host. To find it, go to your assistant in the [Pinecone console](https://app.pinecone.io/organizations/-/assistants). You'll see the assistant **Host** in the sidebar.

### Start the MCP server

Download the `assistant-mcp` Docker image:

```bash  theme={null}
docker pull ghcr.io/pinecone-io/assistant-mcp
```

Start the MCP server, providing your Pinecone API key and Pinecone Assistant host:

```bash  theme={null}
docker run -i --rm \
  -e PINECONE_API_KEY=<PINECONE_API_KEY> \
  -e PINECONE_ASSISTANT_HOST=<PINECONE_ASSISTANT_HOST> \
  pinecone/assistant-mcp
```

### Use with Claude Desktop

1. Open [Claude Desktop](https://claude.ai/download) and go to **Settings**.

2. On the **Developer** tab, click **Edit Config** to open the configuration file.

3. Add the following configuration:

   ```json  theme={null}
   {
     "mcpServers": {
       "pinecone-assistant": {
         "command": "docker",
         "args": [
           "run", 
           "-i", 
           "--rm", 
           "-e", 
           "PINECONE_API_KEY", 
           "-e", 
           "PINECONE_ASSISTANT_HOST", 
           "pinecone/assistant-mcp"
         ],
         "env": {
           "PINECONE_API_KEY": "<YOUR_PINECONE_API_KEY>",
           "PINECONE_ASSISTANT_HOST": "<YOUR_PINECONE_ASSISTANT_HOST>"
         }
       }
     }
   }
   ```

   Replace `<YOUR_PINECONE_API_KEY>` with your Pinecone API key and `<YOUR_PINECONE_ASSISTANT_HOST>` with your Pinecone Assistant host.

4. Save the configuration file and restart Claude Desktop.

5. From the new chat screen, you should see a hammer (MCP) icon appear with the new MCP server available.

### Use with Cursor

1. Open [Cursor](https://www.cursor.com/) and create a `.cursor` directory in your project root if it doesn't exist.

2. Open `.cursor/mcp.json` (create it if necessary). To learn more, refer to [Cursor's MCP documentation](https://docs.cursor.com/context/mcp).

3. Add the following configuration:

   ```json  theme={null}
   {
     "mcpServers": {
       "pinecone-assistant": {
         "command": "docker",
         "args": [
           "run", 
           "-i", 
           "--rm", 
           "-e", 
           "PINECONE_API_KEY", 
           "-e", 
           "PINECONE_ASSISTANT_HOST", 
           "pinecone/assistant-mcp"
         ],
         "env": {
           "PINECONE_API_KEY": "<YOUR_PINECONE_API_KEY>",
           "PINECONE_ASSISTANT_HOST": "<YOUR_PINECONE_ASSISTANT_HOST>"
         }
       }
     }
   }
   ```

   Replace `<YOUR_PINECONE_API_KEY>` with your Pinecone API key and `<YOUR_PINECONE_ASSISTANT_HOST>` with your Pinecone Assistant host.

4. Save the configuration file.

## Next Steps

* Visit the [Pinecone Assistant MCP Server repository](https://github.com/pinecone-io/assistant-mcp) for detailed installation and usage instructions

* Learn about [Model Context Protocol](https://modelcontextprotocol.io/) and how it enables AI agents to interact with tools and data

* Explore [retrieve context snippets](/guides/assistant/retrieve-context-snippets) to understand the underlying API functionality

---

# Source: https://docs.pinecone.io/reference/api/2025-10/assistant/metrics_alignment.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Evaluate an answer

> Evaluate the correctness and completeness of a response from an assistant or a RAG system. The correctness and completeness are evaluated based on the precision and recall of the generated answer with respect to the ground truth answer facts. Alignment is the harmonic mean of correctness and completeness.

For guidance and examples, see [Evaluate answers](https://docs.pinecone.io/guides/assistant/evaluate-answers).

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl https://prod-1-data.ke.pinecone.io/assistant/evaluation/metrics/alignment \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "question": "What are the capital cities of France, England and Spain?",
      "answer": "Paris is the capital city of France and Barcelona of Spain",
      "ground_truth_answer": "Paris is the capital city of France, London of England and Madrid of Spain"
  }'
  ```
</RequestExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/assistant_evaluation_2025-10.oas.yaml post /evaluation/metrics/alignment
openapi: 3.0.3
info:
  title: Evaluation API
  description: Provides endpoints for evaluating RAG systems using various metrics.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://prod-1-data.ke.pinecone.io/assistant
    description: Evaluation US Production API endpoints
  - url: https://prod-eu-data.ke.pinecone.io/assistant
    description: Evaluation EU Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Metrics
    description: Operations related to evaluation metrics calculation
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /evaluation/metrics/alignment:
    post:
      tags:
        - Metrics
      summary: Evaluate an answer
      description: >-
        Evaluate the correctness and completeness of a response from an
        assistant or a RAG system. The correctness and completeness are
        evaluated based on the precision and recall of the generated answer with
        respect to the ground truth answer facts. Alignment is the harmonic mean
        of correctness and completeness.


        For guidance and examples, see [Evaluate
        answers](https://docs.pinecone.io/guides/assistant/evaluate-answers).
      operationId: metrics_alignment
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        description: The request body for the alignment evaluation.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/AlignmentRequest'
        required: true
      responses:
        '200':
          description: The evaluation metrics and reasoning for the generated answer.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/AlignmentResponse'
        '422':
          description: Validation error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/BasicErrorResponse'
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/BasicErrorResponse'
components:
  schemas:
    AlignmentRequest:
      title: AlignmentRequest
      description: The request for the alignment evaluation.
      type: object
      properties:
        question:
          example: What is the capital city of Spain?
          title: Question
          description: The question for which the answer was generated.
          type: string
        answer:
          example: Barcelona.
          title: Answer
          description: The generated answer.
          type: string
        ground_truth_answer:
          example: Madrid.
          title: Ground Truth Answer
          description: The ground truth answer to the question.
          type: string
      required:
        - question
        - answer
        - ground_truth_answer
      additionalProperties: false
    AlignmentResponse:
      title: AlignmentResponse
      description: The response for the alignment evaluation.
      type: object
      properties:
        metrics:
          $ref: '#/components/schemas/Metrics'
        reasoning:
          $ref: '#/components/schemas/Reasoning'
        usage:
          $ref: '#/components/schemas/TokenCounts'
      required:
        - metrics
        - reasoning
        - usage
      additionalProperties: false
    BasicErrorResponse:
      title: BasicErrorResponse
      description: A basic error response that contains a message.
      type: object
      properties:
        message:
          title: Message
          type: string
      required:
        - message
    Metrics:
      title: Metrics
      description: The metrics returned for the alignment evaluation.
      type: object
      properties:
        correctness:
          title: Correctness
          description: The precision of the generated answer.
          type: number
        completeness:
          title: Completeness
          description: The recall of the generated answer.
          type: number
        alignment:
          title: Alignment
          description: The harmonic mean of correctness and completeness.
          type: number
      required:
        - correctness
        - completeness
        - alignment
      additionalProperties: false
    Reasoning:
      title: Reasoning
      description: The reasoning behind the alignment evaluation.
      type: object
      properties:
        evaluated_facts:
          title: Evaluated Facts
          description: The facts that were evaluated.
          type: array
          items:
            $ref: '#/components/schemas/EvaluatedFact'
      required:
        - evaluated_facts
      additionalProperties: false
    TokenCounts:
      title: TokenCounts
      description: Token counts for the input prompt and completion.
      type: object
      properties:
        prompt_tokens:
          title: Prompt Tokens
          type: integer
        completion_tokens:
          title: Completion Tokens
          type: integer
        total_tokens:
          title: Total Tokens
          type: integer
      required:
        - prompt_tokens
        - completion_tokens
        - total_tokens
    EvaluatedFact:
      title: EvaluatedFact
      description: A fact that was evaluated.
      type: object
      properties:
        fact:
          $ref: '#/components/schemas/Fact'
        entailment:
          $ref: '#/components/schemas/Entailment'
      required:
        - fact
        - entailment
      additionalProperties: false
    Fact:
      title: Fact
      description: A fact
      type: object
      properties:
        content:
          title: Content
          description: The content of the fact.
          type: string
      required:
        - content
      additionalProperties: false
    Entailment:
      title: Entailment
      description: The entailment of a fact.
      x-enum:
        - entailed
        - contradicted
        - neutral
      type: string
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/integrations/microsoft-marketplace.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Microsoft Marketplace

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Access Pinecone through our Microsoft Marketplace listing. Microsoft Marketplace allows you to manage Pinecone and other third-party software from a centralized location, and simplifies software licensing and procurement with flexible pricing options and multiple deployment methods.

You can set up pay-as-you-go billing for a Pinecone organization through the Microsoft Marketplace

<PrimarySecondaryCTA primaryLabel="Get started" primaryHref="https://marketplace.microsoft.com/product/saas/pineconesystemsinc1688761585469.pineconesaas" />

---

# Source: https://docs.pinecone.io/guides/indexes/pods/migrate-a-pod-based-index-to-serverless.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Migrate a pod-based index to serverless

> Migrate existing pod indexes to cost-effective serverless

This page shows you how to migrate a pod-based index to [serverless](/guides/get-started/database-architecture). The migration process is free; the standard costs of upserting records to a new serverless index are not applied.

<Warning>
  In most cases, migrating to serverless reduces costs significantly. For read-heavy workloads with more than 1 query per second and for indexes with many records in a single namespace, consider building your serverless indexes on [dedicated read nodes](/guides/index-data/dedicated-read-nodes).

  Before migrating, [contact Pinecone Support](/troubleshooting/contact-support) for help estimating and managing cost implications.
</Warning>

## Limitations

Migration is supported for pod-based indexes with less than 25 million records and 20,000 namespaces across all supported clouds (AWS, GCP, and Azure).

Also, serverless indexes do not support the following features. If you were using these features for your pod-based index, you will need to adapt your code. If you are blocked by these limitations, [contact Pinecone Support](/troubleshooting/contact-support).

* [Selective metadata indexing](/guides/indexes/pods/manage-pod-based-indexes#selective-metadata-indexing)

  * Because high-cardinality metadata in serverless indexes does not cause high memory utilization, this operation is not relevant.
* [Filtering index statistics by metadata](/reference/api/latest/data-plane/describeindexstats)

## How it works

Migrating a pod-based index to serverless is a 2-step process:

<Steps>
  <Step title="Save the pod-based index as a collection" />

  <Step title="Create a new serverless index from the collection" />
</Steps>

After migration, you will have both a new serverless index and the original pod-based index. Once you've switched your workload to the serverless index, you can delete the pod-based index to avoid paying for unused resources.

## 1. Understand cost implications

In most cases, migrating to serverless reduces costs significantly. However, costs can increase for read-heavy workloads with more than 1 query per second and for indexes with many records in a single namespace.

Before migrating, consider [contacting Pinecone Support](/troubleshooting/contact-support) for help estimating and managing cost implications.

## 2. Prepare for migration

Migrating a pod-based index to serverless can take anywhere from a few minutes to several hours, depending on the size of the index. During that time, you can continue reading from the pod-based index. However, all [upserts](/guides/index-data/upsert-data), [updates](/guides/manage-data/update-data), and [deletes](/guides/manage-data/delete-data) to the pod-based index will not automatically be reflected in the new serverless index, so be sure to prepare in one of the following ways:

* **Pause write traffic:** If downtime is acceptable, pause traffic to the pod-based index before starting migration. After migration, you will start sending traffic to the serverless index.

* **Log your writes:** If you need to continue reading from the pod-based index during migration, send read traffic to the pod-based index, but log your writes to a temporary location outside of Pinecone (e.g., S3). After migration, you will replay the logged writes to the new serverless index and start sending all traffic to the serverless index.

## 3. Start migration

<Tabs>
  <Tab title="Pinecone console">
    1. In the [Pinecone console](https://app.pinecone.io/), go to your pod-based index and click the **ellipsis (...) menu > Migrate to serverless**.

       <img className="block max-w-full" noZoom src="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/migrate-to-serverless.png?fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=24217a2e791ffce162686b8fdef47948" data-og-width="1184" width="1184" data-og-height="252" height="252" data-path="images/migrate-to-serverless.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/migrate-to-serverless.png?w=280&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=6ccfc31bcf3e4348c53ed95c0ce4965c 280w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/migrate-to-serverless.png?w=560&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=0cc7ead9e2f5f6552ed7ef29bf0140fb 560w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/migrate-to-serverless.png?w=840&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=780e1acbe7dbdf582d6b46bd8b2a8e4b 840w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/migrate-to-serverless.png?w=1100&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=89bccffcedc7cedecb621f3c453f8938 1100w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/migrate-to-serverless.png?w=1650&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=f732644be14d0a1ff10ef79a5e48ada1 1650w, https://mintcdn.com/pinecone/r0TaYXrfSrAYZYUj/images/migrate-to-serverless.png?w=2500&fit=max&auto=format&n=r0TaYXrfSrAYZYUj&q=85&s=8a77b4f2dd850dd607c84a3ec29a4fff 2500w" />

       <Note>
         The dropdown will not display **Migrate to serverless** if the index has any of the listed [limitations](#limitations).
       </Note>

    2. To save the legacy index and create a new serverless index now, follow the prompts.

       Depending on the size of the index, migration can take anywhere from a few minutes to several hours. While migration is in progress, you'll see the yellow **Initializing** status:
       <img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-initializing.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=acffe738f1cbe2e08e4bc2fad7e957dd" alt="create index from collection - initializing status" data-og-width="2202" width="2202" data-og-height="506" height="506" data-path="images/create-serverless-from-collection-initializing.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-initializing.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6f4464d50c752857a48f220414c61372 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-initializing.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=fa65345532ef65a5f57007e2e233b30d 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-initializing.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=9720ae95b7630f338f021e642c35942d 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-initializing.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=75802502e5b8b22ab0c49728f26c37b7 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-initializing.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=362c1d22e182242489d383254856f612 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-initializing.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=0d439490afc512a8d8e1ddaffbb22544 2500w" />

       When the new serverless index is ready, the status will change to green:

       <img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-ready.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=e04b6b0d892301435be282e5c4fa394b" alt="create index from collection - ready status" data-og-width="2202" width="2202" data-og-height="500" height="500" data-path="images/create-serverless-from-collection-ready.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-ready.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=e9f3525138da438aca544f9f9e91ec97 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-ready.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=f3b774060c9b08569720013a324dd9d0 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-ready.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=195db840282f07a709bff77931856714 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-ready.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=f40328d14f9aa1dac2aacf82115ac53b 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-ready.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=8a50755a42dd444ce4457673473507bc 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/create-serverless-from-collection-ready.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=490d2bf71be87a7560493ea47b586bcb 2500w" />
  </Tab>

  <Tab title="API/SDK">
    1. Use the [`create_collection`](/reference/api/latest/control-plane/create_collection) operation to create a backup of your pod-based index:

       <CodeGroup>
         ```javascript JavaScript theme={null}
         // Requires Node.js SDK v6.1.2 or later
         import { Pinecone } from '@pinecone-database/pinecone'

         const pc = new Pinecone({
           apiKey: 'YOUR_API_KEY'
         });

         await pc.createCollection({
           name: "pod-collection",
           source: "pod-index"
         });
         ```

         ```go Go theme={null}
         // Requires Go SDK v4.1.2 or later
         package main

         import (
             "context"
             "fmt"
             "log"

             "github.com/pinecone-io/go-pinecone/v4/pinecone"
         )

         func main() {
             ctx := context.Background()

             pc, err := pinecone.NewClient(pinecone.NewClientParams{
                 ApiKey: "YOUR_API_KEY",
             })
             if err != nil {
                 log.Fatalf("Failed to create Client: %v", err)
             }

             collection, err := pc.CreateCollection(ctx, &pinecone.CreateCollectionRequest{
                 Name: "pod-collection", 
                 Source: "pod-index",
             })
             if err != nil {
                 log.Fatalf("Failed to create collection: %v", err)
             } else {
                 fmt.Printf("Successfully created collection: %v", collection.Name)
             }
         }
         ```

         ```shell curl theme={null}
         PINECONE_API_KEY="YOUR_API_KEY"

         curl -s POST "https://api.pinecone.io/collections" \
         -H "Accept: application/json" \
         -H "Content-Type: application/json" \
         -H "Api-Key: $PINECONE_API_KEY" \
         -H "X-Pinecone-Api-Version: 2025-10" \
         -d '{
                 "name": "pod-collection",
                 "source": "pod-index"
         }'
         ```
       </CodeGroup>

    2. Use the [`create_index`](/reference/api/latest/control-plane/create_index) operation to create a new serverless index from the collection:

       * Use API verison `2025-04` or later. Creating a serverless index from a collection is not supported in earlier versions.
       * Set `dimension` to the same dimension as the pod-based index. Changing the dimension is not supported.
       * Set `cloud` to the cloud where the pod-based index is hosted. Migrating to a different cloud is not supported.
       * Set `source_collection` to the name of the collection you created in step 1.

       <CodeGroup>
         ```javascript JavaScript theme={null}
         import { Pinecone } from '@pinecone-database/pinecone'

         const pc = new Pinecone({
           apiKey: 'YOUR_API_KEY'
         });

         await pc.createIndex({
           name: 'serverless-index',
           vectorType: 'dense',
           dimension: 1536,
           metric: 'cosine',
           spec: {
             serverless: {
               cloud: 'aws',
               region: 'us-east-1',
               sourceCollection: 'pod-collection'
             }
           }
         });
         ```

         ```go Go theme={null}
         package main

         import (
             "context"
             "fmt"
             "log"

             "github.com/pinecone-io/go-pinecone/v4/pinecone"
         )

         func main() {
             ctx := context.Background()

             pc, err := pinecone.NewClient(pinecone.NewClientParams{
                 ApiKey: "YOUR_API_KEY",
             })
             if err != nil {
                 log.Fatalf("Failed to create Client: %v", err)
             }

             idx, err := pc.CreateServerlessIndex(ctx, &pinecone.CreateServerlessIndexRequest{
                 Name:      "serverless-index",
                 VectorType: "dense",
                 Dimension: 1536,
                 Metric:    pinecone.Cosine,
                 Cloud:     pinecone.Aws,
                 Region:    "us-east-1",
                 SourceCollection: "pod-collection",
             })
             if err != nil {
                 log.Fatalf("Failed to create serverless index: %v", err)
             } else {
                 fmt.Printf("Successfully created serverless index: %v", idx.Name)
             }
         }
         ```

         ```shell curl theme={null}
         PINECONE_API_KEY="YOUR_API_KEY"

         curl -s "https://api.pinecone.io/indexes" \
         -H "Accept: application/json" \
         -H "Content-Type: application/json" \
         -H "Api-Key: $PINECONE_API_KEY" \
         -H "X-Pinecone-Api-Version: 2025-10" \
         -d '{
                 "name": "serverless-index",
                 "vector_type": "dense",
                 "dimension": 1536,
                 "metric": "cosine",
                 "spec": {
                     "serverless": {
                         "cloud": "aws",
                         "region": "us-east-1",
                         "source_collection": "pod-collection"
                     }
                 }
             }'
         ```
       </CodeGroup>
  </Tab>
</Tabs>

## 4. Update SDKs

If you are using an older version of the Python, Node.js, Java, or Go SDK, you must update the SDK to work with serverless indexes.

1. Check your SDK version:

   <CodeGroup>
     ```shell Python theme={null}
     pip show pinecone  
     ```

     ```shell JavaScript theme={null}
     npm list | grep @pinecone-database/pinecone  
     ```

     ```shell Java  theme={null}
     # Check your dependency file or classpath
     ```

     ```shell Go theme={null}
     go list -u -m all | grep go-pinecone
     ```
   </CodeGroup>

2. If your SDK version is less than 3.0.0 for [Python](https://github.com/pinecone-io/pinecone-python-client), 2.0.0 for [Node.js](https://sdk.pinecone.io/typescript/), 1.0.0 for [Java](https://github.com/pinecone-io/pinecone-java-client), or 1.0.0 for [Go](https://github.com/pinecone-io/go-pinecone), upgrade the SDK as follows:

   <CodeGroup>
     ```Python Python theme={null}
     pip install "pinecone[grpc]" --upgrade  
     ```

     ```JavaScript JavaScript theme={null}
     npm install @pinecone-database/pinecone@latest  
     ```

     ```shell Java theme={null}
     # Maven
     <dependency>
       <groupId>io.pinecone</groupId>
       <artifactId>pinecone-client</artifactId>
       <version>5.0.0</version>
     </dependency>

     # Gradle
     implementation "io.pinecone:pinecone-client:5.0.0"
     ```

     ```go Go theme={null}
     go get -u github.com/pinecone-io/go-pinecone/v4/pinecone@latest
     ```
   </CodeGroup>

   If you are using the [.NET SDK](/reference/sdks/dotnet/overview), add a package reference to your project file:

   ```shell C# theme={null}
   dotnet add package Pinecone.Client 
   ```

## 5. Adapt existing code

You must make some minor code changes to work with serverless indexes.

<Warning>
  Serverless indexes do not support some features, as outlined in [Limitations](#limitations). If you were relying on these features for your pod-based index, you’ll need to adapt your code.
</Warning>

1. Change how you import the Pinecone library and authenticate and initialize the client:

   <CodeGroup>
     ```Python Python theme={null}
     from pinecone.grpc import PineconeGRPC as Pinecone
     from pinecone import ServerlessSpec, PodSpec  
     # ServerlessSpec and PodSpec are required only when  
     # creating serverless and pod-based indexes.  
     pc = Pinecone(api_key="YOUR_API_KEY")  
     ```

     ```JavaScript JavaScript theme={null}
     import { Pinecone } from '@pinecone-database/pinecone';  

     const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
     ```

     ```java Java theme={null}
     import io.pinecone.clients.Pinecone;
     import org.openapitools.db_control.client.model.*;

     public class InitializeClientExample {
         public static void main(String[] args) {
             Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
         }
     }
     ```

     ```go Go theme={null}
     package main

     import (
         "context"
         "log"

         "github.com/pinecone-io/go-pinecone/v4/pinecone"
     )

     func main() {
         ctx := context.Background()

         pc, err := pinecone.NewClient(pinecone.NewClientParams{
             ApiKey: "YOUR_API_KEY",
         })
         if err != nil {
             log.Fatalf("Failed to create Client: %v", err)
         }
     }
     ```

     ```csharp C# theme={null}
     using Pinecone;

     var pinecone = new PineconeClient("YOUR_API_KEY");
     ```
   </CodeGroup>

2. [Listing indexes](/guides/manage-data/manage-indexes) now fetches a complete description of each index. If you were relying on the output of this operation, you'll need to adapt your code.

   <CodeGroup>
     ```Python Python theme={null}
     from pinecone.grpc import PineconeGRPC as Pinecone

     pc = Pinecone(api_key="YOUR_API_KEY")

     index_list = pc.list_indexes()

     print(index_list)
     ```

     ```javascript JavaScript theme={null}
     import { Pinecone } from '@pinecone-database/pinecone'

     const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

     const indexList = await pc.listIndexes();

     console.log(indexList);
     ```

     ```java Java theme={null}
     import io.pinecone.clients.Pinecone;
     import org.openapitools.db_control.client.model.*;

     public class ListIndexesExample {
         public static void main(String[] args) {
             Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
             IndexList indexList = pc.listIndexes();
             System.out.println(indexList);
         }
     }
     ```

     ```go Go theme={null}
     package main

     import (
         "context"
         "encoding/json"
         "fmt"
         "log"

         "github.com/pinecone-io/go-pinecone/v4/pinecone"
     )

     func prettifyStruct(obj interface{}) string {
         bytes, _ := json.MarshalIndent(obj, "", "  ")
         return string(bytes)
     }

     func main() {
         ctx := context.Background()

         pc, err := pinecone.NewClient(pinecone.NewClientParams{
             ApiKey: "YOUR_API_KEY",
         })
         if err != nil {
             log.Fatalf("Failed to create Client: %v", err)
         }

         idxs, err := pc.ListIndexes(ctx)
         if err != nil {
             log.Fatalf("Failed to list indexes: %v", err)
         } else {
             for _, index := range idxs {
                 fmt.Printf("index: %v\n", prettifyStruct(index))
             }
         }
     }
     ```

     ```csharp C# theme={null}
     using Pinecone;

     var pinecone = new PineconeClient("YOUR_API_KEY");

     var indexList = await pinecone.ListIndexesAsync();

     Console.WriteLine(indexList);
     ```

     ```shell curl theme={null}
     PINECONE_API_KEY="YOUR_API_KEY"

     curl -i -X GET "https://api.pinecone.io/indexes" \
     -H "Api-Key: $PINECONE_API_KEY" \
     -H "X-Pinecone-Api-Version: 2025-10"
     ```
   </CodeGroup>

   The `list_indexes` operation now returns a response like the following:

   <CodeGroup>
     ```python Python theme={null}
     [{
         "name": "docs-example-sparse",
         "metric": "dotproduct",
         "host": "docs-example-sparse-govk0nt.svc.aped-4627-b74a.pinecone.io",
         "spec": {
             "serverless": {
                 "cloud": "aws",
                 "region": "us-east-1"
             }
         },
         "status": {
             "ready": true,
             "state": "Ready"
         },
         "vector_type": "sparse",
         "dimension": null,
         "deletion_protection": "disabled",
         "tags": {
             "environment": "development"
         }
     }, {
         "name": "docs-example-dense",
         "metric": "cosine",
         "host": "docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io",
         "spec": {
             "serverless": {
                 "cloud": "aws",
                 "region": "us-east-1"
             }
         },
         "status": {
             "ready": true,
             "state": "Ready"
         },
         "vector_type": "dense",
         "dimension": 1536,
         "deletion_protection": "disabled",
         "tags": {
             "environment": "development"
         }
     }]
     ```

     ```javascript JavaScript theme={null}
     {
       indexes: [
         {
           name: 'docs-example-sparse',
           dimension: undefined,
           metric: 'dotproduct',
           host: 'docs-example-sparse-govk0nt.svc.aped-4627-b74a.pinecone.io',
           deletionProtection: 'disabled',
           tags: { environment: 'development', example: 'tag' },
           embed: undefined,
           spec: { pod: undefined, serverless: { cloud: 'aws', region: 'us-east-1' } },
           status: { ready: true, state: 'Ready' },
           vectorType: 'sparse'
         },
         {
           name: 'docs-example-dense',
           dimension: 1536,
           metric: 'cosine',
           host: 'docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io',
           deletionProtection: 'disabled',
           tags: { environment: 'development', example: 'tag' },
           embed: undefined,
           spec: { pod: undefined, serverless: { cloud: 'aws', region: 'us-east-1' } },
           status: { ready: true, state: 'Ready' },
           vectorType: 'dense'
         }
       ]
     }
     ```

     ```java Java theme={null}
     class IndexList {
         indexes: [class IndexModel {
             name: docs-example-sparse
             dimension: null
             metric: dotproduct
             host: docs-example-sparse-govk0nt.svc.aped-4627-b74a.pinecone.io
             deletionProtection: disabled
             tags: {environment=development}
             embed: null
             spec: class IndexModelSpec {
                 pod: null
                 serverless: class ServerlessSpec {
                     cloud: aws
                     region: us-east-1
                     additionalProperties: null
                 }
                 additionalProperties: null
             }
             status: class IndexModelStatus {
                 ready: true
                 state: Ready
                 additionalProperties: null
             }
             vectorType: sparse
             additionalProperties: null
         }, class IndexModel {
             name: docs-example-dense
             dimension: 1536
             metric: cosine
             host: docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io
             deletionProtection: disabled
             tags: {environment=development}
             embed: null
             spec: class IndexModelSpec {
                 pod: null
                 serverless: class ServerlessSpec {
                     cloud: aws
                     region: us-east-1
                     additionalProperties: null
                 }
                 additionalProperties: null
             }
             status: class IndexModelStatus {
                 ready: true
                 state: Ready
                 additionalProperties: null
             }
             vectorType: dense
             additionalProperties: null
         }]
         additionalProperties: null
     }
     ```

     ```go Go theme={null}
     index: {
       "name": "docs-example-sparse",
       "host": "docs-example-sparse-govk0nt.svc.aped-4627-b74a.pinecone.io",
       "metric": "dotproduct",
       "vector_type": "sparse",
       "deletion_protection": "disabled",
       "dimension": null,
       "spec": {
         "serverless": {
           "cloud": "aws",
           "region": "us-east-1"
         }
       },
       "status": {
         "ready": true,
         "state": "Ready"
       },
       "tags": {
         "environment": "development"
       }
     }
     index: {
       "name": "docs-example-dense",
       "host": "docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io",
       "metric": "cosine",
       "vector_type": "dense",
       "deletion_protection": "disabled",
       "dimension": 1536,
       "spec": {
         "serverless": {
           "cloud": "aws",
           "region": "us-east-1"
         }
       },
       "status": {
         "ready": true,
         "state": "Ready"
       },
       "tags": {
         "environment": "development"
       }
     }
     ```

     ```csharp C# theme={null}
     {
       "indexes": [
         {
           "name": "docs-example-sparse",
           "metric": "dotproduct",
           "host": "docs-example-sparse-govk0nt.svc.aped-4627-b74a.pinecone.io",
           "deletion_protection": "disabled",
           "tags": {
             "environment": "development"
           },
           "spec": {
             "serverless": {
               "cloud": "aws",
               "region": "us-east-1"
             }
           },
           "status": {
             "ready": true,
             "state": "Ready"
           },
           "vector_type": "sparse"
         },
         {
           "name": "docs-example-dense",
           "dimension": 1536,
           "metric": "cosine",
           "host": "docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io",
           "deletion_protection": "disabled",
           "tags": {
             "environment": "development"
           },
           "spec": {
             "serverless": {
               "cloud": "aws",
               "region": "us-east-1"
             }
           },
           "status": {
             "ready": true,
             "state": "Ready"
           },
           "vector_type": "dense"
         }
       ]
     }
     ```

     ```json curl theme={null}
     {
       "indexes": [
         {
           "name": "docs-example-sparse",
           "vector_type": "sparse",
           "metric": "dotproduct",
           "dimension": null,
           "status": {
             "ready": true,
             "state": "Ready"
           },
           "host": "docs-example-sparse-govk0nt.svc.aped-4627-b74a.pinecone.io",
           "spec": {
             "serverless": {
               "region": "us-east-1",
               "cloud": "aws"
             }
           },
           "deletion_protection": "disabled",
           "tags": {
             "environment": "development"
           }
         },
         {
           "name": "docs-example-dense",
           "vector_type": "dense",
           "metric": "cosine",
           "dimension": 1536,
           "status": {
             "ready": true,
             "state": "Ready"
           },
           "host": "docs-example-dense-govk0nt.svc.aped-4627-b74a.pinecone.io",
           "spec": {
             "serverless": {
               "region": "us-east-1",
               "cloud": "aws"
             }
           },
           "deletion_protection": "disabled",
           "tags": {
             "environment": "development"
           }
         }
       ]
     }
     ```
   </CodeGroup>

3. [Describing an index](/guides/manage-data/manage-indexes) now returns a description of an index in a different format. It also returns the index host needed to run data plane operations against the index. If you were relying on the output of this operation, you'll need to adapt your code.

   <CodeGroup>
     ```Python Python theme={null}
     from pinecone.grpc import PineconeGRPC as Pinecone

     pc = Pinecone(api_key="YOUR_API_KEY")

     pc.describe_index(name="docs-example")
     ```

     ```JavaScript JavaScript theme={null}
     import { Pinecone } from '@pinecone-database/pinecone';

     const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

     await pc.describeIndex('docs-example');
     ```

     ```java Java theme={null}
     import io.pinecone.clients.Pinecone;
     import org.openapitools.db_control.client.model.*;

     public class DescribeIndexExample {
         public static void main(String[] args) {
             Pinecone pc = new Pinecone.Builder("YOURE_API_KEY").build();
             IndexModel indexModel = pc.describeIndex("docs-example");
             System.out.println(indexModel);
         }
     }
     ```

     ```go Go theme={null}
     package main

     import (
         "context"
         "encoding/json"
         "fmt"
         "log"

         "github.com/pinecone-io/go-pinecone/v4/pinecone"
     )

     func prettifyStruct(obj interface{}) string {
         bytes, _ := json.MarshalIndent(obj, "", "  ")
         return string(bytes)
     }

     func main() {
         ctx := context.Background()

         pc, err := pinecone.NewClient(pinecone.NewClientParams{
             ApiKey: "YOUR_API_KEY",
         })
         if err != nil {
             log.Fatalf("Failed to create Client: %v", err)
         }

         idx, err := pc.DescribeIndex(ctx, "docs-example")
         if err != nil {
             log.Fatalf("Failed to describe index \"%v\": %v", idx.Name, err)
         } else {
             fmt.Printf("index: %v\n", prettifyStruct(idx))
         }
     }
     ```

     ```csharp C# theme={null}
     using Pinecone;

     var pinecone = new PineconeClient("YOUR_API_KEY");

     var indexModel = await pinecone.DescribeIndexAsync("docs-example");

     Console.WriteLine(indexModel);
     ```

     ```bash curl theme={null}
     PINECONE_API_KEY="YOUR_API_KEY"

     curl -i -X GET "https://api.pinecone.io/indexes/docs-example" \
         -H "Api-Key: $PINECONE_API_KEY" \
         -H "X-Pinecone-Api-Version: 2025-10"
     ```
   </CodeGroup>

## 6. Use your new index

When you're ready to cutover to your new serverless index:

1. Your new serverless index has a different name and unique endpoint than your pod-based index. Update your code to target the new serverless index:

   <CodeGroup>
     ```Python Python theme={null}
     index = pc.Index("YOUR_SERVERLESS_INDEX_NAME")  
     ```

     ```JavaScript JavaScript theme={null}
     const index = pc.index("YOUR_SERVERLESS_INDEX_NAME");
     ```

     ```java Java theme={null}
     import io.pinecone.clients.Index;
     import io.pinecone.clients.Pinecone;

     public class TargetIndexExample {
         public static void main(String[] args) {
             Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
             Index index = pc.getIndexConnection("YOUR_SERVERLESS_INDEX_NAME");
     ```

     ```go Go theme={null}
     package main

     import (
         "context"
         "log"

         "github.com/pinecone-io/go-pinecone/v4/pinecone"
     )

     func main() {

         ctx := context.Background()

         pc, err := pinecone.NewClient(pinecone.NewClientParams{
             ApiKey: "YOUR_API_KEY",
         })
         if err != nil {
             log.Fatalf("Failed to create Client: %v", err)
         }

         idx, err := pc.DescribeIndex(ctx, "YOUR_SERVERLESS_INDEX_NAME")
         if err != nil {
             log.Fatalf("Failed to describe index \"%v\": %v", idx.Name, err)
         }

         idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: idx.Host, Namespace: "example-namespace"})
         if err != nil {
             log.Fatalf("Failed to create IndexConnection for Host %v: %v", idx.Host, err)
         }
     }
     ```

     ```csharp C# theme={null}
     using Pinecone;

     var pinecone = new PineconeClient("YOUR_API_KEY");

     var index = pinecone.Index("YOUR_SERVERLESS_INDEX_NAME");
     ```

     ```bash curl theme={null}
     # When using the API directly, you need the unique endpoint for your new serverless index. 
     # See https://docs.pinecone.io/guides/manage-data/target-an-index for details.
     PINECONE_API_KEY="YOUR_API_KEY"
     INDEX_HOST="INDEX_HOST"

     curl -X POST "https://$INDEX_HOST/describe_index_stats" \  
         -H "Api-Key: $PINECONE_API_KEY" \
         -H "X-Pinecone-Api-Version: 2025-10" 
     ```
   </CodeGroup>

2. Reinitialize your clients.

3. If you logged writes to the pod-based index during migration, replay the logged writes to your serverless index.

4. [Delete the pod-based index](/guides/manage-data/manage-indexes#delete-an-index) to avoid paying for unused resources.

   <Warning>
     It is not possible to save a serverless index as a collection, so if you want to retain the option to recreate your pod-based index, be sure to keep the collection you created earlier.
   </Warning>

## See also

* [Limits](/reference/api/database-limits)
* [Serverless architecture](/guides/get-started/database-architecture)
* [Understanding serverless cost](/guides/manage-cost/understanding-cost)

---

# Source: https://docs.pinecone.io/troubleshooting/minimize-latencies.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Minimize latencies

There are many aspects to consider to minimize latencies:

## Slow uploads or high latencies

To minimize latency when accessing Pinecone:

* Switch to a cloud environment. For example: EC2, GCE, [Google Colab](https://colab.research.google.com), [GCP AI Platform Notebook](https://cloud.google.com/ai-platform-notebooks), or [SageMaker Notebook](https://docs.aws.amazon.com/sagemaker/dg/nbi.html). If you experience slow uploads or high query latencies, it might be because you are accessing Pinecone from your home network.
* Consider deploying your application in the same environment as your Pinecone service.
* See [Decrease latency](/guides/optimize/decrease-latency) for more tips.

## High query latencies with batching

If you're batching queries, try reducing the number of queries per call to 1 query vector. You can make these [calls in parallel](/troubleshooting/parallel-queries) and expect roughly the same performance as with batching.

## High latencies with fetch or include\_values

For on-demand indexes, since vector values are retrieved from object storage, operations that return vector values (`fetch` operations or queries with `include_values=true`) may have increased latency. If you don't need the vector values, set `include_values=false` when querying, or use the [`query`](/reference/api/latest/data-plane/query) operation instead of `fetch` if you only need metadata or IDs. See [Decrease latency](/guides/optimize/decrease-latency#avoid-including-vector-values-when-not-needed) for more details.

---

# Source: https://docs.pinecone.io/troubleshooting/module-pinecone-has-no-attribute-init.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Python AttributeError: module pinecone has no attribute init

## Problem

If you are using Pinecone serverless and getting the error `"AttributeError: module 'pinecone' has no attribute 'init'`, first check that you are using the latest version of the Python SDK.

You can check the version of the client by running:

```shell  theme={null}
pip show pinecone
```

## Solution

Serverless requires a minimum version of 3.0. To upgrade to the latest version, run:

```shell  theme={null}
# If you're interacting with Pinecone via HTTP:
pip install pinecone --upgrade

# If you're using gRPC:
# pip install "pinecone[grpc]" --upgrade
```

If you're on the right version and getting this error, you just have to make some slight changes to your code to make use of serverless. Instead of calling:

```python  theme={null}
import pinecone

pinecone.init(api_key=api_key,environment=environment)
```

Use the following if you're interacting with Pinecone via HTTP requests:

```python  theme={null}
from pinecone import Pinecone

pc = Pinecone(api_key=api_key)
```

Or, use the following if you're using gRPC:

```python  theme={null}
from pinecone.grpc import PineconeGRPC as Pinecone

pc = Pinecone(api_key=api_key)
```

You no longer need to specify the cloud environment your index is hosted in; the API key is all you need.

---

# Source: https://docs.pinecone.io/guides/assistant/admin/monitor-spend-and-usage.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Monitor usage and cost

> Set monthly spend alerts and monitor usage across your organization.

## Set monthly spend alerts

You can set up email alerts to monitor your organization's monthly spending. These alerts notify designated recipients when spending reaches specified thresholds. The alerts automatically reset at the start of each monthly billing cycle.

To set a spend alert:

1. Go to [Settings > Spend alerts](https://app.pinecone.io/organizations/-/settings/spend-alerts) in the Pinecone console
2. Click **+ Add Alert**.
3. Enter the dollar amount for the spend alert.
4. Enter the email addresses to send the alert to. [Organization owners](/guides/organizations/understanding-organizations#organization-roles) are listed by default.
5. Click **Create**.

To edit a spend alert:

1. In the row of the spend alert you want to edit, click **ellipsis (...) menu > Edit**.
2. Change the dollar amount and/or email addresses for the spend alert.
3. Click **Update**.

<Note>
  **Auto-spend spike alert**: To protect from unexpected cost increases, Pinecone sends an alert when spending exceeds double your previous month's invoice amount. While the alert threshold is fixed and the alert cannot be deleted, you can modify which email addresses receive the alert and enable or disable the alert notifications.
</Note>

## Monitor organization-level usage

<Note>
  You must be the [organization owner](/guides/organizations/understanding-organizations#organization-owners) to view usage across your Pinecone organization. Also, this feature is available only to organizations on the Standard or Enterprise plans.
</Note>

To view and download a report of your usage and costs for your Pinecone organization, go to [**Settings > Usage**](https://app.pinecone.io/organizations/-/settings/usage) in the Pinecone console.

All dates are given in UTC to match billing invoices.

## Monitor token usage

Requests to the [chat](/reference/api/latest/assistant/chat_assistant), [context retrieval](/reference/api/latest/assistant/context_assistant), and [evaluation](/reference/api/latest/assistant/metrics_alignment) API endpoints return a `usage` parameter with `prompt_tokens`, `completion_tokens`, and `total_tokens` generated.

<Tabs>
  <Tab title="Chat">
    For [chat](/guides/assistant/chat-with-assistant), tokens are defined as follows:

    * `prompt_tokens` are based on the messages sent to the assistant and the context snippets retrieved from the assistant and sent to a model. Messages sent to the assistant can include messages from the [chat history](/guides/assistant/chat-with-assistant#provide-conversation-history) in addition to the newest message.

      `prompt_tokens` appear as **Assistants Input Tokens** on invoices.
    * `completion_tokens` are based on the answer from the model.

      `completion_tokens` appear as **Assistants Output Tokens** on invoices.
    * `total_tokens` is the sum of `prompt_tokens` and `completion_tokens`.

    ```json Example chat response {9-13} theme={null}
    {
        "finish_reason": "stop",
        "message": {
            "role": "assistant",
            "content": "The Chief Financial Officer (CFO) of Netflix is Spencer Neumann."
        },
        "id": "000000000000000030513193ccc52814",
        "model": "gpt-4o-2024-11-20",
        "usage": {
            "prompt_tokens": 23626,
            "completion_tokens": 21,
            "total_tokens": 23647
        },
        "citations": [
            {
                "position": 63,
                "references": [
                    {
                        "file": {
                            "status": "Available",
                            "id": "99305805-3844-41b5-af56-ee693ab80527",
                            "name": "Netflix-10-K-01262024.pdf",
                            "size": 1073470,
                            "metadata": null,
                            "updated_on": "2025-07-29T20:07:53.171752661Z",
                            "created_on": "2025-07-29T20:07:36.361322699Z",
                            "percent_done": 1,
                            "signed_url": "https://storage.googleapis.com/...",
                            "error_message": null
                        },
                        "pages": [
                            78,
                            79,
                            80
                        ],
                        "highlight": null
                    },
                    {
                        "file": {
                            "status": "Available",
                            "id": "99305805-3844-41b5-af56-ee693ab80527",
                            "name": "Netflix-10-K-01262024.pdf",
                            "size": 1073470,
                            "metadata": null,
                            "updated_on": "2025-07-29T20:07:53.171752661Z",
                            "created_on": "2025-07-29T20:07:36.361322699Z",
                            "percent_done": 1,
                            "signed_url": "https://storage.googleapis.com/...",
                            "error_message": null
                        },
                        "pages": [
                            77,
                            78
                        ],
                        "highlight": null
                    }
                ]
            }
        ]
    }
    ```
  </Tab>

  <Tab title="Context retrieval">
    For [context retrieval](/guides/assistant/context-snippets-overview), tokens are defined as follows:

    * `prompt_tokens` are based on the messages sent to the assistant and the context snippets retrieved from the assistant. Messages sent to the assistant can include messages from the chat history in addition to the newest message.

      `prompt_tokens` appear as **Assistants Context Tokens Processed** on invoices.

    * `completion_tokens` do not apply for context retrieval because, unlike for chat, there is no answer from a model. `completion_tokens` will always be 0.

    * `total_tokens` is the sum of `prompt_tokens` and `completion_tokens`.

    ```json Example context response {30-34} theme={null}
    {
        "snippets": [
            {
                "type": "text",
                "content": "edures, or caused such disclosure controls and procedures to be designed under our supervision, to\r\nensure that material information relating to the registrant, including its consolidated subsidiaries, ...",
                "score": 0.86632514,
                "reference": {
                    "type": "pdf",
                    "file": {
                        "status": "Available",
                        "id": "99305805-3844-41b5-af56-ee693ab80527",
                        "name": "Netflix-10-K-01262024.pdf",
                        "size": 1073470,
                        "metadata": null,
                        "updated_on": "2025-07-29T20:07:53.171752661Z",
                        "created_on": "2025-07-29T20:07:36.361322699Z",
                        "percent_done": 1,
                        "signed_url": "https://storage.googleapis.com/...",
                        "error_message": null
                    },
                    "pages": [
                        78,
                        79,
                        80
                    ]
                }
            },
            ...
        ],
        "usage": {
            "prompt_tokens": 22914,
            "completion_tokens": 0,
            "total_tokens": 22914
        },
        "id": "00000000000000007b6ad859184a31b3"
    }
    ```
  </Tab>

  <Tab title="Response evaluation">
    For [response evaluation](/guides/assistant/evaluation-overview), tokens are defined as follows:

    * `prompt_tokens` are based on two requests to a model: The first request contains a question, answer, and ground truth answer, and the second request contains the same details plus generated facts returned by the model for the first request.

      `prompt_tokens` appear as **Assistants Evaluation Tokens Processed** on invoices.
    * `completion_tokens` are based on two responses from a model: The first response contains generated facts, and the second response contains evaluation metrics.

      `completion_tokens` appear as **Assistants Evaluation Tokens Out** on invoices.
    * `total_tokens` is the sum of `prompt_tokens` and `completion_tokens`.

    ```json Response evaluation response {17-21} theme={null}
    {
      "metrics": {
        "correctness": 123,
        "completeness": 123,
        "alignment": 123
      },
      "reasoning": {
        "evaluated_facts": [
          {
            "fact": {
              "content": "<string>"
            },
            "entailment": "entailed"
          }
        ]
      },
      "usage": {
        "prompt_tokens": 123,
        "completion_tokens": 123,
        "total_tokens": 123
      }
    }
    ```
  </Tab>
</Tabs>

---

# Source: https://docs.pinecone.io/guides/manage-cost/monitor-usage-and-costs.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Monitor usage and costs

> Monitor usage and costs for your Pinecone organization and indexes.

## Monitor organization-level usage and costs

<Note>
  To view usage and costs across your Pinecone organization, you must be an [organization owner](/guides/organizations/understanding-organizations#organization-owners). Also, this feature is available only to organizations on the Standard or Enterprise plans.
</Note>

The **Usage** dashboard in the Pinecone console gives you a detailed report of usage and costs across your organization, broken down by each billable SKU or aggregated by project or service. You can view the report in the console or download it as a CSV file for more detailed analysis.

1. Go to [**Settings > Usage**](https://app.pinecone.io/organizations/-/settings/usage) in the Pinecone console.
2. Select the time range to report on. This defaults to the last 30 days.
3. Select the scope for your report:
   * **SKU:** The usage and cost for each billable SKU, for example, read units per cloud region, storage size per cloud region, or tokens per embedding model.
   * **Project:** The aggregated cost for each project in your organization.
   * **Service:** The aggregated cost for each service your organization uses, for example, database (includes serverless back up and restore), assistants, inference (embedding and reranking), and collections.
4. Choose the specific SKUs, projects, or services you want to report on. This defaults to all.
5. To download the report as a CSV file, click **Download**.

   <Tip>
     The CSV download provides more granular detail than the console view, including breakdowns by individual index as well as project and index tags.
   </Tip>

Dates are shown in UTC to match billing invoices. Cost data is delayed up to three days from the actual usage date.

## Monitor index-level usage

You can monitor index-level usage directly in the Pinecone console, or you can pull them into [Prometheus](https://prometheus.io/). For more details, see [Monitoring](/guides/production/monitoring).

## Monitor operation-level usage

### Read units

[Query](/guides/search/search-overview), [fetch](/guides/manage-data/fetch-data), and [list by ID](/guides/manage-data/list-record-ids) requests return a `usage` parameter with the [read unit](/guides/manage-cost/understanding-cost#read-units) consumption of each request that is made.

<Warning>
  While Pinecone tracks read unit usage with decimal precision, the Pinecone API and SDKs round these values up to the nearest whole number in query, fetch, and list responses. For example, if a query uses 0.45 read units, the API and SDKs will report it as 1 read unit.

  For precise read unit reporting, see [index-level metrics](/guides/production/monitoring) or the organization-wide [Usage dashboard](/guides/manage-cost/monitor-usage-and-costs#monitor-organization-level-usage-and-costs).
</Warning>

<Note>
  Indexes built on [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) are not subject to read unit limits for query, fetch, and list operations. For sizing and capacity planning guidance, see the [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) guide.
</Note>

Example query request:

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")
  index = pc.Index("example-index")

  response = index.query(
      vector=[0.22,0.43,0.16,1,...], 
      namespace='example-namespace', 
      top_k=3,
      include_values=False,
      include_metadata=False
  )

  print(response)
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })
  const index = pc.index("example-index")

  const queryResponse = await index.namespace('example-namespace').query({
      vector: [0.22,0.43,0.16,1,...],
      topK: 3,
      includeValues: false,
      includeMetadata: false,
  });

  console.log(queryResponse);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import io.pinecone.unsigned_indices_model.QueryResponseWithUnsignedIndices;

  import java.util.Arrays;
  import java.util.List;

  public class QueryByVector {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index,
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(config, connection, "example-index");

          List<Float> query = Arrays.asList(0.22f,0.43f,0.16f,1f,...);
          QueryResponseWithUnsignedIndices queryResponse = index.query(3, query, null, null, null, "example-namespace", null, false, false);
          System.out.println(queryResponse);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

  	// To get the unique host for an index,
  	// see https://docs.pinecone.io/guides/manage-data/target-an-index
  	idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
  	if err != nil {
  		log.Fatalf("Failed to create IndexConnection for Host: %v", err)
  	}

  	queryVector := []float32{0.22, 0.43, 0.16, 1, ...}

  	res, err := idxConnection.QueryByVectorValues(ctx, &pinecone.QueryByVectorValuesRequest{
  		Vector:        queryVector,
  		TopK:          3,
  		IncludeValues: false,
  	})
  	if err != nil {
  		log.Fatalf("Error encountered when querying by vector: %v", err)
  	} else {
  		fmt.Printf(prettifyStruct(res))
  	}
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var queryResponse = await index.QueryAsync(new QueryRequest {
      Vector = new[] { 0.22f,0.43f,0.16f,1f,... },
      Namespace = "example-namespace",
      TopK = 3,
      IncludeMetadata = false,
  });
  ```
</CodeGroup>

The response looks like this:

<CodeGroup>
  ```python Python theme={null}
  {'matches': [{'id': 'record_193027', 'score': 0.00405937387, 'values': []},
               {'id': 'record_137452', 'score': 0.00405937387, 'values': []},
               {'id': 'record_132264', 'score': 0.00405937387, 'values': []}],
   'namespace': 'example-namespace',
   'usage': {'read_units': 1}}
  ```

  ```javascript JavaScript theme={null}
  {
    matches: [
      {
        id: 'record_186225',
        score: 0.00405937387,
        values: [],
        sparseValues: undefined,
        metadata: undefined
      },
      {
        id: 'record_164994',
        score: 0.00405937387,
        values: [],
        sparseValues: undefined,
        metadata: undefined
      },
      {
        id: 'record_186333',
        score: 0.00405937387,
        values: [],
        sparseValues: undefined,
        metadata: undefined
      }
    ],
    namespace: 'example-namespace',
    usage: { readUnits: 1 }
  }
  ```

  ```java Java theme={null}
  class QueryResponseWithUnsignedIndices {
      matches: [ScoredVectorWithUnsignedIndices {
          score: 0.004059374
          id: record_170370
          values: []
          metadata: 
          sparseValuesWithUnsignedIndices: SparseValuesWithUnsignedIndices {
              indicesWithUnsigned32Int: []
              values: []
          }
      }, ScoredVectorWithUnsignedIndices {
          score: 0.004059374
          id: record_107423
          values: []
          metadata: 
          sparseValuesWithUnsignedIndices: SparseValuesWithUnsignedIndices {
              indicesWithUnsigned32Int: []
              values: []
          }
      }, ScoredVectorWithUnsignedIndices {
          score: 0.004059374
          id: record_171426
          values: []
          metadata: 
          sparseValuesWithUnsignedIndices: SparseValuesWithUnsignedIndices {
              indicesWithUnsigned32Int: []
              values: []
          }
      }]
      namespace: example-index
      usage: read_units: 1
  }
  ```

  ```go Go theme={null}
  {
    "matches": [
      {
        "vector": {
          "id": "record_193027"
        },
        "score": 0.004059374
      },
      {
        "vector": {
          "id": "record_137452"
        },
        "score": 0.004059374
      },
      {
        "vector": {
          "id": "record_132264"
        },
        "score": 0.004059374
      }
    ],
    "usage": {
      "read_units": 1
    },
    "namespace": "example-index"
  }
  ```

  ```csharp C# theme={null}
  {
    "results": [],
    "matches": [
      {
        "id": "record_193027",
        "score": 0.004059374,
        "values": []
      },
      {
        "id": "record_137452",
        "score": 0.004059374,
        "values": []
      },
      {
        "id": "record_132264",
        "score": 0.004059374,
        "values": []
      }
    ],
    "namespace": "example-namespace",
    "usage": {
      "readUnits": 1
    }
  }
  ```
</CodeGroup>

For a more in-depth demonstration of how to use read units to inspect read costs, see [this notebook](https://github.com/pinecone-io/examples/blob/master/docs/read-units-demonstrated.ipynb).

### Embedding tokens

Requests to one of [Pinecone's hosted embedding models](/guides/index-data/create-an-index#embedding-models), either directly via the [`embed` operation](/reference/api/latest/inference/generate-embeddings) or automatically when upserting or querying an [index with integrated embedding](/guides/index-data/indexing-overview#integrated-embedding), return a `usage` parameter with the total tokens generated.

For example, the following request to use the `multilingual-e5-large` model to generate embeddings for sentences related to the word “apple” might return this request and summary of embedding tokens generated:

<CodeGroup>
  ```python Python theme={null}
  # Import the Pinecone library
  from pinecone.grpc import PineconeGRPC as Pinecone
  from pinecone import ServerlessSpec
  import time

  # Initialize a Pinecone client with your API key
  pc = Pinecone(api_key="YOUR_API_KEY")

  # Define a sample dataset where each item has a unique ID and piece of text
  data = [
      {"id": "vec1", "text": "Apple is a popular fruit known for its sweetness and crisp texture."},
      {"id": "vec2", "text": "The tech company Apple is known for its innovative products like the iPhone."},
      {"id": "vec3", "text": "Many people enjoy eating apples as a healthy snack."},
      {"id": "vec4", "text": "Apple Inc. has revolutionized the tech industry with its sleek designs and user-friendly interfaces."},
      {"id": "vec5", "text": "An apple a day keeps the doctor away, as the saying goes."},
      {"id": "vec6", "text": "Apple Computer Company was founded on April 1, 1976, by Steve Jobs, Steve Wozniak, and Ronald Wayne as a partnership."}
  ]

  # Convert the text into numerical vectors that Pinecone can index
  embeddings = pc.inference.embed(
      model="llama-text-embed-v2",
      inputs=[d['text'] for d in data],
      parameters={"input_type": "passage", "truncate": "END"}
  )

  print(embeddings)
  ```

  ```javascript JavaScript theme={null}
  // Import the Pinecone library
  import { Pinecone } from '@pinecone-database/pinecone';

  // Initialize a Pinecone client with your API key
  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  // Define a sample dataset where each item has a unique ID and piece of text
  const data = [
    { id: 'vec1', text: 'Apple is a popular fruit known for its sweetness and crisp texture.' },
    { id: 'vec2', text: 'The tech company Apple is known for its innovative products like the iPhone.' },
    { id: 'vec3', text: 'Many people enjoy eating apples as a healthy snack.' },
    { id: 'vec4', text: 'Apple Inc. has revolutionized the tech industry with its sleek designs and user-friendly interfaces.' },
    { id: 'vec5', text: 'An apple a day keeps the doctor away, as the saying goes.' },
    { id: 'vec6', text: 'Apple Computer Company was founded on April 1, 1976, by Steve Jobs, Steve Wozniak, and Ronald Wayne as a partnership.' }
  ];

  // Convert the text into numerical vectors that Pinecone can index
  const model = 'llama-text-embed-v2';

  const embeddings = await pc.inference.embed(
    model,
    data.map(d => d.text),
    { inputType: 'passage', truncate: 'END' }
  );

  console.log(embeddings);
  ```

  ```java Java theme={null}
  // Import the required classes
  import io.pinecone.clients.Index;
  import io.pinecone.clients.Inference;
  import io.pinecone.clients.Pinecone;
  import org.openapitools.inference.client.ApiException;
  import org.openapitools.inference.client.model.Embedding;
  import org.openapitools.inference.client.model.EmbeddingsList;

  import java.math.BigDecimal;
  import java.util.*;
  import java.util.stream.Collectors;

  public class GenerateEmbeddings {
      public static void main(String[] args) throws ApiException {
          // Initialize a Pinecone client with your API key
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          Inference inference = pc.getInferenceClient();

          // Prepare input sentences to be embedded
          List<DataObject> data = Arrays.asList(
              new DataObject("vec1", "Apple is a popular fruit known for its sweetness and crisp texture."),
              new DataObject("vec2", "The tech company Apple is known for its innovative products like the iPhone."),
              new DataObject("vec3", "Many people enjoy eating apples as a healthy snack."),
              new DataObject("vec4", "Apple Inc. has revolutionized the tech industry with its sleek designs and user-friendly interfaces."),
              new DataObject("vec5", "An apple a day keeps the doctor away, as the saying goes."),
              new DataObject("vec6", "Apple Computer Company was founded on April 1, 1976, by Steve Jobs, Steve Wozniak, and Ronald Wayne as a partnership.")
          );

          List<String> inputs = data.stream()
              .map(DataObject::getText)
              .collect(Collectors.toList());

          // Specify the embedding model and parameters
          String embeddingModel = "llama-text-embed-v2";

          Map<String, Object> parameters = new HashMap<>();
          parameters.put("input_type", "passage");
          parameters.put("truncate", "END");

          // Generate embeddings for the input data
          EmbeddingsList embeddings = inference.embed(embeddingModel, parameters, inputs);

          // Get embedded data
          List<Embedding> embeddedData = embeddings.getData();
      }

      private static List<Float> convertBigDecimalToFloat(List<BigDecimal> bigDecimalValues) {
          return bigDecimalValues.stream()
              .map(BigDecimal::floatValue)
              .collect(Collectors.toList());
      }
  }

  class DataObject {
      private String id;
      private String text;

      public DataObject(String id, String text) {
          this.id = id;
          this.text = text;
      }

      public String getId() {
          return id;
      }
      public String getText() {
          return text;
      }
  }
  ```

  ```go Go theme={null}
  package main

  // Import the required packages
  import (
      "context"
     	"encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  type Data struct {
      ID   string
      Text string
  }

  type Query struct {
  	Text string
  }

  func prettifyStruct(obj interface{}) string {
      bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      // Initialize a Pinecone client with your API key
      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // Define a sample dataset where each item has a unique ID and piece of text
      data := []Data{
          {ID: "vec1", Text: "Apple is a popular fruit known for its sweetness and crisp texture."},
          {ID: "vec2", Text: "The tech company Apple is known for its innovative products like the iPhone."},
          {ID: "vec3", Text: "Many people enjoy eating apples as a healthy snack."},
          {ID: "vec4", Text: "Apple Inc. has revolutionized the tech industry with its sleek designs and user-friendly interfaces."},
          {ID: "vec5", Text: "An apple a day keeps the doctor away, as the saying goes."},
          {ID: "vec6", Text: "Apple Computer Company was founded on April 1, 1976, by Steve Jobs, Steve Wozniak, and Ronald Wayne as a partnership."},
      }

      // Specify the embedding model and parameters
      embeddingModel := "llama-text-embed-v2"

      docParameters := pinecone.EmbedParameters{
          InputType: "passage",
          Truncate:  "END",
      }

      // Convert the text into numerical vectors that Pinecone can index
      var documents []string
      for _, d := range data {
          documents = append(documents, d.Text)
      }

      docEmbeddingsResponse, err := pc.Inference.Embed(ctx, &pinecone.EmbedRequest{
          Model:      embeddingModel,
          TextInputs: documents,
          Parameters: docParameters,
      }) 
      if err != nil {
          log.Fatalf("Failed to embed documents: %v", err)
      } else {
          fmt.Printf(prettifyStruct(docEmbeddingsResponse))
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;
  using System;
  using System.Collections.Generic;

  // Initialize a Pinecone client with your API key
  var pinecone = new PineconeClient("YOUR_API_KEY");

  // Prepare input sentences to be embedded
  var data = new[]
  {
      new
      {
          Id = "vec1",
          Text = "Apple is a popular fruit known for its sweetness and crisp texture."
      },
      new
      {
          Id = "vec2",
          Text = "The tech company Apple is known for its innovative products like the iPhone."
      },
      new
      {
          Id = "vec3",
          Text = "Many people enjoy eating apples as a healthy snack."
      },
      new
      {
          Id = "vec4",
          Text = "Apple Inc. has revolutionized the tech industry with its sleek designs and user-friendly interfaces."
      },
      new
      {
          Id = "vec5",
          Text = "An apple a day keeps the doctor away, as the saying goes."
      },
      new
      {
          Id = "vec6",
          Text = "Apple Computer Company was founded on April 1, 1976, by Steve Jobs, Steve Wozniak, and Ronald Wayne as a partnership."
      }
  };

  // Specify the embedding model and parameters
  var embeddingModel = "llama-text-embed-v2";

  // Generate embeddings for the input data
  var embeddings = await pinecone.Inference.EmbedAsync(new EmbedRequest
  {
      Model = embeddingModel,
      Inputs = data.Select(item => new EmbedRequestInputsItem { Text = item.Text }),
      Parameters = new Dictionary<string, object?>
      {
          ["input_type"] = "passage",
          ["truncate"] = "END"
      }
  });

  Console.WriteLine(embeddings);
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl https://api.pinecone.io/embed \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
        "model": "llama-text-embed-v2",
        "parameters": {
          "input_type": "passage",
          "truncate": "END"
        },
        "inputs": [
          {"text": "Apple is a popular fruit known for its sweetness and crisp texture."},
          {"text": "The tech company Apple is known for its innovative products like the iPhone."},
          {"text": "Many people enjoy eating apples as a healthy snack."},
          {"text": "Apple Inc. has revolutionized the tech industry with its sleek designs and user-friendly interfaces."},
          {"text": "An apple a day keeps the doctor away, as the saying goes."},
          {"text": "Apple Computer Company was founded on April 1, 1976, by Steve Jobs, Steve Wozniak, and Ronald Wayne as a partnership."}
        ]
    }'
  ```
</CodeGroup>

The returned object looks like this:

<CodeGroup>
  ```python Python theme={null}
  EmbeddingsList(
      model='llama-text-embed-v2',
      data=[
          {'values': [0.04925537109375, -0.01313018798828125, -0.0112762451171875, ...]},
          ...
      ],
      usage={'total_tokens': 130}
  )
  ```

  ```javascript JavaScript theme={null}
  EmbeddingsList(1) [
    {
      values: [
        0.04925537109375, 
        -0.01313018798828125, 
        -0.0112762451171875,
        ...
      ]
    },
    ...
    model: 'llama-text-embed-v2',
    data: [ { values: [Array] } ],
    usage: { totalTokens: 130 }
  ]
  ```

  ```java Java theme={null}
  class EmbeddingsList {
      model: llama-text-embed-v2
      data: [class Embedding {
          values: [0.04925537109375, -0.01313018798828125, -0.0112762451171875, ...]
          additionalProperties: null
      }, ...]
      usage: class EmbeddingsListUsage {
          totalTokens: 130
          additionalProperties: null
      }
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  {
    "data": [
      {
        "values": [
          0.03942871,
          -0.010177612,
          -0.046051025,
          ...
        ]
      },
      ...
    ], 
    "model": "llama-text-embed-v2",
    "usage": {
      "total_tokens": 130
    }
  }
  ```

  ```csharp C# theme={null}
  {
    "model": "llama-text-embed-v2",
    "data": [
      {
        "values": [
          0.04913330078125,
          -0.01306915283203125,
          -0.01116180419921875,
          ...
        ]
      },
      ...
    ],
    "usage": {
      "total_tokens": 130
    }
  }
  ```

  ```json curl theme={null}
  {
    "data": [
      {
        "values": [
          0.04925537109375,
          -0.01313018798828125,
          -0.0112762451171875,
          ...
        ]
      }, 
      ...
    ],
    "model": "llama-text-embed-v2",
    "usage": {
      "total_tokens": 130
    }
  }
  ```
</CodeGroup>

## See also

* [Understanding cost](/guides/manage-cost/understanding-cost)
* [Manage cost](/guides/manage-cost/manage-cost)

---

# Source: https://docs.pinecone.io/guides/production/monitoring.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Monitor performance

> Monitor performance metrics in the Pinecone console or with Prometheus or Datadog.

Pinecone generates time-series performance metrics for each Pinecone index. You can monitor these metrics directly in the Pinecone console or with tools like Prometheus or Datadog.

## Monitor in the Pinecone Console

To view performance metrics in the Pinecone console:

1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).
2. Select the project containing the index you want to monitor.
3. Go to **Database > Indexes**.
4. Select the index.
5. Go to the **Metrics** tab.

## Monitor with Datadog

To monitor Pinecone with Datadog, use Datadog's [Pinecone integration](/integrations/datadog).

<Note>
  This feature is available on [Standard and Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

## Monitor with Prometheus

<Note>
  This feature is available on [Standard and  Enterprise plans](https://www.pinecone.io/pricing/). When using [Bring Your Own Cloud](/guides/production/bring-your-own-cloud), you must configure Prometheus monitoring within your VPC.
</Note>

To monitor all serverless indexes in a project, insert the following snippet into the [`scrape_configs`](https://prometheus.io/docs/prometheus/latest/configuration/configuration/#scrape_config) section of your `prometheus.yml` file and update it with values for your Prometheus integration:

<Note>
  This method uses [HTTP service discovery](https://prometheus.io/docs/prometheus/latest/http_sd/) to automatically discover and target all serverless indexes across all regions in a project.
</Note>

```YAML  theme={null}
global:
  scrape_interval: 15s

scrape_configs:
  - job_name: 'pinecone-serverless-metrics'
    http_sd_configs:
      - url: https://api.pinecone.io/prometheus/projects/PROJECT_ID/metrics/discovery
        refresh_interval: 1m
        authorization:
          type: Bearer
          credentials: API_KEY
    authorization:
      type: Bearer
      credentials: API_KEY
```

* Replace `PROJECT_ID` with the unique ID of the project you want to monitor. You can [find the project ID](/guides/projects/understanding-projects#project-ids) in the Pinecone console.

* Replace both instances of `API_KEY` with an API key for the project you want to monitor. The first instance is for service discovery, and the second instance is for the discovered targets. If necessary, you can [create an new API key](/guides/projects/manage-api-keys) in the Pinecone console.

For more configuration details, see the [Prometheus docs](https://prometheus.io/docs/prometheus/latest/configuration/configuration/).

### Available metrics

The following metrics are available when you integrate Pinecone with Prometheus:

<div className="border border-gray-200 dark:border-gray-700 rounded-lg p-4 my-4">
  <Tabs>
    <Tab title="Metrics">
      | Name                                 | Type    | Description                                                                                                 |
      | :----------------------------------- | :------ | :---------------------------------------------------------------------------------------------------------- |
      | `pinecone_db_record_total`           | gauge   | The total number of records in the index.                                                                   |
      | `pinecone_db_storage_size_bytes`     | gauge   | The total size of the index in bytes.                                                                       |
      | `pinecone_db_op_upsert_count`        | counter | The number of [upsert](/guides/index-data/upsert-data) requests.                                            |
      | `pinecone_db_op_upsert_duration_sum` | counter | The total time taken processing [upsert](/guides/index-data/upsert-data) requests in milliseconds.          |
      | `pinecone_db_op_query_count`         | counter | The number of [query](/guides/search/search-overview) requests.                                             |
      | `pinecone_db_op_query_duration_sum`  | counter | The total time taken processing [query](/guides/search/search-overview) requests in milliseconds.           |
      | `pinecone_db_op_fetch_count`         | counter | The number of [fetch](/guides/manage-data/fetch-data) requests.                                             |
      | `pinecone_db_op_fetch_duration_sum`  | counter | The total time taken processing [fetch](/guides/manage-data/fetch-data) requests in milliseconds.           |
      | `pinecone_db_op_update_count`        | counter | The number of [update](/guides/manage-data/update-data) requests.                                           |
      | `pinecone_db_op_update_duration_sum` | counter | The total time taken processing [update](/guides/manage-data/update-data) requests in milliseconds.         |
      | `pinecone_db_op_delete_count`        | counter | The number of [delete](/guides/manage-data/delete-data) requests.                                           |
      | `pinecone_db_op_delete_duration_sum` | counter | The total time taken processing [delete](/guides/manage-data/delete-data) requests in milliseconds.         |
      | `pinecone_db_write_unit_count`       | counter | The total number of [write units](/guides/manage-cost/understanding-cost#write-units) consumed by an index. |
      | `pinecone_db_read_unit_count`        | counter | The total number of [read units](/guides/manage-cost/understanding-cost#read-units) consumed by an index.   |
    </Tab>

    <Tab title="Metric name migration">
      Some metric names changed on December 19, 2025. The following metrics were renamed:

      | Previous name (before Dec 19, 2025)    | Current name                                 |
      | :------------------------------------- | :------------------------------------------- |
      | `pinecone_db_record_total`             | `pinecone_db_record_total` (no change)       |
      | `pinecone_db_storage_size_bytes`       | `pinecone_db_storage_size_bytes` (no change) |
      | `pinecone_db_op_upsert_total`          | `pinecone_db_op_upsert_count`                |
      | `pinecone_db_op_upsert_duration_total` | `pinecone_db_op_upsert_duration_sum`         |
      | `pinecone_db_op_query_total`           | `pinecone_db_op_query_count`                 |
      | `pinecone_db_op_query_duration_total`  | `pinecone_db_op_query_duration_sum`          |
      | `pinecone_db_op_fetch_total`           | `pinecone_db_op_fetch_count`                 |
      | `pinecone_db_op_fetch_duration_total`  | `pinecone_db_op_fetch_duration_sum`          |
      | `pinecone_db_op_update_total`          | `pinecone_db_op_update_count`                |
      | `pinecone_db_op_update_duration_total` | `pinecone_db_op_update_duration_sum`         |
      | `pinecone_db_op_delete_total`          | `pinecone_db_op_delete_count`                |
      | `pinecone_db_op_delete_duration_total` | `pinecone_db_op_delete_duration_sum`         |
      | `pinecone_db_write_unit_total`         | `pinecone_db_write_unit_count`               |
      | `pinecone_db_read_unit_total`          | `pinecone_db_read_unit_count`                |
    </Tab>
  </Tabs>
</div>

### Metric labels

Each metric contains the following labels:

| Label           | Description                                                  |
| :-------------- | :----------------------------------------------------------- |
| `index_name`    | Name of the index to which the metric applies.               |
| `cloud`         | Cloud where the index is deployed: `aws`, `gcp`, or `azure`. |
| `region`        | Region where the index is deployed.                          |
| `capacity_mode` | Type of index: `serverless` or `byoc`.                       |
| `instance`      | Server instance (only available for counter metrics).        |

### Example queries

Return the total number of records per index:

```shell  theme={null}
sum by (index_name) (pinecone_db_record_total)
```

Return the total number of records in Pinecone index `docs-example`:

```shell  theme={null}
pinecone_db_record_total{index_name="docs-example"}
```

For each index, return the total number of upsert requests per second:

```shell  theme={null}
sum by (index_name) (rate(pinecone_db_op_upsert_count[5m]))
```

Return the average processing time in milliseconds for upsert requests per index:

```shell  theme={null}
(sum by (index_name) (rate(pinecone_db_op_upsert_duration_sum[1m])))/(sum by (index_name) (rate(pinecone_db_op_upsert_count[1m])))
```

For each index, return the total number of read units consumed per second:

```shell  theme={null}
sum by (index_name) (rate(pinecone_db_read_unit_count[5m]))
```

Return the total write units consumed per second for the Pinecone index `docs-example`:

```shell  theme={null}
sum (rate(pinecone_db_write_unit_count{index_name="docs-example"}[5m]))
```

---

# Source: https://docs.pinecone.io/guides/assistant/multimodal.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Multimodal context for assistants

> Process images and charts in PDFs with multimodal assistants.

<Note>
  This feature is in [public preview](/release-notes/feature-availability).
</Note>

Pinecone assistants support multimodal context, allowing them to understand and respond to questions about images embedded in PDF documents.

This enables use cases like:

* Analyzing charts, graphs, and diagrams in financial reports
* Understanding infographics and visual data in research papers
* Interpreting visual layouts in technical documentation

When working with multimodal PDFs, assistants attempt to filter out purely decorative images (such as example logos, background graphics, generic stock photos), so they can focus on images that contain meaningful information.

Additionally, assistants use Optical Character Recognition (OCR) to extract text from images. This allows them to read and analyze scanned PDFs (PDFs that contain images of text, but no actual embedded text).

## How it works

When you enable multimodal context for a PDF:

1. Pinecone extracts text and images (raster or vector) from the file and analyzes their contents. For each image, the assistant generates a descriptive caption and set of keywords. Additionally, when it makes sense, the assistant captures data points found in the image (for example, values from a table or chart).
2. During chat or context queries, the assistant searches for relevant text and image context it captured when analyzing the PDF. Image context can include the original image data (base64-encoded).
3. The assistant passes this context to the LLM, which uses it to generate responses.

<Note>
  For an overview of how Pinecone Assistant works, see [Pinecone Assistant architecture](/reference/architecture/assistant-architecture).
</Note>

## Try it out

The following steps demonstrate how to create an assistant, provide it with a PDF that contains images, and then query that assistant using chat and context APIs.

<Note>
  All versions of Pinecone's Assistant API allow you to upload multimodal PDFs.
</Note>

### 1. Create an assistant

First, if you don't have one, [create an assistant](/reference/api/2025-10/assistant/create_assistant):

<CodeGroup>
  ```Python Python theme={null}
  from pprint import pprint
  from pinecone import Pinecone

  pc = Pinecone("YOUR_API_KEY") 
  assistant = pc.assistant.create_assistant(
      assistant_name="example-assistant-multimodal", 
      instructions="You are a helpful assistant that can understand both text and images in documents.",
      region="us",
      timeout=30
  )

  print(f"Type: {type(assistant).__name__}")
  pprint(assistant)
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://api.pinecone.io/assistant/assistants" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
          "name": "example-assistant-multimodal",
          "instructions": "You are a helpful assistant that can understand both text and images in documents.",
          "region": "us"
        }'
  ```
</CodeGroup>

Response:

<CodeGroup>
  ```shell Python theme={null}
  Type: AssistantModel
  {'created_at': '2025-08-28T23:35:26.917953498Z',
    'host': 'https://prod-1-data.ke.pinecone.io',
    'instructions': 'You are a helpful assistant that can understand both text '
                    'and images in documents.',
    'metadata': {},
    'name': 'example-assistant-multimodal',
    'status': 'Ready',
    'updated_at': '2025-08-28T23:35:28.507639215Z'}
  ```

  ```json curl theme={null}
  {
    "name": "example-assistant-multimodal",
    "instructions": "You are a helpful assistant that can understand both text and images in documents.",
    "metadata": null,
    "status": "Initializing",
    "host": "https://prod-1-data.ke.pinecone.io",
    "created_at": "2025-08-18T23:18:52.858197495Z",
    "updated_at": "2025-08-18T23:18:52.858198077Z"
  }
  ```
</CodeGroup>

<Note>
  You don't need to create a new assistant to use multimodal context. Existing assistants can enable multimodal context for newly uploaded PDFs, as described in [the next section](#2-upload-a-multimodal-pdf).
</Note>

### 2. Upload a multimodal PDF

To enable multimodal context for a PDF, when [uploading the file](/reference/api/2025-10/assistant/upload_file), set the `multimodal` URL parameter to true (defaults to false).

<Tip>
  To improve retrieval accuracy for images found in uploaded PDFs, include relevant contextual text on the same page as each image. For example:

  * **Graphs and charts**: Include nearby text explaining the experiment, methodology, or data being visualized.
  * **Diagrams**: Add descriptive labels or explanations adjacent to technical diagrams.
  * **Tables**: Provide context about what the data represents and any relevant methodology.

  The assistant uses surrounding text when generating captions that are later passed to the LLM, so placing relevant context near images improves caption quality and retrieval accuracy.
</Tip>

<CodeGroup>
  ```Python Python theme={null}
  from pprint import pprint
  from pinecone import Pinecone

  pc = Pinecone("YOUR_API_KEY") 
  assistant = pc.assistant.Assistant(assistant_name="example-assistant-multimodal")

  # timeout=None allows the SDK to wait for file processing to complete before returning.
  # This parameter is only available in the SDK, not in direct API calls.
  file_model = assistant.upload_file(
      file_path="./document.pdf",
      multimodal=True,
      timeout=None
  )

  pprint(file_model)
  ```

  ```bash curl theme={null}
  ASSISTANT_HOST="YOUR_ASSISTANT_HOST"
  ASSISTANT_NAME="example-assistant-multimodal"
  PINECONE_API_KEY="YOUR_API_KEY"
  LOCAL_FILE_PATH="/path/to/your/document.pdf"

  curl -X POST "https://$ASSISTANT_HOST/assistant/files/$ASSISTANT_NAME?multimodal=true" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -F "file=@$LOCAL_FILE_PATH"
  ```
</CodeGroup>

Response:

<CodeGroup>
  ```shell Python theme={null}
  # Formatted for readability
  FileModel(
    name='document.pdf', 
    id='9c322597-58d6-4ebc-84b5-a398b620da01', 
    metadata=None, 
    created_on='2025-08-28T23:41:41.982805815Z', 
    updated_on='2025-08-28T23:42:09.562949544Z', 
    status='Available', 
    percent_done=1.0, 
    signed_url=None, 
    error_message=None, 
    size=1236044.0, 
    multimodal=True
  )
  ```

  ```json curl theme={null}
  {
    "status": "Processing",
    "id": "a89f678c-9ceb-40ec-8fc7-23913e560b37",
    "name": "document.pdf",
    "size": 1236044,
    "metadata": null,
    "multimodal": true,
    "updated_on": "2025-08-18T23:21:36.516310298Z",
    "created_on": "2025-08-18T23:21:36.516310771Z",
    "percent_done": 0.0,
    "signed_url": null,
    "error_message": null
  }
  ```
</CodeGroup>

<Note>
  * The `multimodal` parameter is only available for PDF files.
  * To check the status of a file, use the [describe a file upload](/reference/api/2025-10/assistant/describe_file) endpoint.
  * If upload processing fails, you'll need to re-upload the file.
</Note>

### 3. Chat with the assistant

Now, [chat with your assistant](/reference/api/2025-10/assistant/chat_assistant). To tell the assistant to provide image-related context to the LLM:

* Set the `multimodal` request parameter to true (default) in the `context_options` object. Setting `multimodal` to false means the LLM only receives text snippets.
* When `multimodal` is true, use `include_binary_content` to specify what image context the LLM should receive: base64 image data and captions (true) or captions only (false).

<Note>
  Sending image-related context to the LLM (whether captions, base64 data, or both) increases token usage. Learn about [monitoring spend and usage](/guides/assistant/admin/monitor-spend-and-usage).
</Note>

<CodeGroup>
  ```Python Python theme={null}
  from pprint import pprint
  from pinecone import Pinecone
  from pinecone_plugins.assistant.models.chat import Message

  pc = Pinecone("YOUR_API_KEY") 
  assistant = pc.assistant.Assistant(assistant_name="example-assistant-multimodal")

  msg = Message(
      role="user", 
      content="Describe the symbol on the paper tray that indicates the maximum fill level."
  )

  chat_response = assistant.chat(
    messages=[msg],
    context_options={
        "multimodal": True,
        "include_binary_content": True,
        "top_k": 10,
        "snippet_size": 2048
    }
  )

  pprint(chat_response)
  ```

  ```bash curl theme={null}
  ASSISTANT_HOST="YOUR_ASSISTANT_HOST"
  ASSISTANT_NAME="example-assistant-multimodal"
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://$ASSISTANT_HOST/assistant/chat/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
          "messages": [
            {
              "role": "user",
              "content": "Describe the symbol on the paper tray that indicates the maximum fill level."
            }
          ],
          "context_options": {
            "multimodal": true,
            "include_binary_content": true,
            "top_k": 10,
            "snippet_size": 2048
          }
        }'
  ```
</CodeGroup>

Response:

<CodeGroup>
  ```shell Python theme={null}
  # Formatted for readability
  ChatResponse(
    id='00000000000000000fe49626f3ee5164', 
    model='gpt-4o-2024-11-20', 
    usage=Usage(
      prompt_tokens=8703, 
      completion_tokens=41, 
      total_tokens=8744
    ), 
    message=Message(
      content='The symbol on the paper tray that indicates...', 
      role='assistant'
    ), 
    finish_reason='stop', 
    citations=[
      Citation(
        position=209, 
        references=[
          Reference(
              file=FileModel(
                name='document.pdf', 
                id='9c322597-58d6-4ebc-84b5-a398b620da01', 
                metadata=None, 
                created_on='2025-08-28T23:41:41.982805815Z', 
                updated_on='2025-08-28T23:42:09.562949544Z', 
                status='Available', 
                percent_done=1.0, 
                signed_url='https://storage.googleapis.com/...', 
                error_message=None, 
                size=1236044.0, 
                multimodal=True
            ), 
            pages=[3, 4, 5, 6, 7, 8, 9, 10, 11], 
            highlight=None
          )
        ]
      )
    ]
  )
  ```

  ```json curl theme={null}
  {
    "finish_reason": "stop",
    "message": {
      "role": "assistant",
      "content": "The symbol on the paper tray that indicates..."
    },
    "id": "0000000000000000d904dfd3dd4f4597"
    "model": "gpt-4o-2024-11-20",
    "usage": {
      "prompt_tokens": 8414,
      "completion_tokens": 42,
      "total_tokens": 8456
    },
    "citations": [
      {
        "position": 213,
        "references": [
          {
            "file": {
              "status": "Available",
              "id": "a89f678c-9ceb-40ec-8fc7-23913e560b37",
              "name": "document.pdf",
              "size": 1236044,
              "metadata": null,
              "multimodal": true,
              "updated_on": "2025-08-18T23:21:59.697988967Z",
              "created_on": "2025-08-18T23:21:36.498381046Z",
              "percent_done": 1.0,
              "signed_url": "https://storage.googleapis.com/...",
              "error_message": null
            },
            "pages": [ 3, 4, 5, 6, 7, 8, 9, 10, 11 ],
            "highlight": null
          }
        ]
      }
    ]
  }
  ```
</CodeGroup>

<Warning>
  If your assistant uses multimodal context snippets to generate a response, no [highlights](/guides/assistant/chat-with-assistant#include-citation-highlights-in-the-response) are returned—even when `include_highlights` is true.
</Warning>

### 4. Query for context

To query context for a custom RAG workflow, you can [retrieve context snippets](/reference/api/2025-10/assistant/context_assistant) directly. Then, you can pass these snippets to an LLM as context.

To fetch image-related context snippets (as well as text snippets), set the `multimodal` request parameter to true (default). When `multimodal` is true, use `include_binary_content` to specify what image context you'd like to receive: base64 image data and captions (true) or captions only (false).

<CodeGroup>
  ```Python Python theme={null}
  from pprint import pprint
  from pinecone import Pinecone

  pc = Pinecone("PINECONE_API_KEY") 
  assistant = pc.assistant.Assistant(assistant_name="example-assistant-multimodal")
  context_response = assistant.context(
      query="Describe the symbol on the paper tray that indicates the maximum fill level.",
      multimodal=True,
      include_binary_content=True
  )

  pprint(context_response)
  ```

  ```bash curl theme={null}
  ASSISTANT_HOST="YOUR_ASSISTANT_HOST"
  ASSISTANT_NAME="example-assistant-multimodal"
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://$ASSISTANT_HOST/assistant/chat/$ASSISTANT_NAME/context" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "accept: application/json" \
    -H "Content-Type: application/json" \
    -d '{
      "query": "Describe the symbol on the paper tray that indicates the maximum fill level.",
      "multimodal": true,
      "include_binary_content": true
    }'
  ```
</CodeGroup>

<Note>
  If you set `multimodal` to true and `include_binary_content` to false, image objects are not returned in the snippets. If you set `multimodal` to false, only text snippets are returned.
</Note>

Response:

<CodeGroup>
  ```shell Python theme={null}
  # Formatted for readability
  ContextResponse(
    id='00000000000000001e3ef84bd493e612', 
    snippets=[
      MultimodalSnippet(
        type='multimodal', 
        content=[
          TextBlock(type='text', text="..."), 
          ImageBlock(
            type='image', 
            caption='...', 
            image=Image(mime_type='image/jpeg', data='...', type='base64')), 
          // ...
        ], 
        score=0.16321887, 
        reference=PdfReference(
          type='pdf', 
          pages=[3, 4, 5, 6, 7, 8, 9, 10, 11], 
          file=FileModel(
            name='document.pdf', 
            id='9c322597-58d6-4ebc-84b5-a398b620da01', 
            metadata=None, 
            created_on='2025-08-28T23:41:41.982805815Z', 
            updated_on='2025-08-28T23:42:09.562949544Z', 
            status='Available', 
            percent_done=1.0, 
            signed_url='https://storage.googleapis.com/...', 
            error_message=None, 
            size=1236044, 
            multimodal=True
          )
        )
      ), 
      // ...
    ], 
    usage=TokenCounts(
      prompt_tokens=7061, 
      completion_tokens=0, 
      total_tokens=7061
    )
  )
  ```

  ```json curl theme={null}
  {
    "snippets": [
      {
        "type": "multimodal",
        "content": [
          {
            "type": "text",
            "text": "# The Online User's Guide..."
          },
          {
            "type": "image",
            "caption": "An image of a control panel...",
            "image": {
              "type": "base64",
              "mime_type": "image/jpeg",
              "data": "..."
            }
          },
          // ...
        ],
        "score": 0.16775002,
        "reference": {
          "type": "pdf",
          "file": {
            "status": "Available",
            "id": "a89f678c-9ceb-40ec-8fc7-23913e560b37",
            "name": "document.pdf",
            "size": 1236044,
            "metadata": null,
            "multimodal": true,
            "updated_on": "2025-08-18T23:21:59.697988967Z",
            "created_on": "2025-08-18T23:21:36.498381046Z",
            "percent_done": 1.0,
            "signed_url": "https://storage.googleapis.com/...",
            "error_message": null
          },
          "pages": [ 3, 4, 5, 6, 7, 8, 9, 10, 11 ]
        }
      },
      // ...
    ],
    "usage": {
      "prompt_tokens": 6778,
      "completion_tokens": 0,
      "total_tokens": 6778
    },
    "id": "000000000000000005b9cd91b1c5446d"
  } 
  ```
</CodeGroup>

<Note>
  Snippets are returned based on their semantic relevance to the provided query. When you set `multimodal` to true, you'll receive the most relevant snippets, regardless of the types of content they contain. You can receive text snippets, multimodal snippets, or both.
</Note>

## Limits

Multimodal context for assistants is only available for PDF files. Additionally, the following limits apply:

| Metric                        | Starter plan | Standard plan | Enterprise plan |
| :---------------------------- | :----------- | :------------ | :-------------- |
| Max file size                 | 10 MB        | 50 MB         | 50 MB           |
| Page limit                    | 100          | 100           | 100             |
| Multimodal PDFs per assistant | 10           | 20            | 20              |

To learn about other assistant-related limits, see [Pinecone Assistant limits](/guides/assistant/pricing-and-limits).

---

# Source: https://docs.pinecone.io/guides/assistant/quickstart/n8n-quickstart.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Pinecone Assistant: n8n quickstart

> Create an n8n workflow to chat with documents using Pinecone Assistant and OpenAI.

Create an [n8n](https://docs.n8n.io/choose-n8n/) workflow that that downloads files via HTTP, uploads them to Pinecone Assistant, and enables you to chat with documents using OpenAI.

## 1. Get API keys

Your n8n workflow will need API keys for Pinecone and OpenAI.

<Steps>
  <Step title="Get a Pinecone API key">
    Create a new API key in the [Pinecone console](https://app.pinecone.io/organizations/-/keys), or use the widget below to generate a key. If you don't have a Pinecone account, the widget will sign you up for the free [Starter plan](https://www.pinecone.io/pricing/).

    <div style={{minWidth: '450px', minHeight:'152px'}}>
      <div id="pinecone-connect-widget">
        <div class="connect-widget-skeleton">
          <div class="skeleton-content" />
        </div>
      </div>
    </div>

    Your generated API key:

    ```shell  theme={null}
    "{{YOUR_API_KEY}}"
    ```
  </Step>

  <Step title="Get an OpenAI API key">
    Create a new API key in the [OpenAI console](https://platform.openai.com/api-keys).
  </Step>
</Steps>

## 2. Create an assistant

[Create an assistant](https://app.pinecone.io/organizations/-/projects/-/assistant) in the Pinecone console:

* Name your assistant `n8n-assistant`.

## 3. Set up n8n

<Steps>
  <Step title="Install the Pinecone Assistant node">
    In your n8n account, [install the Pinecone Assistant node](https://docs.n8n.io/integrations/community-nodes/installation/verified-install/) using the nodes panel.

    <Note>
      Restart your n8n instance if the Pinecone Assistant node doesn't appear in the nodes panel.
    </Note>
  </Step>

  <Step title="Create a new workflow">
    In your n8n account, [create a new workflow](https://docs.n8n.io/workflows/create/).
  </Step>

  <Step title="Import a workflow template">
    Copy this workflow template URL:

    ```shell  theme={null}
    https://raw.githubusercontent.com/pinecone-io/n8n-templates/refs/heads/main/assistant-quickstart/assistant-quickstart.json
    ```

    Paste the URL anywhere in the workflow editor and then click **Import** to add the workflow.
  </Step>

  <Step title="Add credentials to the workflow">
    * Add your Pinecone credentials:
      * In the **Upload file to Assistant** node, select **Credential to connect with > Create new credential** and paste in your Pinecone API key.

    * Add your OpenAI credentials:
      * In the **OpenAI Chat Model** node, select **Credential to connect with > Create new credential** and paste in your OpenAI API key.
  </Step>

  <Step title="Execute the workflow">
    By default, the workflow downloads recent Pinecone release notes and uploads them to your assistant. Click **Execute workflow** to start uploading documents.

    <Tip>
      You can add your own files to the workflow by changing the URLs in the **Set file urls** node.
    </Tip>
  </Step>
</Steps>

## 4. Chat with your docs

Once the documents are uploaded, you can chat with your assistant. In the n8n workflow, use the **Chat input** node to ask questions like:

```
What support does Pinecone have for MCP?
```

## Next steps

* Customize the workflow for your own use case:
  * Change the urls in the **Set file urls** node to use your own files.
  * Customize the system message on the **AI Agent** node to indicate what kind of knowledge is stored in Pinecone Assistant.
  * To help manage token consumption, add the Top K and/or Snippet Size parameters to the **Get context from Assistant** node.
  * Filter the context snippets even further by adding metadata filters to the **Get context from Assistant** node.
* Use n8n, Pinecone Assistant, and OpenAI to [chat with your Google Drive documents](https://n8n.io/workflows/9942-rag-powered-document-chat-with-google-drive-openai-and-pinecone-assistant/).
* Learn more about [Pinecone Assistant](/guides/assistant/overview).
* Get help in the [Pinecone Discord community](https://discord.gg/tJ8V62S3sH).

---

# Source: https://docs.pinecone.io/integrations/new-relic.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# New Relic

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

New Relic is an all-in-one observability platform and provides the industry’s first APM solution for AI-powered applications. New Relic is pioneering AI observability with AIM to provide engineers unprecedented visibility and insights across the AI application stack, making it easier to troubleshoot and optimize their AI applications for performance, quality, cost, and responsible use of AI.

Implement monitoring and integrate your Pinecone application with New Relic for performance analysis and insights. The New Relic for Pinecone (Prometheus) quickstart contains one dashboard. These interactive visualizations let you easily explore your data, understand context, and resolve problems faster. It also includes three alerts to detect changes in key performance metrics. Integrate these alerts with your favorite tools (like Slack, PagerDuty, etc.) and New Relic will let you know when something needs your attention.

<PrimarySecondaryCTA primaryLabel={"Get started"} primaryHref={"https://newrelic.com/instant-observability/pinecone-prometheus"} />

---

# Source: https://docs.pinecone.io/integrations/nexla.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Nexla

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Nexla is a Data + AI Integration Platform that makes it easy for users to build data pipelines in a no-code/low-code manner.

The Pinecone integration with Nexla makes it easy for enterprise users to ingest data from systems like Sharepoint, OneDrive, Cloud Storage, Data Warehouses, and 500+ other connectors that Nexla supports natively.

<PrimarySecondaryCTA primaryHref={"https://docs.nexla.com/tutorials/write-to-vector-db/"} primaryLabel={"Get started"} secondaryHref={"https://nexla.com/free-trial/"} secondaryLabel={"Sign up for Nexla"} />

---

# Source: https://docs.pinecone.io/reference/node-sdk.md

# Node.js SDK

<Tip>
  See the [Pinecone Node.js SDK
  documentation](https://sdk.pinecone.io/typescript/) for full installation
  instructions, usage examples, and reference information.

  To make a feature request or report an issue, please [file an issue](https://github.com/pinecone-io/pinecone-ts-client/issues).
</Tip>

## Requirements

The Pinecone Node SDK requires TypeScript 4.1 or later and Node 18.x or later.

## SDK versions

SDK versions are pinned to specific [API versions](/reference/api/versioning). When a new API version is released, a new version of the SDK is also released.

The mappings between API versions and Node.js SDK versions are as follows:

| API version        | SDK version |
| :----------------- | :---------- |
| `2025-04` (latest) | v6.x        |
| `2025-01`          | v5.x        |
| `2024-10`          | v4.x        |
| `2024-07`          | v3.x        |
| `2024-04`          | v2.x        |

When a new stable API version is released, you should upgrade your SDK to the latest version to ensure compatibility with the latest API changes.

## Install

To install the latest version of the [Node.js SDK](https://github.com/pinecone-io/pinecone-ts-client), written in TypeScript, run the following command:

```Shell  theme={null}
npm install @pinecone-database/pinecone
```

To check your SDK version, run the following command:

```Shell  theme={null}
npm list | grep @pinecone-database/pinecone
```

## Upgrade

If you already have the Node.js SDK, upgrade to the latest version as follows:

```Shell  theme={null}
npm install @pinecone-database/pinecone@latest
```

## Initialize

Once installed, you can import the library and then use an [API key](/guides/projects/manage-api-keys) to initialize a client instance:

```JavaScript  theme={null}
import { Pinecone } from '@pinecone-database/pinecone';

const pc = new Pinecone({
    apiKey: 'YOUR_API_KEY'
});
```

## Proxy configuration

If your network setup requires you to interact with Pinecone through a proxy, you can pass a custom `ProxyAgent` from the [`undici` library](https://undici.nodejs.org/#/). Below is an example of how to construct an `undici` `ProxyAgent` that routes network traffic through a [`mitm` proxy server](https://mitmproxy.org/) while hitting Pinecone's `/indexes` endpoint.

<Note>
  The following strategy relies on Node's native [`fetch`](https://nodejs.org/docs/latest/api/globals.html#fetch) implementation, released in Node v16 and stabilized in Node v21. If you are running Node versions 18-21, you may experience issues stemming from the instability of the feature. There are currently no known issues related to proxying in Node v18+.
</Note>

```JavaScript JavaScript theme={null}
import {
  Pinecone,
  type PineconeConfiguration,
} from '@pinecone-database/pinecone';
import { Dispatcher, ProxyAgent } from 'undici';
import * as fs from 'fs';

const cert = fs.readFileSync('path/to/mitmproxy-ca-cert.pem');

const client = new ProxyAgent({
  uri: 'https://your-proxy.com',
  requestTls: {
    port: 'YOUR_PROXY_SERVER_PORT',
    ca: cert,
    host: 'YOUR_PROXY_SERVER_HOST',
  },
});

const customFetch = (
  input: string | URL | Request,
  init: RequestInit | undefined
) => {
  return fetch(input, {
    ...init,
    dispatcher: client as Dispatcher,
    keepalive: true,  # optional
  });
};

const config: PineconeConfiguration = {
  apiKey:
    'YOUR_API_KEY',
  fetchApi: customFetch,
};

const pc = new Pinecone(config);

const indexes = async () => {
  return await pc.listIndexes();
};

indexes().then((response) => {
  console.log('My indexes: ', response);
});
```

---

# Source: https://docs.pinecone.io/troubleshooting/nodejs-troubleshooting.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Node.js Troubleshooting

There could be several reasons why a [Node.js application](/reference/sdks/node/overview) works in development mode but not in deployment.

In order to troubleshoot the issue, it's important to identify where the application is failing and compare the development and deployment environments to see what differences exist. It's also important to review any error messages or logs that are generated to help identify the issue.

You may also reach out to our [community of Pinecone users](https://community.pinecone.io) for help.

Here are a few aspects to troubleshoot:

## Dependency version mismatch

Sometimes, different environments have different versions of dependencies installed. If the application was developed using a specific version of a dependency, and that version is not installed on the deployment environment, the application may not work as expected.

## Environment configuration

The development environment may have different configurations from the deployment environment. For example, the development environment may have different environment variables set or different network settings. If the application relies on specific configuration settings that are not present in the deployment environment, it may not work.

## Permissions

The application may require permissions to access certain resources that are only granted in the development environment. For example, if the application needs to write to a specific directory, the permissions to write to that directory may only be granted in the development environment.

## Database connection

If the application relies on a database connection, it's possible that the connection settings are different in the deployment environment. For example, the database may have a different hostname or port number.

## Code optimization

During development, the application may have been running on a development server that did not optimize the code. However, when deployed, the application may be running on a production server that is optimized for performance. If there are code issues or performance bottlenecks, they may only appear when the application is deployed.

## Install fetch

It may be necessary to install the `fetch` Python library for compatibility with node.js.

---

# Source: https://docs.pinecone.io/examples/notebooks.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Notebooks

export const UtilityExampleCard = ({title, text, link}) => {
  return <a href={link} className="example-card group">
            <h2 className="font-bold" style={{
    fontSize: "0.875rem"
  }}>{title}</h2>
            <p style={{
    fontSize: "0.875rem",
    marginTop: "0"
  }}>{text}</p>
        </a>;
};

export const ExampleCard = ({title, text, link, children, arrow, vectors, namespaces}) => {
  return <a href={link} className="example-card group">
            <h2 className="font-semibold text-base">{title}</h2>
            <p>{text}</p>

            {children && <div className="tags">{children}</div>}

            {arrow && <svg xmlns="http://www.w3.org/2000/svg" className="arrow" width="24" height="24" viewBox="0 0 24 24" fill="none">
                <path d="M5.30739 20L4 18.6926L16.8249 5.8677H5.05837V4H20V18.9416H18.1323V7.1751L5.30739 20Z" fill="var(--text-secondary)" />
            </svg>}

            {(vectors || namespaces) && <div className="vectors">
                {vectors && <span>{vectors} vectors</span>}
                {namespaces && <span>{namespaces} namespaces</span>}
            </div>}
        </a>;
};

export const Tag = ({text, icon}) => {
  return <span className="card-tag">
            {icon && <img src={icon} className={`w-4 h-4 object-contain ${icon.includes("openai") ? "dark-inverted" : ""}`} />}
            {text}
        </span>;
};

<h2 className="examples-h2">Search</h2>

<div className="card-grid not-prose">
  <ExampleCard title="Semantic search" text="Implement semantic search over a dense index to find records that are similar in meaning to a given query." link="https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/semantic-search.ipynb" arrow>
    <Tag text="Pinecone Integrated Inference" />

    <Tag text="Hugging Face Datasets" icon="/images/examples/huggingface-icon.svg" />

    <Tag text="llama-text-embed-v2" />
  </ExampleCard>

  <ExampleCard title="Lexical search" text="Implement lexical search over a sparse index to find records that most exactly match the words or phrases in a query." link="https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/lexical-search.ipynb" arrow>
    <Tag text="Pinecone Integrated Inference" />

    <Tag text="pinecone-sparse-english-v0" />

    <Tag text="bge-reranker-v2-m3" />
  </ExampleCard>

  <ExampleCard title="Cascading retrieval" text="Implement cascading retrieval (hybrid search with two indexes) to combine the benefits of semantic and lexical search." link="https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/cascading-retrieval.ipynb" arrow>
    <Tag text="Pinecone Integrated Inference" />

    <Tag text="llama-text-embed-v2" />

    <Tag text="pinecone-sparse-english-v0" />

    <Tag text="bge-reranker-v2-m3" />
  </ExampleCard>

  <ExampleCard title="Reranking search results" text="Use Pinecone's reranking feature to enhance the accuracy of search results." link="https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/pinecone-reranker.ipynb" arrow>
    <Tag text="Pinecone Inference" />

    <Tag text="bge-reranker-v2-m3" icon="https://cdn.sanity.io/images/vr8gru94/production/40b1d05ee1325e6d9e4886af4e76ff06d844faff-188x188.jpg" />
  </ExampleCard>
</div>

<h2 className="examples-h2">Retrieval-augmented generation (RAG)</h2>

<div className="card-grid not-prose">
  <ExampleCard title="RAG with hybrid search and Claude" text="Implement simple retrieval-augmented generation with hybrid search and Anthropic's Claude models" link="https://colab.research.google.com/github/pinecone-io/examples/blob/master/learn/generation/traditional-rag/traditional-rag-with-claude-and-hybrid.ipynb" arrow>
    <Tag text="claude-3-5-haiku-latest" />

    <Tag text="llama-text-embed-v2" />

    <Tag text="pinecone-sparse-english-v0" />

    <Tag text="Dense Indexes" />

    <Tag text="Sparse Indexes" />

    <Tag text="Hybrid Search" />
  </ExampleCard>

  <ExampleCard title="Agentic RAG with Claude" text="Build an agentic RAG pipeline that uses tools to retrieve data from web search and Pinecone semantic search, then generates responses using Anthropic's Claude models" link="https://colab.research.google.com/github/pinecone-io/examples/blob/master/learn/generation/agentic-rag/agentic-rag-with-claude.ipynb" arrow>
    <Tag text="claude-3-5-haiku-latest" />

    <Tag text="llama-text-embed-v2" />

    <Tag text="Dense Indexes" />

    <Tag text="Tool use" />
  </ExampleCard>

  <ExampleCard title="RAG with LangChain and OpenAI" text="Learn how RAG can be used with Pinecone to reduce hallucinations, by grounding responses using our own release notes" link="https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/langchain-retrieval-augmentation.ipynb" arrow>
    <Tag text="gpt-5" icon="/images/examples/openai-icon.svg" />

    <Tag text="text-embedding-3-small" icon="/images/examples/openai-icon.svg" />

    <Tag text="OpenAI" />

    <Tag text="LangChain" />
  </ExampleCard>

  <ExampleCard title="RAG with cascading retrieval and OpenAI " text="Investigate research papers by implementing cascading retrieval, then pass results to OpenAI to generate answers" link="https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/gen-qa-openai.ipynb" arrow>
    <Tag text="gpt-4o-mini" icon="/images/examples/openai-icon.svg" />

    <Tag text="llama-text-embed-v2" />

    <Tag text="pinecone-sparse-english-v0" />

    <Tag text="Dense Indexes" />

    <Tag text="Sparse Indexes" />

    <Tag text="Cascading Retrieval" />
  </ExampleCard>

  <ExampleCard title="Retrieval Agents with Pinecone Assistant, LangChain and LangGraph" text="Create a study guide generator using agentic retrieval and the Pinecone Assistant Context API" link="https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/langchain-retrieval-agent.ipynb" arrow>
    <Tag text="gpt-4o-mini" icon="/images/examples/openai-icon.svg" />

    <Tag text="gpt-4.1" icon="/images/examples/openai-icon.svg" />

    <Tag text="LangChain" />

    <Tag text="LangGraph" />

    <Tag text="Agentic Retrieval" />

    <Tag text="Structured Generation" />
  </ExampleCard>
</div>

<h2 className="examples-h2">Miscellaneous</h2>

<div className="card-grid not-prose">
  <ExampleCard title="Import from object storage" text="Import records from Parquet files in Amazon S3 bucket into a serverless index." link="https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/pinecone-import.ipynb" arrow>
    <Tag text="Amazon S3" />
  </ExampleCard>
</div>

---

# Source: https://docs.pinecone.io/integrations/nuclia.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Nuclia

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

[Nuclia](https://nuclia.com/) RAG-as-a-Service automatically indexes files and documents from both internal and external sources, powering diverse company use cases with large language models (LLMs). This comprehensive indexing capability ensures that organizations can leverage unstructured data effectively, transforming it into actionable insights. With Nuclia's modular Retrieval-Augmented Generation (RAG) system, you can deploy solutions tailored to various operational needs across different deployment options, enhancing flexibility and efficiency.

The modular RAG system from Nuclia is designed to fit specific use cases, allowing you to customize your RAG pipeline to meet your unique requirements. Whether it's defining your own retrieval and chunking strategies or choosing from various embedding models, Nuclia's RAG-as-a-Service makes it easy to bring your tailored solutions into production. This customization not only improves the value of your products but also helps you stay competitive by automating tasks and making your data smarter with LLMs, saving hundreds of hours in the process.

When you create a knowledge box at Nuclia, choose to store the index in Pinecone. This is especially useful for large datasets where full text search is not key on the retrieval phase.

<PrimarySecondaryCTA primaryLabel={"Get started"} primaryHref={"https://docs.nuclia.dev/docs/management/third-party-vector-databases/pinecone/"} />

---

# Source: https://docs.pinecone.io/integrations/octoai.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# OctoAI

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Harness value from the latest AI innovations by delievering efficient, reliable, and customizable AI systems for your apps. Run your models or checkpoints on OctoAI's cost-effective API endpoints, or run OctoAI's optimized GenAI stack in your environment.

Choose from the best models that OctoAI has to offer, including GTE Large embedding model, the best foundational open source LLMs such as Mixtral-8x7B from Mistral AI, Llama2 from Meta, and highly capable model fine tunes like Nous Hermes 2 Pro Mistral from Nous Research.

As a fully open source solution, Pinecone Canopy and OctoAI is one of the fastest ways and more affordable ways to get started on your RAG journey. Canopy uses Pinecone vector database for storage and retrieval, which is free to use for up to 100k vectors (that's about 30k pages of text).

<PrimarySecondaryCTA primaryLabel={"Get started"} primaryHref={"https://octo.ai/docs/integrations/pinecone"} />

---

# Source: https://docs.pinecone.io/guides/production/configure-single-sign-on/okta.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Configure SSO with Okta

> Configure SAML SSO with Okta for enterprise.

This page describes how to set up Pinecone with Okta as the single sign-on (SSO) provider. These instructions can be adapted for any provider with SAML 2.0 support.

<Note>SSO is available on Standard and Enterprise plans.</Note>

## Before you begin

This page assumes you have the following:

* Access to your organization's [Pinecone console](https://login.pinecone.io) as an [organization owner](/guides/organizations/understanding-organizations#organization-owners).
* Access to your organization's [Okta Admin console](https://login.okta.com/).

## 1. Start SSO setup in Pinecone

First, start setting up SSO in Pinecone. In this step, you'll capture a couple values necessary for configuring Okta in [Step 2](#2-create-an-app-integration-in-okta).

1. In the Pinecone console, go to [**Settings > Manage**](https://app.pinecone.io/organizations/-/settings/manage).
2. In the **Single Sign-On** section, click **Enable SSO**.
3. In the **Setup SSO** dialog, copy the **Entity ID** and the **Assertion Consumer Service (ACS) URL**. You'll need these values in [Step 2](#2-create-an-app-integration-in-okta).
4. Click **Next**.

Keep this window or browser tab open. You'll come back to it in [Step 4](#4-complete-sso-setup-in-pinecone).

## 2. Create an app integration in Okta

In [Okta](https://login.okta.com/), follow these steps to create and configure a Pinecone app integration:

1. If you're not already on the Okta Admin console, navigate there by clicking the **Admin** button.

2. Navigate to **Applications > Applications**.

3. Click **Create App Integration**.

4. Select **SAML 2.0**.

5. Click **Next**.

6. Enter the **General Settings**:

   * **App name**: `Pinecone`
   * **App logo**: (optional)
   * **App visibility**: Set according to your organization's needs.

7. Click **Next**.

8. For **SAML Settings**, enter values you copied in [Step 1](#1-start-sso-setup-in-pinecone):

   * **Single sign-on URL**: Your **Assertion Consumer Service (ACS) URL**
   * **Audience URI (SP Entity ID)**: Your **Entity ID**
   * **Name ID format**: `EmailAddress`
   * **Application username**: `Okta username`
   * **Update application username on**: `Create and update`

9. In the **Attribute Statements** section, create the following attribute:

   * **Name**: `email`
   * **Value**: `user.email`

10. Click **Next**.

11. Click **Finish**.

## 3. Get the sign on URL and certificate from Okta

Next, in Okta, get the URL and certificate for the Pinecone application you just created. You'll use these in [Step 4](#4-complete-sso-setup-in-pinecone).

1. In the Okta Admin console, navigate to **Applications > Pinecone > Sign On**. If you're continuing from the previous step, you should already be on the right page.
2. In the **SAML 2.0** section, expand **More details**.
3. Copy the **Sign on URL**.
4. Download the **Signing Certificate**.

   <Warning>
     Download the certificate, don't copy it. The downloaded version contains necessary `-----BEGIN CERTIFICATE-----` and `-----END CERTIFICATE-----` lines.
   </Warning>

## 4. Complete SSO setup in Pinecone

In the browser tab or window you kept open in [Step 1](#1-start-sso-setup-in-pinecone), complete the SSO setup in Pinecone:

1. In the **SSO Setup** window, enter the following values:

   * **Login URL**: The URL copied in [Step 3](#3-get-the-sign-on-url-and-certificate-from-okta).
   * **Email domain**: Your company's email domain. To target multiple domains, enter each domain separated by a comma.
   * **Certificate**: The contents of the certificate file you copied in [Step 3](#3-get-the-sign-on-url-and-certificate-from-okta).

     <Warning>
       When pasting the certificate, be sure to include the `-----BEGIN CERTIFICATE-----` and `-----END CERTIFICATE-----` lines.
     </Warning>

2. Choose whether or not to **Enforce SSO for all users**.

   * If enabled, all members of your organization must use SSO to log in to Pinecone.
   * If disabled, members can choose to log in with SSO or with their Pinecone credentials.

3. Click **Next**.

4. Select a **Default role** for all users who log in with SSO. You can change user roles later.

   <Warning>
     When users first log in via SSO, they receive the default SSO role regardless of their previous role. Subsequent SSO logins do not change the role. If the default is **User**, existing owners will lose owner access on their first SSO login.

     To prevent losing access to organization management features:

     * **Sole owner**: Temporarily set the default to **Owner**, log in via SSO to retain owner access, then change the default back to **User**. After changing it back, check your organization's user list to verify no one else logged in via SSO while the default was **Owner**—if they did, adjust their roles accordingly.
     * **Multiple owners**: Keep at least one owner signed in via email while others log in via SSO. That owner can restore roles as needed, then log in via SSO last.

     If all owners lose access, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).
   </Warning>

Okta is now ready to be used for single sign-on. Follow the [Okta docs](https://help.okta.com/en-us/content/topics/users-groups-profiles/usgp-main.htm) to learn how to add users and groups.

---

# Source: https://docs.pinecone.io/integrations/openai.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# OpenAI

> Using OpenAI and Pinecone to combine deep learning capabilities for embedding generation with efficient vector storage and retrieval

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

OpenAI's large language models (LLMs) enhance semantic search or “long-term memory” for LLMs. This combo utilizes LLMs' embedding and completion (or generation) endpoints alongside Pinecone's vector search capabilities for nuanced information retrieval.

By integrating OpenAI's LLMs with Pinecone, you can combine deep learning capabilities for embedding generation with efficient vector storage and retrieval. This approach surpasses traditional keyword-based search, offering contextually-aware, precise results.

<PrimarySecondaryCTA secondaryHref={"#setup-guide"} secondaryLabel={"View setup guide"} />

## Setup guide

[View source](https://github.com/pinecone-io/examples/blob/master/integrations/openai/)

[Open in Colab](https://colab.research.google.com/github/pinecone-io/examples/blob/master/integrations/openai/semantic_search_openai.ipynb)

This guide covers the integration of OpenAI's Large Language Models (LLMs) with Pinecone (referred to as the **OP stack**), enhancing semantic search or 'long-term memory' for LLMs. This combo utilizes LLMs' embedding and completion (or generation) endpoints alongside Pinecone's vector search capabilities for nuanced information retrieval.

LLMs like OpenAI's `text-embedding-ada-002` generate vector embeddings, i.e., numerical representations of text semantics. These embeddings facilitate semantic-based rather than literal textual matches. Additionally, LLMs like `gpt-4` or `gpt-3.5-turbo` can predict text completions based on information provided from these contexts.

Pinecone is a vector database designed for storing and querying high-dimensional vectors. It provides fast, efficient semantic search over these vector embeddings.

By integrating OpenAI's LLMs with Pinecone, we combine deep learning capabilities for embedding generation with efficient vector storage and retrieval. This approach surpasses traditional keyword-based search, offering contextually-aware, precise results.

There are many ways of integrating these two tools and we have several guides focusing on specific use-cases. If you already know what you'd like to do you can jump to these specific materials:

* [ChatGPT Plugins Walkthrough](https://youtu.be/hpePPqKxNq8)
* [Ask Lex ChatGPT Plugin](https://github.com/pinecone-io/examples/tree/master/learn/generation/openai/chatgpt/plugins/ask-lex)
* [Generative Question-Answering](https://github.com/pinecone-io/examples/blob/master/docs/gen-qa-openai.ipynb)
* [Retrieval Augmentation using LangChain](https://github.com/pinecone-io/examples/blob/master/learn/generation/langchain/handbook/05-langchain-retrieval-augmentation.ipynb)

### Introduction to Embeddings

At the core of the OP stack we have embeddings which are supported via the [OpenAI Embedding API](https://beta.openai.com/docs/guides/embeddings). We index those embeddings in the [Pinecone vector database](https://www.pinecone.io) for fast and scalable retrieval augmentation of our LLMs or other information retrieval use-cases.

*This example demonstrates the core OP stack. It is the simplest workflow and is present in each of the other workflows, but is not the only way to use the stack. Please see the links above for more advanced usage.*

The OP stack is built for semantic search, question-answering, threat-detection, and other applications that rely on language models and a large corpus of text data.

The basic workflow looks like this:

* Embed and index
  * Use the OpenAI Embedding API to generate vector embeddings of your documents (or any text data).
  * Upload those vector embeddings into Pinecone, which can store and index millions/billions of these vector embeddings, and search through them at ultra-low latencies.
* Search
  * Pass your query text or document through the OpenAI Embedding API again.
  * Take the resulting vector embedding and send it as a [query](/guides/search/search-overview) to Pinecone.
  * Get back semantically similar documents, even if they don't share any keywords with the query.

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6a3ea5a-pinecone-openai-overview.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=7bb594ae2f7a2c13846b2993cd1b3899" alt="Basic workflow of OpenAI and Pinecone" data-og-width="1902" width="1902" data-og-height="864" height="864" data-path="images/6a3ea5a-pinecone-openai-overview.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6a3ea5a-pinecone-openai-overview.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=dc5a682918633d319dbeeecb46e50d85 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6a3ea5a-pinecone-openai-overview.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=16c2456e6c25afe985222aaf7c7945a4 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6a3ea5a-pinecone-openai-overview.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=405832c780b51b5f471c934d4bf12a58 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6a3ea5a-pinecone-openai-overview.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=9c0f212e45a2a47423a2ade78d2d533e 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6a3ea5a-pinecone-openai-overview.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=bbd84486e6e4bbb88effa28075301d9c 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6a3ea5a-pinecone-openai-overview.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=86a650955b6d155da093e7c20ceccb8d 2500w" />

Let's get started...

### Environment Setup

We start by installing the OpenAI and Pinecone clients, we will also need HuggingFace *Datasets* for downloading the TREC dataset that we will use in this guide.

```Bash Bash theme={null}
!pip install -qU \
    pinecone[grpc]==7.3.0 \
    openai==1.93.0 \
    datasets==3.6.0
```

#### Creating Embeddings

To create embeddings we must first initialize our connection to OpenAI Embeddings, we sign up for an API key at [OpenAI](https://beta.openai.com/signup).

```Python Python theme={null}
from openai import OpenAI

client = OpenAI(
    api_key="OPENAI_API_KEY"
)  # get API key from platform.openai.com
```

We can now create embeddings with the OpenAI v3 small embedding model like so:

```Python Python theme={null}
MODEL = "text-embedding-3-small"

res = client.embeddings.create(
    input=[
        "Sample document text goes here",
        "there will be several phrases in each batch"
    ], model=MODEL
)
```

In `res` we should find a JSON-like object containing two 1536-dimensional embeddings, these are the vector representations of the two inputs provided above. To access the embeddings directly we can write:

```Python Python theme={null}
# we can extract embeddings to a list
embeds = [record.embedding for record in res.data]
len(embeds)
```

We will use this logic when creating our embeddings for the **T**ext **RE**trieval **C**onference (TREC) question classification dataset later.

#### Initializing a Pinecone Index

Next, we initialize an index to store the vector embeddings. For this we need a Pinecone API key, [sign up for one here](https://app.pinecone.io).

```Python Python theme={null}
import time
from pinecone.grpc import PineconeGRPC as Pinecone
from pinecone import ServerlessSpec

pc = Pinecone(api_key="...")

spec = ServerlessSpec(cloud="aws", region="us-east-1")

index_name = 'semantic-search-openai'

# check if index already exists (it shouldn't if this is your first run)
if index_name not in pc.list_indexes().names():
    # if does not exist, create index
    pc.create_index(
        index_name,
        dimension=len(embeds[0]),  # dimensionality of text-embed-3-small
        metric='dotproduct',
        spec=spec
    )

# connect to index
index = pc.Index(index_name)
time.sleep(1)
# view index stats
index.describe_index_stats()
```

#### Populating the Index

With both OpenAI and Pinecone connections initialized, we can move onto populating the index. For this, we need the TREC dataset.

```Python Python theme={null}
from datasets import load_dataset

# load the first 1K rows of the TREC dataset
trec = load_dataset('trec', split='train[:1000]')
```

Then we create a vector embedding for each question using OpenAI (as demonstrated earlier), and `upsert` the ID, vector embedding, and original text for each phrase to Pinecone.

<Warning>
  High-cardinality metadata values (like the unique text values we use here)\
  can reduce the number of vectors that fit on a single pod. See\
  [Known limitations](/reference/api/known-limitations) for more.
</Warning>

```Python Python theme={null}
from tqdm.auto import tqdm

count = 0  # we'll use the count to create unique IDs
batch_size = 32  # process everything in batches of 32
for i in tqdm(range(0, len(trec['text']), batch_size)):
    # set end position of batch
    i_end = min(i+batch_size, len(trec['text']))
    # get batch of lines and IDs
    lines_batch = trec['text'][i: i+batch_size]
    ids_batch = [str(n) for n in range(i, i_end)]
    # create embeddings
    res = client.embeddings.create(input=lines_batch, model=MODEL)
    embeds = [record.embedding for record in res.data]
    # prep metadata and upsert batch
    meta = [{'text': line} for line in lines_batch]
    to_upsert = zip(ids_batch, embeds, meta)
    # upsert to Pinecone
    index.upsert(vectors=list(to_upsert))
```

#### Querying

With our data indexed, we're now ready to move onto performing searches. This follows a similar process to indexing. We start with a text `query`, that we would like to use to find similar sentences. As before we encode this with OpenAI's text similarity Babbage model to create a *query vector* `xq`. We then use `xq` to query the Pinecone index.

```Python Python theme={null}
query = "What caused the 1929 Great Depression?"

xq = client.embeddings.create(input=query, model=MODEL).data[0].embedding
```

Now we query.

```Python Python theme={null}
res = index.query(vector=xq, top_k=5, include_metadata=True)
```

The response from Pinecone includes our original text in the `metadata` field, let's print out the `top_k` most similar questions and their respective similarity scores.

```Python Python theme={null}
for match in res['matches']:
    print(f"{match['score']:.2f}: {match['metadata']['text']}")
```

```[Out]:  theme={null}
0.75: Why did the world enter a global depression in 1929 ?
0.60: When was `` the Great Depression '' ?
0.37: What crop failure caused the Irish Famine ?
0.32: What were popular songs and types of songs in the 1920s ?
0.32: When did World War I start ?
```

Looks good, let's make it harder and replace *"depression"* with the incorrect term *"recession"*.

```Python Python theme={null}
query = "What was the cause of the major recession in the early 20th century?"

# create the query embedding
xq = client.embeddings.create(input=query, model=MODEL).data[0].embedding

# query, returning the top 5 most similar results
res = index.query(vector=xq, top_k=5, include_metadata=True)

for match in res['matches']:
    print(f"{match['score']:.2f}: {match['metadata']['text']}")
```

```[Out]:  theme={null}
0.63: Why did the world enter a global depression in 1929 ?
0.55: When was `` the Great Depression '' ?
0.34: What were popular songs and types of songs in the 1920s ?
0.33: What crop failure caused the Irish Famine ?
0.29: What is considered the costliest disaster the insurance industry has ever faced ?
```

Let's perform one final search using the definition of depression rather than the word or related words.

```Python Python theme={null}
query = "Why was there a long-term economic downturn in the early 20th century?"

# create the query embedding
xq = client.embeddings.create(input=query, model=MODEL).data[0].embedding

# query, returning the top 5 most similar results
res = index.query(vector=xq, top_k=5, include_metadata=True)

for match in res['matches']:
    print(f"{match['score']:.2f}: {match['metadata']['text']}")
```

```[Out]:  theme={null}
0.62: Why did the world enter a global depression in 1929 ?
0.54: When was `` the Great Depression '' ?
0.34: What were popular songs and types of songs in the 1920s ?
0.33: What crop failure caused the Irish Famine ?
0.32: What do economists do ?
```

It's clear from this example that the semantic search pipeline is clearly able to identify the meaning between each of our queries. Using these embeddings with Pinecone allows us to return the most semantically similar questions from the already indexed TREC dataset.

Once we're finished with the index we delete it to save resources.

```Python Python theme={null}
pc.delete_index(name=index_name)
```

## Related articles

* [Generative Question-Answering with Long-Term Memory](https://www.pinecone.io/learn/openai-gen-qa)
* [OpenAI's Text Embeddings v3](https://www.pinecone.io/learn/openai-embeddings-v3/)

---

# Source: https://docs.pinecone.io/guides/assistant/admin/organizations-overview.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Organizations overview

> Understand organization structure, projects, and billing.

A Pinecone organization is a set of [projects](/guides/assistant/admin/projects-overview) that use the same billing. Organizations allow one or more users to control billing and project permissions for all of the projects belonging to the organization. Each project belongs to an organization.

<Note>
  While an email address can be associated with multiple organizations, it cannot be used to create more than one organization. For information about managing organization members, see [Manage organization members](/guides/assistant/admin/manage-organization-members).
</Note>

## Projects in an organization

Each organization contains one or more projects that share the same organization owners and billing settings. Each project belongs to exactly one organization. If you need to move a project from one organization to another, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

## Billing settings

All of the projects in an organization share the same billing method and settings. The billing settings for the organization are controlled by the organization owners.

Organization owners can update the billing contact information, update the payment method, and view and download invoices using the [Pinecone console](https://app.pinecone.io/organizations/-/settings/billing).

## Organization roles

Organization owners can manage access to their organizations and projects by assigning roles to organization members and service accounts. The role determines the entity's permissions within Pinecone. The organization roles are as follows:

* **Organization owner**: Organization owners have global permissions across the organization. This includes managing billing details, organization members, and all projects. Organization owners are automatically [project owners](/guides/assistant/admin/projects-overview#project-roles) and, therefore, have all project owner permissions as well.

* **Organization user**: Organization users have restricted organization-level permissions. When inviting organization users, you also choose the projects they belong to and the project role they should have. Organization users are automatically [project owners](/guides/assistant/admin/projects-overview#project-roles) and, therefore, have all project owner permissions as well.

* **Billing admin**: Billing admins have permissions to view and update billing details, but they cannot manage organization members. Billing admins cannot manage projects unless they are also [project owners](/guides/assistant/admin/projects-overview#project-roles).

The following table summarizes the permissions for each organization role:

| Permission                           | Org Owner | Org User | Billing Admin |
| ------------------------------------ | --------- | -------- | ------------- |
| View account details                 | ✓         | ✓        | ✓             |
| Update organization name             | ✓         |          |               |
| Delete the organization              | ✓         |          |               |
| View billing details                 | ✓         |          | ✓             |
| Update billing details               | ✓         |          | ✓             |
| View usage details                   | ✓         |          | ✓             |
| View support plans                   | ✓         |          | ✓             |
| Invite members to the organization   | ✓         |          |               |
| Delete pending member invites        | ✓         |          |               |
| Remove members from the organization | ✓         |          |               |
| Update organization member roles     | ✓         |          |               |
| Create projects                      | ✓         | ✓        |               |

## Organization single sign-on (SSO)

SSO allows organizations to manage their teams' access to Pinecone through their identity management solution. Once your integration is configured, you can specify a default role for teammates when they sign up.

For more information, see [Configure single sign-on](/guides/assistant/admin/configure-sso-with-okta).

<Note>SSO is available on Standard and Enterprise plans.</Note>

## Service accounts

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

[Service accounts](/guides/assistant/admin/manage-organization-service-accounts) enable programmatic access to Pinecone's Admin API, which can be used to create and manage projects and API keys.

Use service accounts to automate infrastructure management and integrate Pinecone into your deployment workflows, rather than through manual actions in the Pinecone console. Service accounts use the [organization roles](/guides/assistant/admin/organizations-overview#organization-roles) and [project role](/guides/assistant/admin/projects-overview#project-roles) for permissioning, and provide a secure and auditable way to handle programmatic access.

## See also

* [Manage organization members](/guides/assistant/admin/manage-organization-members)
* [Manage project members](/guides/assistant/admin/manage-project-members)
* [Project overview](/guides/assistant/admin/projects-overview)

---

# Source: https://docs.pinecone.io/reference/sdks/rust/overview.md

# Source: https://docs.pinecone.io/reference/sdks/python/overview.md

# Source: https://docs.pinecone.io/reference/sdks/node/overview.md

# Source: https://docs.pinecone.io/reference/sdks/java/overview.md

# Source: https://docs.pinecone.io/reference/sdks/go/overview.md

# Source: https://docs.pinecone.io/reference/sdks/dotnet/overview.md

# Source: https://docs.pinecone.io/reference/cli/overview.md

# Source: https://docs.pinecone.io/models/overview.md

# Source: https://docs.pinecone.io/integrations/overview.md

# Source: https://docs.pinecone.io/guides/get-started/overview.md

# Source: https://docs.pinecone.io/guides/assistant/overview.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Pinecone Assistant

> Pinecone Assistant is a service that allow you to build production-grade chat and agent-based applications quickly.

<CardGroup cols={2}>
  <Card title="Assistant quickstart" icon="comments" href="/guides/assistant/quickstart/sdk-quickstart">
    Create an AI assistant that answers complex questions about your proprietary data
  </Card>

  <Card title="Database quickstart" icon="database" href="/guides/get-started/quickstart">
    Set up a fully managed vector database for high-performance semantic search
  </Card>
</CardGroup>

## Use cases

Pinecone Assistant is useful for a variety of tasks, especially for the following:

* Prototyping and deploying an AI assistant quickly.
* Providing context-aware answers about your proprietary data without training an LLM.
* Retrieving answers grounded in your data, with references.

## SDK support

You can use the [Assistant API](/reference/api/latest/assistant/) directly, through the [Pinecone Python SDK](/reference/sdks/python/overview), or through the [Pinecone Node.js SDK](/reference/sdks/node/overview).

## Workflow

You can use the Pinecone Assistant through the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/assistant) or [Pinecone API](/reference/api/latest/assistant/list_assistants).

<Tabs>
  <Tab title="Overview">
    The following steps outline the general Pinecone Assistant workflow:

    <Steps>
      <Step title="Create an assistant">
        [Create an assistant](/guides/assistant/create-assistant) to answer questions about your documents.
      </Step>

      <Step title="Upload documents">
        [Upload documents](/guides/assistant/upload-files) to your assistant. Your assistant manages chunking, embedding, and storage for you.
      </Step>

      <Step title="Chat with an assistant">
        [Chat with your assistant](/guides/assistant/chat-with-assistant) and receive responses as a JSON object or as a text stream. For each chat, your assistant queries a large language model (LLM) with context from your documents to ensure the LLM provides grounded responses.
      </Step>

      <Step title="Evaluate answers">
        [Evaluate the assistant's responses](/guides/assistant/evaluation-overview) for correctness and completeness.
      </Step>

      <Step title="Optimize performance">
        [Use custom instructions](https://www.pinecone.io/learn/assistant-api-deep-dive/#Custom-Instructions) to tailor your assistant's behavior and responses to specific use cases or requirements. [Filter by metadata associated with files](https://www.pinecone.io/learn/assistant-api-deep-dive/#Using-Metadata) to reduce latency and improve the accuracy of responses.
      </Step>

      <Step title="Retrieve context snippets">
        [Retrieve context snippets](/guides/assistant/retrieve-context-snippets) to understand what relevant data snippets Pinecone Assistant is using to generate responses. You can use the retrieved snippets with your own LLM, RAG application, or agentic workflow.
      </Step>
    </Steps>

    <Note>
      For information on how the Pinecone Assistant works, see [Assistant architecture](/reference/architecture/assistant-architecture).
    </Note>
  </Tab>

  <Tab title="Code sample">
    The following code samples outline the Pinecone Assistant workflow using either the [Pinecone Python SDK](/reference/sdks/python/overview) and [Pinecone Assistant plugin](/reference/sdks/python/overview#install-the-pinecone-assistant-python-plugin) or the [Pinecone Node.js SDK](/reference/sdks/node/overview).

    <CodeGroup>
      ```python Python theme={null}
      # pip install pinecone
      # pip install pinecone-plugin-assistant

      from pinecone import Pinecone
      import requests
      from pinecone_plugins.assistant.models.chat import Message

      pc = Pinecone(api_key="YOUR_API_KEY")

      # Create an assistant.
      assistant = pc.assistant.create_assistant(
          assistant_name="example-assistant", 
          instructions="Use American English for spelling and grammar.", # Description or directive for the assistant to apply to all responses.
          region="us", # Region to deploy assistant. Options: "us" (default) or "eu".    
          timeout=30 # Maximum seconds to wait for assistant status to become "Ready" before timing out.
      )

      # Upload a file to your assistant.
      response = assistant.upload_file(
          file_path="/Users/jdoe/Downloads/Netflix-10-K-01262024.pdf",
          metadata={"company": "netflix", "document_type": "form 10k"},
          timeout=None
      )

      # Set up for evaluation later.
      payload = {
          "question": "Who is the CFO of Netflix?", # Question to ask the assistant.
          "ground_truth_answer": "Spencer Neumann" # Expected answer to evaluate the assistant's response.
      }

      # Chat with the assistant.
      msg = Message(role="user", content=payload["question"])
      resp = assistant.chat(messages=[msg], model="gpt-4o")
      print(resp)

      # {
      #    'id': '0000000000000000163008a05b317b7b', 
      #    'model': 'gpt-4o-2024-05-13', 
      #    'usage': {
      #        'prompt_tokens': 9259, 
      #        'completion_tokens': 30, 
      #        'total_tokens': 9289
      #        }, 
      #        'message': {
      #            'content': 'The Chief Financial Officer (CFO) of Netflix is Spencer Neumann.', 
      #            'role': '"assistant"'
      #            }, 
      #            'finish_reason': 'stop', 
      #            'citations': [
      #                {
      #                    'position': 63, 
      #                    'references': [
      #                        {
      #                            'pages': [78, 72, 79], 
      #                            'file': {
      #                                'name': 'Netflix-10-K-01262024.pdf', 
      #                                'id': '76a11dd1...', 
      #                                'metadata': {
      #                                    'company': 'netflix', 
      #                                    'document_type': 'form 10k'
      #                                    }, 
      #                                    'created_on': '2024-12-06T01:29:07.369208590Z', 
      #                                    'updated_on': '2024-12-06T01:29:50.923493799Z', 
      #                                    'status': 'Available', 
      #                                    'percent_done': 1.0, 
      #                                    'signed_url': 'https://storage.googleapis.com/...', 
      #                                    'error_message': None,
      #                                    'size': 1073470.0
      #                                }
      #                            }
      #                        ]
      #                    }
      #                ]
      #            }

      # Evaluate the assistant's response.
      payload["answer"] = resp.message.content

      headers = {
          "Api-Key": "YOUR_API_KEY",
          "Content-Type": "application/json"
      }

      url = "https://prod-1-data.ke.pinecone.io/assistant/evaluation/metrics/alignment"

      response = requests.request("POST", url, json=payload, headers=headers)

      print(response.text)

      # {
      #    "metrics":
      #    {
      #        "correctness":1.0,
      #        "completeness":1.0,
      #        "alignment":1.0
      #    },
      #    "reasoning":
      #    {
      #        "evaluated_facts":
      #        [
      #            {
      #                "fact":
      #                {
      #                    "content":"Spencer Neumann is the CFO of Netflix."
      #                    },
      #                    "entailment":"entailed"
      #                }
      #            ]
      #        },
      #        "usage":
      #        {
      #            "prompt_tokens":1221,
      #            "completion_tokens":24,
      #            "total_tokens":1245
      #            }
      #        }
      ```

      ```javascript JavaScript theme={null}
      import { Pinecone } from "@pinecone-database/pinecone";

      function sleep(ms) {
        return new Promise((resolve) => setTimeout(resolve, ms));
      }

      async function testPinecone() {
        try {
          console.log("Initializing Pinecone client...");

          const pc = new Pinecone({
            apiKey: "YOUR_API_KEY",
          });

          console.log("Pinecone client initialized successfully.");

          const assistantName = "test-assistant";

          // Create a new assistant.
          console.log(`Creating new assistant: ${assistantName}...`);
          await pc.createAssistant({
            name: assistantName,
            region: "us",
            metadata: { 'test-key': 'test-value' },
          });

          // Validate Assistant was created through describe.
          const asstDesc = await pc.describeAssistant(assistantName);
          console.log(`Described Assistant: ${JSON.stringify(asstDesc)}`);

          // Delay to ensure the Assistant is ready.
          await sleep(4000);

          // Upload file
          const assistant = pc.Assistant(assistantName);
          await assistant.uploadFile({
            path: '/Users/jdoe/Downloads/Netflix-10-K-01262024.pdf',
            metadata: { 'test-key': 'test-value' },
          });
          console.log("File uploaded. Processsing...");

          // Delay to ensure file is available.
          await sleep(45000);

          // Chat
          const chatResp = await assistant.chat({
            messages: [{ role: 'user', content: 'Who is the CFO of Netflix?' }]
          });
          console.log(chatResp);
          
        // Error handling
        } catch (error) {
          console.error("Error:", error);
        }
      }

      // Run the sample code
      testAssistant();
      ```
    </CodeGroup>
  </Tab>
</Tabs>

## Learn more

<CardGroup cols={3}>
  <Card title="API Reference" icon="code-simple" href="/reference">
    Comprehensive details about the Pinecone APIs, SDKs, utilities, and architecture.
  </Card>

  <Card title="Blog" icon="blog" href="https://www.pinecone.io/learn/assistant-api-deep-dive/">
    Four features of the Assistant API you aren't using - but should
  </Card>

  <Card title="Releases" icon="party-horn" href="/release-notes">
    News about features and changes in Pinecone and related tools.
  </Card>
</CardGroup>

---

# Source: https://docs.pinecone.io/troubleshooting/parallel-queries.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Parallel queries

There are many approaches to perform parallel queries in your application, from using the Python SDK to making REST calls. Below is one example of an approach using multi-threaded, asynchronous requests in Python. For guidance on using `asyncio` for single-threaded, asynchronous requests in Python, see [Async requests](/reference/sdks/python/overview#async-requests).

This example assumes the following:

* You have a 1536-dimension serverless index called `docs-example`.
* You have the [Pinecone Python SDK](/reference/sdks/python/overview) and [`concurrent.futures`](https://docs.python.org/3/library/concurrent.futures.html#module-concurrent.futures) and [`numpy`](https://numpy.org/) packages installed.

```python  theme={null}
import os
from pinecone import Pinecone
from concurrent.futures import ThreadPoolExecutor

# Get the API key from the environment variable and initialize Pinecone
api_key = os.environ.get("PINECONE_API_KEY")
pc = Pinecone(api_key=api_key)

# Define the index name
index_name = "docs-example"

# Define the index
index = pc.Index(index_name)

# Define the function to run parallel queries
def run_parallel_queries(vectors):
    """
    Run a list of vectors in parallel using ThreadPoolExecutor.
    
    Parameters:
    vectors (list): A list of vectors.
    
    Returns:
    list: A list of query results.
    """
    
    # Define the maximum number of concurrent queries
    MAX_CONCURRENT_QUERIES = 4

    def run_query(vector):
        """
        Run a single query.
        """
        return index.query(
            namespace="",
            vector=vector,
            top_k=3,
            include_values=True
        )
    
    # Run the queries in parallel
    with ThreadPoolExecutor(max_workers=MAX_CONCURRENT_QUERIES) as executor:
        """
        Run the queries in parallel.
        """
        results = list(executor.map(run_query, vectors))
    
    return results

def test_parallel_queries():
    """
    Test the run_parallel_queries function with 20 random vectors.
    """
    import numpy as np

    # Generate 20 random vectors of size 1536 and convert them to lists
    vectors = [np.random.rand(1536).tolist() for _ in range(20)]

    # Define the batch size
    QUERY_BATCH_SIZE = 20
    
    # Run the parallel queries
    results = run_parallel_queries(vectors)

    # Print the results
    for i, result in enumerate(results):
        print(f"Query {i+1} results: {result}")

if __name__ == "__main__":
    test_parallel_queries()
```

---

# Source: https://docs.pinecone.io/troubleshooting/pinecone-attribute-errors-with-langchain.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# PineconeAttribute errors with LangChain

## Problem

When using an outdated version of LangChain, you may encounter errors like the following:

```console  theme={null}
Pinecone has no attribute 'from_texts'
```

```console  theme={null}
Pinecone has no attribute `from_documents'
```

## Solution

Previously, the Python classes for both LangChain and Pinecone had objects named `Pinecone`, but this is no longer an issue in the latest LangChain version. To resolve these errors, upgrade LangChain to >=0.0.3:

```shell  theme={null}
pip install --upgrade langchain-pinecone
```

Depending on which version of LangChain you are upgrading from, you may need to update your code. You can find more information about using LangChain with Pinecone in our [documentation](/integrations/langchain#4-initialize-a-langchain-vector-store).

---

# Source: https://docs.pinecone.io/reference/pinecone-sdks.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Introduction

## Pinecone SDKs

Official Pinecone SDKs provide convenient access to the [Pinecone APIs](/reference/api/introduction).

<CardGroup cols={3}>
  <Card title="Python SDK" icon="python" href="/reference/sdks/python/overview" />

  <Card title="Node.js SDK" icon="node-js" href="/reference/sdks/node/overview" />

  <Card title="Java SDK" icon="java" href="/reference/sdks/java/overview" />

  <Card title="Go SDK" icon="golang" href="/reference/sdks/go/overview" />

  <Card title=".NET SDK" icon="microsoft" href="/reference/sdks/dotnet/overview" />

  <Card title="Rust SDK" icon="rust" href="/reference/sdks/rust/overview" />
</CardGroup>

## SDK versions

SDK versions are pinned to specific [API versions](/reference/api/versioning). When a new API version is released, a new version of the SDK is also released.

The mappings between API versions and SDK versions are as follows:

|                                               | `2025-04` | `2025-01` | `2024-10` | `2024-07`     | `2024-04` |
| --------------------------------------------- | :-------- | :-------- | :-------- | :------------ | :-------- |
| [Python SDK](/reference/sdks/python/overview) | v7.x      | v6.x      | v5.3.x    | v5.0.x-v5.2.x | v4.x      |
| [Node.js SDK](/reference/sdks/node/overview)  | v6.x      | v5.x      | v4.x      | v3.x          | v2.x      |
| [Java SDK](/reference/sdks/java/overview)     | v5.x      | v4.x      | v3.x      | v2.x          | v1.x      |
| [Go SDK](/reference/sdks/go/overview)         | v4.x      | v3.x      | v2.x      | v1.x          | v0.x      |
| [.NET SDK](/reference/sdks/dotnet/overview)   | v4.x      | v3.x      | v2.x      | v1.x          | v0.x      |

When a new stable API version is released, you should upgrade your SDK to the latest version to ensure compatibility with the latest API changes.

<Note>
  SDKs that target API version `2025-10` will be available soon.
</Note>

## Limitations

While Pinecone tracks read unit usage with decimal precision, the Pinecone API and SDKs round these values up to the nearest whole number in query, fetch, and list responses. For example, if a query uses 0.45 read units, the API and SDKs will report it as 1 read unit.

For precise read unit reporting, see [index-level metrics](/guides/production/monitoring) or the organization-wide [Usage dashboard](/guides/manage-cost/monitor-usage-and-costs#monitor-organization-level-usage-and-costs).

## Community SDKs

Find community-contributed SDKs for Pinecone. These libraries are not supported by Pinecone.

* [Ruby SDK](https://github.com/ScotterC/pinecone) (contributed by [ScotterC](https://github.com/ScotterC))

* [Scala SDK](https://github.com/cequence-io/pinecone-scala) (contributed by [cequence-io](https://github.com/cequence-io))

* [PHP SDK](https://github.com/probots-io/pinecone-php) (contributed by [protobots-io](https://github.com/probots-io))

---

# Source: https://docs.pinecone.io/reference/tools/pinecone-spark-connector.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Spark-Pinecone connector

Use the [`spark-pinecone` connector](https://github.com/pinecone-io/spark-pinecone/) to efficiently create, ingest, and update [vector embeddings](https://www.pinecone.io/learn/vector-embeddings/) at scale with [Databricks and Pinecone](/integrations/databricks).

## Install the Spark-Pinecone connector

<Tabs>
  <Tab title="Databricks platform">
    1. [Install the Spark-Pinecone connector as a library](https://docs.databricks.com/en/libraries/cluster-libraries.html#install-a-library-on-a-cluster).
    2. Configure the library as follows:
       1. Select **File path/S3** as the **Library Source**.

       2. Enter the S3 URI for the Pinecone assembly JAR file:

          ```
          s3://pinecone-jars/1.1.0/spark-pinecone-uberjar.jar  
          ```

          <Note>
            Databricks platform users must use the Pinecone assembly jar listed above to ensure that the proper dependecies are installed.
          </Note>

       3. Click **Install**.
  </Tab>

  <Tab title="Databricks on AWS">
    1. [Install the Spark-Pinecone connector as a library](https://docs.databricks.com/en/libraries/cluster-libraries.html#install-a-library-on-a-cluster).
    2. Configure the library as follows:
       1. Select **File path/S3** as the **Library Source**.

       2. Enter the S3 URI for the Pinecone assembly JAR file:

          ```
          s3://pinecone-jars/1.1.0/spark-pinecone-uberjar.jar  
          ```

       3. Click **Install**.
  </Tab>

  <Tab title="Databricks on GCP / Azure">
    1. [Install the Spark-Pinecone connector as a library](https://docs.databricks.com/en/libraries/cluster-libraries.html#install-a-library-on-a-cluster).
    2. Configure the library as follows:
       1. [Download the Pinecone assembly JAR file](https://repo1.maven.org/maven2/io/pinecone/spark-pinecone_2.12/1.1.0/).
       2. Select **Workspace** as the **Library Source**.
       3. Upload the JAR file.
       4. Click **Install**.
  </Tab>
</Tabs>

## Batch upsert

To batch upsert embeddings to Pinecone:

<CodeGroup>
  ```python Python theme={null}
  from pyspark import SparkConf
  from pyspark.sql import SparkSession
  from pyspark.sql.types import StructType, StructField, ArrayType, FloatType, StringType, LongType

  # Your API key and index name
  api_key = "PINECONE_API_KEY"
  index_name = "PINECONE_INDEX_NAME"
  source_tag = "PINECONE_SOURCE_TAG"

  COMMON_SCHEMA = StructType([
      StructField("id", StringType(), False),
      StructField("namespace", StringType(), True),
      StructField("values", ArrayType(FloatType(), False), False),
      StructField("metadata", StringType(), True),
      StructField("sparse_values", StructType([
          StructField("indices", ArrayType(LongType(), False), False),
          StructField("values", ArrayType(FloatType(), False), False)
      ]), True)
  ])

  # Initialize Spark
  spark = SparkSession.builder.getOrCreate()

  # Read the file and apply the schema
  df = spark.read \
      .option("multiLine", value = True) \
      .option("mode", "PERMISSIVE") \
      .schema(COMMON_SCHEMA) \
      .json("src/test/resources/sample.jsonl")

  # Show if the read was successful
  df.show()

  # Write the dataFrame to Pinecone in batches 
  df.write \
      .option("pinecone.apiKey", api_key) \
      .option("pinecone.indexName", index_name) \
      .option("pinecone.sourceTag", source_tag) \
      .format("io.pinecone.spark.pinecone.Pinecone") \
      .mode("append") \
      .save()
  ```

  ```scala Scala theme={null}
  import io.pinecone.spark.pinecone.{COMMON_SCHEMA, PineconeOptions}
  import org.apache.spark.SparkConf
  import org.apache.spark.sql.{SaveMode, SparkSession}

  object MainApp extends App {
    // Your API key and index name
    val apiKey = "PINECONE_API_KEY"
    val indexName = "PINECONE_INDEX_NAME"
    val sourceTag = "PINECONE_SOURCE_TAG"

    // Configure Spark to run locally with all available cores
    val conf = new SparkConf()
      .setMaster("local[*]")

    // Create a Spark session with the defined configuration
    val spark = SparkSession.builder().config(conf).getOrCreate()

    // Read the JSON file into a DataFrame, applying the COMMON_SCHEMA
    val df = spark.read
      .option("multiLine", value = true)
      .option("mode", "PERMISSIVE")
      .schema(COMMON_SCHEMA)
      .json("src/test/resources/sample.jsonl") // path to sample.jsonl

    // Define Pinecone options as a Map
    val pineconeOptions = Map(
      PineconeOptions.PINECONE_API_KEY_CONF -> apiKey,
      PineconeOptions.PINECONE_INDEX_NAME_CONF -> indexName,
      PineconeOptions.PINECONE_SOURCE_TAG_CONF -> sourceTag
    )

    // Show if the read was successful
    df.show(df.count().toInt)
    
    // Write the DataFrame to Pinecone using the defined options in batches
    df.write
      .options(pineconeOptions)
      .format("io.pinecone.spark.pinecone.Pinecone")
      .mode(SaveMode.Append)
      .save()
  }
  ```
</CodeGroup>

<Tip>
  For a guide on how to set up batch upserts, refer to the [Databricks integration page](/integrations/databricks#setup-guide).
</Tip>

## Stream upsert

To stream upsert embeddings to Pinecone:

<CodeGroup>
  ```python Python theme={null}
  from pyspark.sql import SparkSession
  from pyspark.sql.types import StructType, StructField, ArrayType, FloatType, StringType, LongType
  import os

  # Your API key and index name
  api_key = "PINECONE_API_KEY"
  index_name = "PINECONE_INDEX_NAME"
  source_tag = "PINECONE_SOURCE_TAG"

  COMMON_SCHEMA = StructType([
      StructField("id", StringType(), False),
      StructField("namespace", StringType(), True),
      StructField("values", ArrayType(FloatType(), False), False),
      StructField("metadata", StringType(), True),
      StructField("sparse_values", StructType([
          StructField("indices", ArrayType(LongType(), False), False),
          StructField("values", ArrayType(FloatType(), False), False)
      ]), True)
  ])

  # Initialize Spark session
  spark = SparkSession.builder \
      .appName("StreamUpsertExample") \
      .config("spark.sql.shuffle.partitions", 3) \
      .master("local") \
      .getOrCreate()

  # Read the stream of JSON files, applying the schema from the input directory
  lines = spark.readStream \
      .option("multiLine", True) \
      .option("mode", "PERMISSIVE") \
      .schema(COMMON_SCHEMA) \
      .json("path/to/input/directory/")

  # Write the stream to Pinecone using the defined options
  upsert = lines.writeStream \
      .format("io.pinecone.spark.pinecone.Pinecone") \
      .option("pinecone.apiKey", api_key) \
      .option("pinecone.indexName", index_name) \
      .option("pinecone.sourceTag", source_tag) \
      .option("checkpointLocation", "path/to/checkpoint/dir") \
      .outputMode("append") \
      .start()

  upsert.awaitTermination()
  ```

  ```scala Scala theme={null}
  import io.pinecone.spark.pinecone.{COMMON_SCHEMA, PineconeOptions}
  import org.apache.spark.SparkConf
  import org.apache.spark.sql.{SaveMode, SparkSession}

  object MainApp extends App {
    // Your API key and index name
    val apiKey = "PINECONE_API_KEY"
    val indexName = "PINECONE_INDEX_NAME"

    // Create a Spark session
    val spark = SparkSession.builder()
      .appName("StreamUpsertExample")
      .config("spark.sql.shuffle.partitions", 3)
      .master("local")
      .getOrCreate()

    // Read the JSON files into a DataFrame, applying the COMMON_SCHEMA from input directory
    val lines = spark.readStream
      .option("multiLine", value = true)
      .option("mode", "PERMISSIVE")
      .schema(COMMON_SCHEMA)
      .json("path/to/input/directory/")

    // Define Pinecone options as a Map
    val pineconeOptions = Map(
      PineconeOptions.PINECONE_API_KEY_CONF -> System.getenv("PINECONE_API_KEY"),
      PineconeOptions.PINECONE_INDEX_NAME_CONF -> System.getenv("PINECONE_INDEX"),
      PineconeOptions.PINECONE_SOURCE_TAG_CONF -> System.getenv("PINECONE_SOURCE_TAG")
    )

    // Write the stream to Pinecone using the defined options
    val upsert = lines
      .writeStream
      .format("io.pinecone.spark.pinecone.Pinecone")
      .options(pineconeOptions)
      .option("checkpointLocation", "path/to/checkpoint/dir")
      .outputMode("append")
      .start()

    upsert.awaitTermination()
  }
  ```
</CodeGroup>

## Learn more

* [Spark-Pinecone connector setup guide](/integrations/databricks#setup-guide)
* [GitHub](https://github.com/pinecone-io/spark-pinecone)

---

# Source: https://docs.pinecone.io/troubleshooting/pinecone-support-slas.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Pinecone Support SLAs

<Note>
  New first-response SLAs went into effect on September 16th, 2024. See the [pricing page](https://www.pinecone.io/pricing/?plans=support) for more details.
</Note>

Pinecone Support has first-response SLAs based on the support plan of the ticket requester's organization and the selected severity. These SLAs are as follows:

### Premium

* **Sev-1**: 30 minutes
* **Sev-2**: 2 business hours
* **Sev-3**: 8 business hours
* **Sev-4**: 12 business hours

### Pro

* **Sev-1**: 2 hours
* **Sev-2**: 4 business hours
* **Sev-3**: 12 business hours
* **Sev-4**: 2 business days

### Developer

* **Sev-1**: 8 business hours
* **Sev-2**: 12 business hours
* **Sev-3**: 2 business days
* **Sev-4**: 3 business days

The current business hours for Pinecone Support are 8 AM to 8 PM Eastern Time. SLAs only apply outside of our business hours for Sev-1 tickets created by users subscribed to Pro or Premium support. All first-response SLAs only apply to tickets created by users in an organization subscribed to a [support plan](https://www.pinecone.io/pricing/?plans=support).

Pinecone Support is reserved for customers on the Standard billing plan. However, you may find helpful resources on our [community page](https://community.pinecone.io). This is a great place to ask questions and find answers from other Pinecone users and our community moderators.

---

# Source: https://docs.pinecone.io/guides/assistant/pricing-and-limits.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Pricing and limits

> Understand Pinecone Assistant pricing and service limits.

Pricing and limits vary based on [subscription plan](https://www.pinecone.io/pricing/).

## Pricing

The cost of using Pinecone Assistant is determined by the following factors:

* Monthly usage
* Hourly rate
* Tokens used
* Storage

### Minimum usage

The Standard and Enterprise [pricing plans](https://www.pinecone.io/pricing/) include a monthly minimum usage committment:

| Plan       | Minimum usage |
| ---------- | ------------- |
| Starter    | \$0/month     |
| Standard   | \$50/month    |
| Enterprise | \$500/month   |

Beyond the monthly minimum, customers are charged for what they use each month.

**Examples**

<AccordionGroup>
  <Accordion title="Usage below monthly minimum">
    * You are on the Standard plan.
    * Your usage for the month of August amounts to \$20.
    * Your usage is below the \$50 monthly minimum, so your total for the month is \$50.

    In this case, the August invoice would include line items for each service you used (totaling \$20), plus a single line item covering the rest of the minimum usage commitment (\$30).
  </Accordion>

  <Accordion title="Usage exceeds monthly minimum">
    * You are on the Standard plan.
    * Your usage for the month of August amounts to \$100.
    * Your usage exceeds the \$50 monthly minimum, so your total for the month is \$100.

    In this case, the August invoice would only show line items for each service you used (totaling \$100). Since your usage exceeds the minimum usage commitment, you are only charged for your actual usage and no additional minimum usage line item appears on your invoice.
  </Accordion>
</AccordionGroup>

### Hourly rate

For paid plans, you are charged an hourly rate for each assistant, regardless of assistant activity.

| Plan       | Hourly rate |
| ---------- | ----------- |
| Starter    | Free        |
| Standard   | \$0.05/hour |
| Enterprise | \$0.05/hour |

### Tokens

For paid plans, you are charged for the number of tokens used by each assistant.

#### Chat tokens

[Chatting with an assistant](/guides/assistant/chat-with-assistant) involves both input and output tokens:

* **Input tokens** are based on the messages sent to the assistant and the context snippets retrieved from the assistant and sent to a model. Messages sent to the assistant can include messages from the [chat history](/guides/assistant/chat-with-assistant#provide-conversation-history) in addition to the newest message.

* **Output tokens** are based on the answer from the model.

| Plan       | Input token rate            | Output token rate           |
| ---------- | --------------------------- | --------------------------- |
| Starter    | Free (1.5M max per project) | Free (200k max per project) |
| Standard   | \$8/million tokens          | \$15/million tokens         |
| Enterprise | \$8/million tokens          | \$15/million tokens         |

<Note>
  Chat input tokens appear as "Assistants Input Tokens" on invoices and `prompt_tokens` in API responses. Chat output tokens appear as "Assistants Output Tokens"" on invoices and `completion_tokens` in API responses.
</Note>

#### Context tokens

When you [retrieve context snippets](/guides/assistant/context-snippets-overview), tokens are based on the messages sent to the assistant and the context snippets retrieved from the assistant. Messages sent to the assistant can include messages from the [chat history](/guides/assistant/chat-with-assistant#provide-conversation-history) in addition to the newest message.

| Plan       | Token rate                  |
| ---------- | --------------------------- |
| Starter    | Free (500k max per project) |
| Standard   | \$5/million tokens          |
| Enterprise | \$5/million tokens          |

<Note>
  Context retrieval tokens appear as **Assistants Context Tokens Processed** on invoices and `prompt_tokens` in API responses. In API responses, `completion_tokens` will always be 0 because, unlike for chat, there is no answer from a model.
</Note>

#### Evaluation tokens

[Evaluating responses](/guides/assistant/evaluation-overview) involves both input and output tokens:

* **Input tokens** are based on two requests to a model: The first request contains a question, answer, and ground truth answer, and the second request contains the same details plus generated facts returned by the model for the first request.
* **Output tokens** are based on two responses from a model: The first response contains generated facts, and the second response contains evaluation metrics.

| Plan       | Input token rate   | Output token rate   |
| ---------- | ------------------ | ------------------- |
| Starter    | Not available      | Not available       |
| Standard   | \$8/million tokens | \$15/million tokens |
| Enterprise | \$8/million tokens | \$15/million tokens |

<Note>
  Evaluation input tokens appear as **Assistants Evaluation Tokens Processed** on invoices and `prompt_tokens` in API responses. Evaluation output tokens appear as as **Assistants Evaluation Tokens Out** on invoices and `completion_tokens` in API responses.
</Note>

### Storage

For paid plans, you are charged for the size of each assistant.

| Plan       | Storage rate                |
| ---------- | --------------------------- |
| Starter    | Free (1 GB max per project) |
| Standard   | \$3/GB per month            |
| Enterprise | \$3/GB per month            |

## Limits

Pinecone Assistant limits vary based on [subscription plan](https://www.pinecone.io/pricing/).

### Object limits

Object limits are restrictions on the number or size of assistant-related objects.

| Metric                               | Starter plan  | Standard plan | Enterprise plan |
| :----------------------------------- | :------------ | :------------ | :-------------- |
| Assistants per project               | 5             | Unlimited     | Unlimited       |
| File storage per project             | 1 GB          | Unlimited     | Unlimited       |
| Chat input tokens per project        | 1,500,000     | Unlimited     | Unlimited       |
| Chat output tokens per project       | 200,000       | Unlimited     | Unlimited       |
| Context retrieval tokens per project | 500,000       | Unlimited     | Unlimited       |
| Evaluation input tokens per project  | Not available | 150,000       | 500,000         |
| Files per assistant                  | 100           | 10,000        | 10,000          |
| File size (.docx, .json, .md, .txt)  | 10 MB         | 10 MB         | 10 MB           |
| File size (.pdf)                     | 10 MB         | 100 MB        | 100 MB          |
| Metadata size per file               | 16 KB         | 16 KB         | 16 KB           |

Additionally, the following limits apply to [multimodal PDFs](/guides/assistant/multimodal) (currently in [public preview](/release-notes/feature-availability)):

| Metric                        | Starter plan | Standard plan | Enterprise plan |
| :---------------------------- | :----------- | :------------ | :-------------- |
| Max file size                 | 10 MB        | 50 MB         | 50 MB           |
| Page limit                    | 100          | 100           | 100             |
| Multimodal PDFs per assistant | 10           | 20            | 20              |

### Rate limits

Rate limits help protect your applications from misuse and maintain the health of our shared infrastructure. These limits are designed to support typical production workloads while ensuring reliable performance for all users.

**Most rate limits can be adjusted upon request.** If you need higher limits to scale your application, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket) with details about your use case.

Requests that exceed a rate limit fail and return a `429 - TOO_MANY_REQUESTS` status.

<Tip>To handle rate limits, implement [retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic).</Tip>

| Metric                                      | Starter plan | Standard plan | Enterprise plan |
| :------------------------------------------ | :----------- | :------------ | :-------------- |
| Assistant list/get requests per minute      | 40           | 100           | 500             |
| Assistant create/update requests per minute | 20           | 50            | 100             |
| Assistant delete requests per minute        | 20           | 50            | 100             |
| File get requests per minute                | 100          | 300           | 6,000           |
| File list requests per minute               | 50           | 150           | 3,000           |
| File upload requests per minute             | 5            | 20            | 300             |
| File delete requests per minute             | 5            | 20            | 300             |
| Chat input tokens per minute                | 100,000      | 300,000       | 1,000,000       |
| Chat history tokens per query               | 64,000       | 64,000        | 64,000          |

---

# Source: https://docs.pinecone.io/guides/production/production-checklist.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Production checklist

> Prepare your indexes for production with best practices.

This page provides recommendations and best practices for preparing your Pinecone indexes for production, anticipating production issues, and enabling reliability and growth.

## Prepare your project structure

One of the first steps towards building a production-ready Pinecone index is configuring your project correctly.

* Consider [creating a separate project](/guides/projects/create-a-project) for your development and production indexes, to allow for testing changes to your index before deploying them to production.
* Ensure that you have properly [configured user access](/guides/projects/understanding-projects#project-roles) to the Pinecone console, so that only those users who need to access the production index can do so.
* Ensure that you have properly configured access through the API by [managing API keys](/guides/projects/manage-api-keys) and using API key permissions.

Consider how best to [manage the API keys](/guides/projects/manage-api-keys) associated with your production project. In order to [make calls to the Pinecone API](/guides/get-started/quickstart), you must provide a valid API key for the relevant Pinecone project.

## Enforce security

Use Pinecone's [security features](/guides/production/security-overview) to protect your production data:

* Data security
  * Private endpoints
  * Customer-managed encryption keys (CMEK)
* Authorization
  * API keys
  * Role-based access control (RBAC)
  * Organization single sign-on (SSO)
* Audit logs
* Bring your own cloud

## Design your indexes for scale

Follow these best practices when designing and populating your indexes:

* **Data ingestion**: For large datasets (10M+ records), [import from object storage](/guides/index-data/import-data) for the most efficient and cost-effective ingestion. For ongoing ingestion, [upsert in batches](/guides/index-data/upsert-data#upsert-in-batches) to optimize speed and efficiency. See the [data ingestion overview](/guides/index-data/data-ingestion-overview) for details.
* **Dimensionality**: Consider the dimensionality of your vectors. Higher dimensions can offer more accuracy but require more resources.
* **Data modeling**: Use [structured IDs](/guides/index-data/data-modeling#use-structured-ids) (e.g., `document_id#chunk_number`) for efficient operations. Design [metadata](/guides/index-data/data-modeling#include-metadata) to support filtering, linking related chunks, and traceability. See the [data modeling guide](/guides/index-data/data-modeling) for details.
* **Namespaces**: When indexing, try to [use namespaces to keep your data among tenants separate](/guides/index-data/implement-multitenancy), and do not use multiple indexes for this purpose. Namespaces are more efficient and more affordable in the long run.

## Understand database limits

Architect your application to work within Pinecone's [database limits](/reference/api/database-limits):

* **Rate limits**: Serverless indexes have per-second operation limits for queries, upserts, updates, and deletes. [Implement error handling with exponential backoff](/guides/production/error-handling) to handle rate limit errors gracefully.
* **Size limits**: Be aware of constraints on vector dimensionality, metadata size per record, record ID length, maximum `top_k` values, and query result sizes. Design your [data model](/guides/index-data/data-modeling) accordingly.
* **Index limits**: Plan for index capacity based on your [plan tier](https://www.pinecone.io/pricing/). Use [namespaces](/guides/index-data/implement-multitenancy) to partition data within indexes rather than creating multiple indexes.
* **Plan limits**: Starter plans have monthly read/write unit limits. Upgrade to Standard or Enterprise for unlimited read/write units and higher throughput needs.

<Note>
  Indexes built on [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) are not subject to read unit limits for query, fetch, and list operations. For sizing and capacity planning guidance, see the [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) guide.
</Note>

## Test your query results

Before you move your index to production, make sure that your index is returning accurate results in the context of your application by [identifying the appropriate metrics](https://www.pinecone.io/learn/offline-evaluation/) for evaluating your results.

## Optimize performance

Before serving production workloads, optimize your Pinecone implementation:

* **Increase search relevance**: Use techniques like reranking, metadata filtering, hybrid search, and chunking strategies to improve result quality. See [increase search relevance](/guides/optimize/increase-relevance) for details.
* **Increase throughput**: Import from object storage, upsert in batches, use parallel operations, and leverage Python SDK optimizations like gRPC. See [increase throughput](/guides/optimize/increase-throughput) for details.
* **Decrease latency**: Use namespaces, filter by metadata, target indexes by host, reuse connections, and deploy in the same cloud region as your index. See [decrease latency](/guides/optimize/decrease-latency) for details.

## Backup up your indexes

In order to enable long-term retention, compliance archiving, and deployment of new indexes, consider backing up your production indexes by [creating a backup or collection](/guides/manage-data/back-up-an-index).

## Implement error handling

Prepare your application to handle errors gracefully:

* Implement [error handling and retry logic](/guides/production/error-handling) with exponential backoff
* Handle different error types appropriately (4xx vs 5xx)
* Monitor error rates and set up alerts
* Check [status.pinecone.io](https://status.pinecone.io) before escalating issues

## Configure monitoring

Prepare to [monitor the production performance and availability of your indexes](/guides/production/monitoring).

## Configure CI/CD

Use [Pinecone in CI/CD](/guides/production/automated-testing) to safely test changes before deploying them to production.

## Know how to get support

If you need help, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket), or talk to the [Pinecone community](https://www.pinecone.io/community/). Ensure that your [plan tier](https://www.pinecone.io/pricing/) matches the support and availability SLAs you need. This may require you to upgrade to Enterprise.

---

# Source: https://docs.pinecone.io/guides/assistant/admin/projects-overview.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Projects overview

> Learn about projects, roles, and collaboration.

A Pinecone project belongs to an [organization](/guides/assistant/admin/organizations-overview) and contains a number of [assistants](/guides/assistant/overview) and users. Only a user who belongs to the project can access the indexes in that project. Each project also has at least one project owner.

## Project roles

If you are an [organization owner](/guides/assistant/admin/organizations-overview#organization-roles) or project owner, you can manage members in your project. You assign project members a specific role that determines the member's permissions within the Pinecone console.

When you invite a member at the project-level, you assign one of the following roles:

* **Project owner**: Project owners have global permissions across projects they own.

* **Project user**: Project users have restricted permissions for the specific projects they are invited to.

The following table summarizes the permissions for each project role:

| Permission                  | Owner | User |
| :-------------------------- | ----- | ---- |
| Update project names        | ✓     |      |
| Delete projects             | ✓     |      |
| View project members        | ✓     | ✓    |
| Update project member roles | ✓     |      |
| Delete project members      | ✓     |      |
| View API keys               | ✓     | ✓    |
| Create API keys             | ✓     |      |
| Delete API keys             | ✓     |      |
| View indexes                | ✓     | ✓    |
| Create indexes              | ✓     | ✓    |
| Delete indexes              | ✓     | ✓    |
| Upsert vectors              | ✓     | ✓    |
| Query vectors               | ✓     | ✓    |
| Fetch vectors               | ✓     | ✓    |
| Update a vector             | ✓     | ✓    |
| Delete a vector             | ✓     | ✓    |
| List vector IDs             | ✓     | ✓    |
| Get index stats             | ✓     | ✓    |

Specific to pod-based indexes:

<Warning>
  Customers who sign up for a Standard or Enterprise plan on or after August 18, 2025 cannot create pod-based indexes. Instead, create [serverless indexes](/guides/index-data/create-an-index), and consider using [dedicated read nodes](/guides/index-data/dedicated-read-nodes) for large workloads (millions of records or more, and moderate or high query rates).
</Warning>

| Permission                | Owner | User |
| :------------------------ | ----- | ---- |
| Update project pod limits | ✓     |      |
| View project pod limits   | ✓     | ✓    |
| Update index size         | ✓     | ✓    |

## API keys

Each Pinecone [project](/guides/assistant/admin/projects-overview) has one or more API keys. In order to [make calls to the Pinecone API](/guides/assistant/quickstart/sdk-quickstart), you must provide a valid API key for the relevant Pinecone project.

For more information, see [Manage API keys](/guides/assistant/admin/manage-api-keys).

## Service accounts

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

[Service accounts](/guides/assistant/admin/manage-organization-service-accounts) enable programmatic access to Pinecone's Admin API, which can be used to create and manage projects and API keys.

Use service accounts to automate infrastructure management and integrate Pinecone into your deployment workflows, rather than through manual actions in the Pinecone console. Service accounts use the [organization roles](/guides/assistant/admin/organizations-overview#organization-roles) and [project role](/guides/assistant/admin/projects-overview#project-roles) for permissioning, and provide a secure and auditable way to handle programmatic access.

To use service accounts, [add the account to your organization](/guides/assistant/admin/manage-organization-service-accounts) before [connecting it to a project](/guides/assistant/admin/manage-project-service-accounts).

## Project IDs

Each Pinecone project has a unique product ID.

To find the ID of a project, go to the project list in the [Pinecone console](https://app.pinecone.io/organizations/-/projects).

## See also

* [Create a project](guides/assistant/admin/create-a-project)
* [Manage project members](guides/assistant/admin/manage-project-members)
* [Organizations overview](guides/assistant/admin/organizations-overview)

---

# Source: https://docs.pinecone.io/integrations/pulumi.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Pulumi

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Pulumi is an infrastructure as code platform that allows you to use familiar programming languages and tools to build, deploy, and manage cloud infrastructure. Pulumi is free, open source, and optionally pairs with the Pulumi Cloud to make managing infrastructure secure, reliable, and hassle-free.

This Pulumi Pinecone Provider enables you to manage your Pinecone collections and indexes using any language of Pulumi Infrastructure as Code.

<PrimarySecondaryCTA primaryLabel={"Get started"} primaryHref={"https://www.pulumi.com/registry/packages/pinecone/"} />

---

# Source: https://docs.pinecone.io/reference/python-sdk.md

# Python SDK

The Pinecone Python SDK is distributed on PyPI using the package name `pinecone`. By default, the `pinecone` package has a minimal set of dependencies and interacts with Pinecone via HTTP requests. However, you can install the following extras to unlock additional functionality:

* `pinecone[grpc]` adds dependencies on `grpcio` and related libraries needed to run data operations such as upserts and queries over [gRPC](https://grpc.io/) for a modest performance improvement.

* `pinecone[asyncio]` adds a dependency on `aiohttp` and enables usage of `async` methods for use with [asyncio](https://docs.python.org/3/library/asyncio.html). For more details, see [Asyncio support](#async-requests).

<Tip>
  See the [Pinecone Python SDK
  documentation](https://sdk.pinecone.io/python/)
  for full installation instructions, usage examples, and reference information.

  To make a feature request or report an issue, please [file an issue](https://github.com/pinecone-io/pinecone-python-client/issues).
</Tip>

## Requirements

The Pinecone Python SDK requires Python 3.9 or later. It has been tested with CPython versions from 3.9 to 3.13.

## SDK versions

SDK versions are pinned to specific [API versions](/reference/api/versioning). When a new API version is released, a new version of the SDK is also released.

The mappings between API versions and Python SDK versions are as follows:

| API version        | SDK version   |
| :----------------- | :------------ |
| `2025-04` (latest) | v7.x          |
| `2025-01`          | v6.x          |
| `2024-10`          | v5.3.x        |
| `2024-07`          | v5.0.x-v5.2.x |
| `2024-04`          | v4.x          |

When a new stable API version is released, you should upgrade your SDK to the latest version to ensure compatibility with the latest API changes.

## Install

To install the latest version of the [Python SDK](https://github.com/pinecone-io/pinecone-python-client), run the following command:

```shell  theme={null}
# Install the latest version
pip install pinecone

# Install the latest version with gRPC extras
pip install "pinecone[grpc]"

# Install the latest version with asyncio extras
pip install "pinecone[asyncio]"
```

To install a specific version of the Python SDK, run the following command:

```shell pip theme={null}
# Install a specific version
pip install pinecone==<version>

# Install a specific version with gRPC extras
pip install "pinecone[grpc]"==<version>

# Install a specific version with asyncio extras
pip install "pinecone[asyncio]"==<version>
```

To check your SDK version, run the following command:

```shell pip theme={null}
pip show pinecone
```

<Note>
  To use the [Inference API](/reference/api/introduction#inference), you must be on version 5.0.0 or later.
</Note>

### Install the Pinecone Assistant Python plugin

As of Python SDK v7.0.0, the `pinecone-plugin-assistant` package is included by default. It is only necessary to install the package if you are using a version of the Python SDK prior to v7.0.0.

```shell HTTP theme={null}
pip install --upgrade pinecone pinecone-plugin-assistant
```

## Upgrade

<Warning>
  Before upgrading to `v6.0.0`, update all relevant code to account for the breaking changes explained [here](https://github.com/pinecone-io/pinecone-python-client/blob/main/docs/upgrading.md).

  Also, make sure to upgrade using the `pinecone` package name instead of `pinecone-client`; upgrading with the latter will not work as of `v6.0.0`.
</Warning>

If you already have the Python SDK, upgrade to the latest version as follows:

```shell  theme={null}
# Upgrade to the latest version
pip install pinecone --upgrade

# Upgrade to the latest version with gRPC extras
pip install "pinecone[grpc]" --upgrade

# Upgrade to the latest version with asyncio extras
pip install "pinecone[asyncio]" --upgrade
```

## Initialize

Once installed, you can import the library and then use an [API key](/guides/projects/manage-api-keys) to initialize a client instance:

<CodeGroup>
  ```Python HTTP theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")
  ```

  ```python gRPC theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")
  ```
</CodeGroup>

When [creating an index](/guides/index-data/create-an-index), import the `ServerlessSpec` or `PodSpec` class as well:

<CodeGroup>
  ```Python Serverless index theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone
  from pinecone import ServerlessSpec

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.create_index(
    name="docs-example",
    dimension=1536,
    metric="cosine",
    spec=ServerlessSpec(
      cloud="aws",
      region="us-east-1"
    )
  )
  ```

  ```Python Pod-based index theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone
  from pinecone import PodSpec

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.create_index(
    name="docs-example",
    dimension=1536,
    metric="cosine",
    spec=PodSpec(
      environment="us-west-1-gcp",
      pod_type="p1.x1",
      pods=1
    )
  )
  ```
</CodeGroup>

## Proxy configuration

If your network setup requires you to interact with Pinecone through a proxy, you will need to pass additional configuration using optional keyword parameters:

* `proxy_url`: The location of your proxy. This could be an HTTP or HTTPS URL depending on your proxy setup.
* `proxy_headers`: Accepts a python dictionary which can be used to pass any custom headers required by your proxy. If your proxy is protected by authentication, use this parameter to pass basic authentication headers with a digest of your username and password. The `make_headers` utility from `urllib3` can be used to help construct the dictionary. **Note:** Not supported with Asyncio.
* `ssl_ca_certs`: By default, the client will perform SSL certificate verification using the CA bundle maintained by Mozilla in the [`certifi`](https://pypi.org/project/certifi/) package. If your proxy is using self-signed certicates, use this parameter to specify the path to the certificate (PEM format).
* `ssl_verify`: SSL verification is enabled by default, but it is disabled when set to `False`. It is not recommened to go into production with SSL verification disabled.

<CodeGroup>
  ```python HTTP theme={null}
  from pinecone import Pinecone
  import urllib3 import make_headers

  pc = Pinecone(
      api_key="YOUR_API_KEY",
      proxy_url='https://your-proxy.com',
      proxy_headers=make_headers(proxy_basic_auth='username:password'),
      ssl_ca_certs='path/to/cert-bundle.pem'
  )
  ```

  ```python gRPC theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone
  import urllib3 import make_headers

  pc = Pinecone(
      api_key="YOUR_API_KEY",
      proxy_url='https://your-proxy.com',
      proxy_headers=make_headers(proxy_basic_auth='username:password'),
      ssl_ca_certs='path/to/cert-bundle.pem'
  )
  ```

  ```python asyncio theme={null}
  import asyncio
  from pinecone import PineconeAsyncio
  	
  async def main():
      async with PineconeAsyncio(
          api_key="YOUR_API_KEY",
          proxy_url='https://your-proxy.com',
          ssl_ca_certs='path/to/cert-bundle.pem'
      ) as pc:
          # Do async things
          await pc.list_indexes()

  asyncio.run(main())
  ```
</CodeGroup>

## Async requests

Pinecone Python SDK versions 6.0.0 and later provide `async` methods for use with [asyncio](https://docs.python.org/3/library/asyncio.html). Asyncio support makes it possible to use Pinecone with modern async web frameworks such as [FastAPI](https://fastapi.tiangolo.com/), [Quart](https://quart.palletsprojects.com/en/latest/), and [Sanic](https://sanic.dev/en/), and should significantly increase the efficiency of running requests in parallel.

Use the [`PineconeAsyncio`](https://sdk.pinecone.io/python/asyncio.html) class to create and manage indexes and the [`IndexAsyncio`](https://sdk.pinecone.io/python/asyncio.html#pinecone.db_data.IndexAsyncio) class to read and write index data. To ensure that sessions are properly closed, use the `async with` syntax when creating `PineconeAsyncio` and `IndexAsyncio` objects.

<CodeGroup>
  ```python Manage indexes theme={null}
  # pip install "pinecone[asyncio]"
  import asyncio
  from pinecone import PineconeAsyncio, ServerlessSpec

  async def main():
      async with PineconeAsyncio(api_key="YOUR_API_KEY") as pc:
          if not await pc.has_index(index_name):
              desc = await pc.create_index(
                  name="docs-example",
                  dimension=1536,
                  metric="cosine",
                  spec=ServerlessSpec(
                      cloud="aws",
                      region="us-east-1"
                  ),
                  deletion_protection="disabled",
                  tags={
                      "environment": "development"
                  }
              )

  asyncio.run(main())
  ```

  ```python Read and write index data theme={null}
  # pip install "pinecone[asyncio]"
  import asyncio
  from pinecone import Pinecone

  async def main():
      pc = Pinecone(api_key="YOUR_API_KEY")
      async pc.IndexAsyncio(host="INDEX_HOST") as idx:
          await idx.upsert_records(
              namespace="example-namespace",
              records=[
                  {
                      "id": "1",
                      "title": "The Great Gatsby",
                      "author": "F. Scott Fitzgerald",
                      "description": "The story of the mysteriously wealthy Jay Gatsby and his love for the beautiful Daisy Buchanan.",
                      "year": 1925,
                  },
                  {
                      "id": "2",
                      "title": "To Kill a Mockingbird",
                      "author": "Harper Lee",
                      "description": "A young girl comes of age in the segregated American South and witnesses her father's courageous defense of an innocent black man.",
                      "year": 1960,
                  },
                  {
                      "id": "3",
                      "title": "1984",
                      "author": "George Orwell",
                      "description": "In a dystopian future, a totalitarian regime exercises absolute control through pervasive surveillance and propaganda.",
                      "year": 1949,
                  },
              ]
          )

  asyncio.run(main())
  ```
</CodeGroup>

## Query across namespaces

Each query is limited to a single [namespace](/guides/index-data/indexing-overview#namespaces). However, the Pinecone Python SDK provides a `query_namespaces` utility method to run a query in parallel across multiple namespaces in an index and then merge the result sets into a single ranked result set with the `top_k` most relevant results.

The `query_namespaces` method accepts most of the same arguments as `query` with the addition of a required `namespaces` parameter.

<Tabs>
  <Tab title="Python SDK without gRPC">
    When using the Python SDK without gRPC extras, to get good performance, it is important to set values for the `pool_threads` and `connection_pool_maxsize` properties on the index client. The `pool_threads` setting is the number of threads available to execute requests, while `connection_pool_maxsize` is the number of cached http connections that will be held. Since these tasks are not computationally heavy and are mainly i/o bound, it should be okay to have a high ratio of threads to cpus.

    The combined results include the sum of all read unit usage used to perform the underlying queries for each namespace.

    ```python Python theme={null}
    from pinecone import Pinecone

    pc = Pinecone(api_key="YOUR_API_KEY")
    index = pc.Index(
        name="docs-example",
        pool_threads=50,             # <-- make sure to set these
        connection_pool_maxsize=50,  # <-- make sure to set these
    )

    query_vec = [ 0.1, ...] # an embedding vector with same dimension as the index
    combined_results = index.query_namespaces(
        vector=query_vec,
        namespaces=['ns1', 'ns2', 'ns3', 'ns4'],
        metric="cosine",
        top_k=10,
        include_values=False,
        include_metadata=True,
        filter={"genre": { "$eq": "comedy" }},
        show_progress=False,
    )

    for scored_vec in combined_results.matches:
        print(scored_vec)
    print(combined_results.usage)
    ```
  </Tab>

  <Tab title="Python SDK with gRPC">
    When using the Python SDK with gRPC extras, there is no need to set the `connection_pool_maxsize` because grpc makes efficient use of open connections by default.

    ```python Python theme={null}
    from pinecone.grpc import PineconeGRPC

    pc = PineconeGRPC(api_key="YOUR_API_KEY")
    index = pc.Index(
        name="docs-example",
        pool_threads=50, # <-- make sure to set this
    )

    query_vec = [ 0.1, ...] # an embedding vector with same dimension as the index
    combined_results = index.query_namespaces(
        vector=query_vec,
        namespaces=['ns1', 'ns2', 'ns3', 'ns4'],
        metric="cosine",
        top_k=10,
        include_values=False,
        include_metadata=True,
        filter={"genre": { "$eq": "comedy" }},
        show_progress=False,
    )

    for scored_vec in combined_results.matches:
        print(scored_vec)
    print(combined_results.usage)
    ```
  </Tab>
</Tabs>

## Upsert from a dataframe

To quickly ingest data when using the [Python SDK](/reference/python-sdk), use the `upsert_from_dataframe` method. The method includes retry logic and`batch_size`, and is performant especially with Parquet file data sets.

The following example upserts the `uora_all-MiniLM-L6-bm25` dataset as a dataframe.

```Python Python theme={null}
from pinecone import Pinecone, ServerlessSpec
from pinecone_datasets import list_datasets, load_dataset

pc = Pinecone(api_key="API_KEY")

dataset = load_dataset("quora_all-MiniLM-L6-bm25")

pc.create_index(
  name="docs-example",
  dimension=384,
  metric="cosine",
  spec=ServerlessSpec(
    cloud="aws",
    region="us-east-1"
  )
)

# To get the unique host for an index, 
# see https://docs.pinecone.io/guides/manage-data/target-an-index
index = pc.Index(host="INDEX_HOST")

index.upsert_from_dataframe(dataset.drop(columns=["blob"]))
```

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/query.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Search with a vector

> Search a namespace using a query vector. It retrieves the ids of the most similar items in a namespace, along with their similarity scores.

For guidance, examples, and limits, see [Search](https://docs.pinecone.io/guides/search/search-overview).

<RequestExample>
  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/query" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "namespace": "example-namespace",
      "vector": [0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3],
      "filter": {"genre": {"$eq": "documentary"}},
      "topK": 3,
      "includeValues": true
    }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "matches":[
      {
        "id": "vec3",
        "score": 0,
        "values": [0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3]
      },
      {
        "id": "vec2",
        "score": 0.0800000429,
        "values": [0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2]
      },
      {
        "id": "vec4",
        "score": 0.0799999237,
        "values": [0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4]
      }
    ],
    "namespace": "example-namespace",
    "usage": {"read_units": 6}
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml post /query
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /query:
    post:
      tags:
        - Vector Operations
      summary: Search with a vector
      description: >-
        Search a namespace using a query vector. It retrieves the ids of the
        most similar items in a namespace, along with their similarity scores.


        For guidance, examples, and limits, see
        [Search](https://docs.pinecone.io/guides/search/search-overview).
      operationId: queryVectors
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/QueryRequest'
        required: true
      responses:
        '200':
          description: A successful response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/QueryResponse'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    QueryRequest:
      description: The request for the `query` operation.
      type: object
      properties:
        namespace:
          example: example-namespace
          description: The namespace to query.
          type: string
        topK:
          example: 10
          description: The number of results to return for each query.
          type: integer
          minimum: 1
          maximum: 10000
          required:
            - top_k
          format: int64
        filter:
          example:
            genre:
              $in:
                - comedy
                - documentary
                - drama
            year:
              $eq: 2019
          description: >-
            The filter to apply. You can use vector metadata to limit your
            search. See [Understanding
            metadata](https://docs.pinecone.io/guides/index-data/indexing-overview#metadata).
          type: object
        includeValues:
          example: true
          description: >-
            Indicates whether vector values are included in the response. For
            on-demand indexes, setting this to `true` may increase latency,
            especially with higher `topK` values, because vector values are
            retrieved from object storage. Unless you need vector values, set
            this to `false` for better performance.
          default: false
          type: boolean
        includeMetadata:
          example: true
          description: >-
            Indicates whether metadata is included in the response as well as
            the ids.
          default: false
          type: boolean
        queries:
          deprecated: true
          description: DEPRECATED. Use `vector` or `id` instead.
          type: array
          items:
            $ref: '#/components/schemas/QueryVector'
          minLength: 1
          maxLength: 10
        vector:
          example:
            - 0.1
            - 0.2
            - 0.3
            - 0.4
            - 0.5
            - 0.6
            - 0.7
            - 0.8
          description: >-
            The query vector. This should be the same length as the dimension of
            the index being queried. Each `query` request can contain only one
            of the parameters `id` or `vector`.
          type: array
          items:
            type: number
            format: float
          minLength: 1
          maxLength: 20000
        sparseVector:
          $ref: '#/components/schemas/SparseValues'
        id:
          example: example-vector-1
          description: >-
            The unique ID of the vector to be used as a query vector. Each
            request  can contain either the `vector` or `id` parameter.
          type: string
          maxLength: 512
      required:
        - topK
    QueryResponse:
      description: >-
        The response for the `query` operation. These are the matches found for
        a particular query vector. The matches are ordered from most similar to
        least similar.
      type: object
      properties:
        results:
          deprecated: true
          description: >-
            DEPRECATED. The results of each query. The order is the same as
            `QueryRequest.queries`.
          type: array
          items:
            $ref: '#/components/schemas/SingleQueryResults'
        matches:
          description: The matches for the vectors.
          type: array
          items:
            $ref: '#/components/schemas/ScoredVector'
        namespace:
          description: The namespace for the vectors.
          type: string
        usage:
          $ref: '#/components/schemas/Usage'
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    QueryVector:
      deprecated: true
      description: A single query vector within a `QueryRequest`.
      type: object
      properties:
        values:
          example:
            - 0.1
            - 0.2
            - 0.3
            - 0.4
            - 0.5
            - 0.6
            - 0.7
            - 0.8
          description: >-
            The query vector values. This should be the same length as the
            dimension of the index being queried.
          type: array
          required:
            - values
          items:
            type: number
            format: float
          minLength: 1
          maxLength: 20000
        sparseValues:
          $ref: '#/components/schemas/SparseValues'
        topK:
          example: 10
          description: >-
            An override for the number of results to return for this query
            vector.
          type: integer
          format: int64
          minimum: 1
          maximum: 10000
        namespace:
          example: example-namespace
          description: An override the namespace to search.
          type: string
        filter:
          example:
            genre:
              $in:
                - comedy
                - documentary
                - drama
            year:
              $eq: 2019
          description: >-
            An override for the metadata filter to apply. This replaces the
            request-level filter.
          type: object
      required:
        - values
    SparseValues:
      description: >-
        Vector sparse data. Represented as a list of indices and a list of 
        corresponded values, which must be with the same length.
      type: object
      properties:
        indices:
          example:
            - 1
            - 312
            - 822
            - 14
            - 980
          description: The indices of the sparse data.
          type: array
          required:
            - indices
          items:
            type: integer
            format: int64
          minLength: 1
          maxLength: 1000
        values:
          example:
            - 0.1
            - 0.2
            - 0.3
            - 0.4
            - 0.5
          description: >-
            The corresponding values of the sparse data, which must be with the
            same length as the indices.
          type: array
          required:
            - values
          items:
            type: number
            format: float
          minLength: 1
          maxLength: 1000
      required:
        - indices
        - values
    SingleQueryResults:
      title: The query results for a single `QueryVector`
      type: object
      properties:
        matches:
          description: The matches for the vectors.
          type: array
          items:
            $ref: '#/components/schemas/ScoredVector'
        namespace:
          example: example-namespace
          description: The namespace for the vectors.
          type: string
    ScoredVector:
      type: object
      properties:
        id:
          example: example-vector-1
          description: This is the vector's unique id.
          type: string
          required:
            - id
          minLength: 1
          maxLength: 512
        score:
          example: 0.08
          description: >-
            This is a measure of similarity between this vector and the query
            vector.  The higher the score, the more they are similar.
          type: number
          format: float
        values:
          example:
            - 0.1
            - 0.2
            - 0.3
            - 0.4
            - 0.5
            - 0.6
            - 0.7
            - 0.8
          description: This is the vector data, if it is requested.
          type: array
          items:
            type: number
            format: float
        sparseValues:
          $ref: '#/components/schemas/SparseValues'
        metadata:
          example:
            genre: documentary
            year: 2019
          description: This is the metadata, if it is requested.
          type: object
      required:
        - id
    Usage:
      type: object
      properties:
        readUnits:
          example: 5
          description: The number of read units consumed by this operation.
          type: integer
          format: int64
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/cli/quickstart.md

# Source: https://docs.pinecone.io/guides/get-started/quickstart.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Quickstart

> Get started with Pinecone manually, with AI assistance, or with no-code tools.

This quickstart walks you through creating a Pinecone index and building a sample application for semantic search, recommendations, or RAG.

## Get set up

To get started, you'll need a Pinecone account and API key.

### 1. Create a Pinecone account

If you're new to Pinecone, sign up at [app.pinecone.io](https://app.pinecone.io) and choose a free plan:

* [Starter plan](https://pinecone.io/pricing/): Free access to most features, but you're limited to one cloud region and need to stay under Starter plan [limits](/reference/api/database-limits).

* [Standard plan trial](/guides/organizations/manage-billing/standard-trial): 21 days and \$300 in credits with access to Standard plan [features](https://www.pinecone.io/pricing/) and [higher limits](/reference/api/database-limits) that let you test Pinecone at scale.

<Note>
  If you're already on a Starter plan, you can activate a Standard plan trial at any time (one trial per organization).
</Note>

After signing up, you'll receive an API key in the console. Save this key. You'll need it to authenticate your requests to Pinecone.

### 2. Get a Pinecone API key

Create a new API key in the [Pinecone console](https://app.pinecone.io/organizations/-/keys), or use the widget below to generate a key. If you don't have a Pinecone account, the widget will sign you up for the free [Starter plan](https://www.pinecone.io/pricing/).

<div style={{minWidth: '450px', minHeight:'152px'}}>
  <div id="pinecone-connect-widget">
    <div class="connect-widget-skeleton">
      <div class="skeleton-content" />
    </div>
  </div>
</div>

Your generated API key:

```shell  theme={null}
"{{YOUR_API_KEY}}"
```

## Build with Pinecone

Choose your approach to build with Pinecone below. Each approach achieves the same result—building a sample app for semantic search—but uses different tools and workflows.

<Tabs>
  <Tab title="Manual">
    Use [Pinecone's SDKs](/reference/pinecone-sdks) to manually create indexes, upsert data, and run queries.

    ### 1. Install an SDK

    Install the SDK for your preferred language:

    <CodeGroup>
      ```shell Python theme={null}
      pip install pinecone
      ```

      ```shell JavaScript theme={null}
      npm install @pinecone-database/pinecone
      ```

      ```shell Java theme={null}
      # Maven
      <dependency>
        <groupId>io.pinecone</groupId>
        <artifactId>pinecone-client</artifactId>
        <version>5.0.0</version>
      </dependency>

      # Gradle
      implementation "io.pinecone:pinecone-client:5.0.0"
      ```

      ```shell Go theme={null}
      go get github.com/pinecone-io/go-pinecone/v4/pinecone
      ```

      ```shell C# theme={null}
      # .NET Core CLI
      dotnet add package Pinecone.Client 

      # NuGet CLI
      nuget install Pinecone.Client
      ```
    </CodeGroup>

    <Tip>
      To get started in your browser, use the [Quickstart colab notebook](https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/pinecone-quickstart.ipynb).
    </Tip>

    ### 2. Create an index

    In Pinecone, there are two types of indexes for storing vector data: [Dense indexes](/guides/index-data/indexing-overview#dense-indexes) store <Tooltip tip="Each number in a dense vector is a point in a multidimensional space. Vectors that are closer together in that space are semantically similar.">dense vectors</Tooltip> for semantic search, and [sparse indexes](/guides/index-data/indexing-overview#sparse-indexes) store <Tooltip tip="A sparse vector has a very large number of dimensions, where only a small proportion of values are non-zero. The dimensions represent words from a dictionary, and the values represent the importance of these words in the document.">sparse vectors</Tooltip> for lexical/keyword search.

    For this quickstart, [create a dense index](/guides/index-data/create-an-index#create-a-dense-index) that is integrated with an [embedding model hosted by Pinecone](/guides/index-data/create-an-index#embedding-models). With integrated models, you upsert and search with text and have Pinecone generate vectors automatically.

    <Note>
      If you prefer to use external embedding models, see [Bring your own vectors](/guides/index-data/indexing-overview#bring-your-own-vectors).
    </Note>

    <CodeGroup>
      ```python Python theme={null}
      # Import the Pinecone library
      from pinecone import Pinecone

      # Initialize a Pinecone client with your API key
      pc = Pinecone(api_key="{{YOUR_API_KEY}}")

      # Create a dense index with integrated embedding
      index_name = "quickstart-py"
      if not pc.has_index(index_name):
          pc.create_index_for_model(
              name=index_name,
              cloud="aws",
              region="us-east-1",
              embed={
                  "model":"llama-text-embed-v2",
                  "field_map":{"text": "chunk_text"}
              }
          )
      ```

      ```javascript JavaScript theme={null}
      // Import the Pinecone library
      import { Pinecone } from '@pinecone-database/pinecone'

      // Initialize a Pinecone client with your API key
      const pc = new Pinecone({ apiKey: '{{YOUR_API_KEY}}' });

      // Create a dense index with integrated embedding
      const indexName = 'quickstart-js';
      await pc.createIndexForModel({
        name: indexName,
        cloud: 'aws',
        region: 'us-east-1',
        embed: {
          model: 'llama-text-embed-v2',
          fieldMap: { text: 'chunk_text' },
        },
        waitUntilReady: true,
      });
      ```

      ```java Java theme={null}
      import io.pinecone.clients.Index;
      import io.pinecone.clients.Pinecone;
      import org.openapitools.db_control.client.ApiException;
      import org.openapitools.db_control.client.model.CreateIndexForModelRequest;
      import org.openapitools.db_control.client.model.CreateIndexForModelRequestEmbed;
      import org.openapitools.db_control.client.model.DeletionProtection;
      import org.openapitools.db_control.client.model.IndexModel;
      import org.openapitools.db_data.client.model.SearchRecordsRequestQuery;
      import org.openapitools.db_data.client.model.SearchRecordsResponse;
      import io.pinecone.proto.DescribeIndexStatsResponse;

      import java.util.*;

      public class Quickstart {
          public static void main(String[] args) throws ApiException {
              Pinecone pc = new Pinecone.Builder("{{YOUR_API_KEY}}").build();
              String indexName = "quickstart-java";
              String region = "us-east-1";
              HashMap<String, String> fieldMap = new HashMap<>();
              fieldMap.put("text", "chunk_text");
              CreateIndexForModelRequestEmbed embed = new CreateIndexForModelRequestEmbed()
                      .model("llama-text-embed-v2")
                      .fieldMap(fieldMap);
              IndexModel index = pc.createIndexForModel(
                      indexName,
                      CreateIndexForModelRequest.CloudEnum.AWS,
                      region,
                      embed,
                      DeletionProtection.DISABLED,
                      null
              );
          }
      }
      ```

      ```go Go theme={null}
      package main

      import (
          "context"
          "encoding/json"
          "fmt"
          "log"

          "github.com/pinecone-io/go-pinecone/v4/pinecone"
      )

      func main() {
          ctx := context.Background()

          pc, err := pinecone.NewClient(pinecone.NewClientParams{
              ApiKey: "{{YOUR_API_KEY}}",
          })
          if err != nil {
              log.Fatalf("Failed to create Client: %v", err)
          }

        	indexName := "quickstart-go"
          index, err := pc.CreateIndexForModel(ctx, &pinecone.CreateIndexForModelRequest{
              Name:   indexName,
              Cloud:  pinecone.Aws,
              Region: "us-east-1",
              Embed: pinecone.CreateIndexForModelEmbed{
                  Model:    "llama-text-embed-v2",
                  FieldMap: map[string]interface{}{"text": "chunk_text"},
              },
          })
          if err != nil {
              log.Fatalf("Failed to create serverless index: %v", err)
          } else {
              fmt.Printf("Successfully created serverless index: %v", idx.Name)
          }
      }

      // Function to prettify responses
      func prettifyStruct(obj interface{}) string {
        	bytes, _ := json.MarshalIndent(obj, "", "  ")
          return string(bytes)
      }
      ```

      ```csharp C# theme={null}
      using Pinecone;

      var pinecone = new PineconeClient("{{YOUR_API_KEY}}");

      var indexName = "quickstart-dotnet";

      var createIndexRequest = await pinecone.CreateIndexForModelAsync(
          new CreateIndexForModelRequest
          {
              Name = indexName,
              Cloud = CreateIndexForModelRequestCloud.Aws,
              Region = "us-east-1",
              Embed = new CreateIndexForModelRequestEmbed
              {
                  Model = "llama-text-embed-v2",
                  FieldMap = new Dictionary<string, object?>() 
                  { 
                      { "text", "chunk_text" } 
                  }
              }
      );
      ```
    </CodeGroup>

    ### 3. Upsert text

    Prepare a sample dataset of factual statements from different domains like history, physics, technology, and music. [Model the data](/guides/index-data/data-modeling) as records with an ID, text, and category.

    <CodeGroup>
      ```python Python [expandable] theme={null}
      records = [
          { "_id": "rec1", "chunk_text": "The Eiffel Tower was completed in 1889 and stands in Paris, France.", "category": "history" },
          { "_id": "rec2", "chunk_text": "Photosynthesis allows plants to convert sunlight into energy.", "category": "science" },
          { "_id": "rec3", "chunk_text": "Albert Einstein developed the theory of relativity.", "category": "science" },
          { "_id": "rec4", "chunk_text": "The mitochondrion is often called the powerhouse of the cell.", "category": "biology" },
          { "_id": "rec5", "chunk_text": "Shakespeare wrote many famous plays, including Hamlet and Macbeth.", "category": "literature" },
          { "_id": "rec6", "chunk_text": "Water boils at 100°C under standard atmospheric pressure.", "category": "physics" },
          { "_id": "rec7", "chunk_text": "The Great Wall of China was built to protect against invasions.", "category": "history" },
          { "_id": "rec8", "chunk_text": "Honey never spoils due to its low moisture content and acidity.", "category": "food science" },
          { "_id": "rec9", "chunk_text": "The speed of light in a vacuum is approximately 299,792 km/s.", "category": "physics" },
          { "_id": "rec10", "chunk_text": "Newton's laws describe the motion of objects.", "category": "physics" },
          { "_id": "rec11", "chunk_text": "The human brain has approximately 86 billion neurons.", "category": "biology" },
          { "_id": "rec12", "chunk_text": "The Amazon Rainforest is one of the most biodiverse places on Earth.", "category": "geography" },
          { "_id": "rec13", "chunk_text": "Black holes have gravitational fields so strong that not even light can escape.", "category": "astronomy" },
          { "_id": "rec14", "chunk_text": "The periodic table organizes elements based on their atomic number.", "category": "chemistry" },
          { "_id": "rec15", "chunk_text": "Leonardo da Vinci painted the Mona Lisa.", "category": "art" },
          { "_id": "rec16", "chunk_text": "The internet revolutionized communication and information sharing.", "category": "technology" },
          { "_id": "rec17", "chunk_text": "The Pyramids of Giza are among the Seven Wonders of the Ancient World.", "category": "history" },
          { "_id": "rec18", "chunk_text": "Dogs have an incredible sense of smell, much stronger than humans.", "category": "biology" },
          { "_id": "rec19", "chunk_text": "The Pacific Ocean is the largest and deepest ocean on Earth.", "category": "geography" },
          { "_id": "rec20", "chunk_text": "Chess is a strategic game that originated in India.", "category": "games" },
          { "_id": "rec21", "chunk_text": "The Statue of Liberty was a gift from France to the United States.", "category": "history" },
          { "_id": "rec22", "chunk_text": "Coffee contains caffeine, a natural stimulant.", "category": "food science" },
          { "_id": "rec23", "chunk_text": "Thomas Edison invented the practical electric light bulb.", "category": "inventions" },
          { "_id": "rec24", "chunk_text": "The moon influences ocean tides due to gravitational pull.", "category": "astronomy" },
          { "_id": "rec25", "chunk_text": "DNA carries genetic information for all living organisms.", "category": "biology" },
          { "_id": "rec26", "chunk_text": "Rome was once the center of a vast empire.", "category": "history" },
          { "_id": "rec27", "chunk_text": "The Wright brothers pioneered human flight in 1903.", "category": "inventions" },
          { "_id": "rec28", "chunk_text": "Bananas are a good source of potassium.", "category": "nutrition" },
          { "_id": "rec29", "chunk_text": "The stock market fluctuates based on supply and demand.", "category": "economics" },
          { "_id": "rec30", "chunk_text": "A compass needle points toward the magnetic north pole.", "category": "navigation" },
          { "_id": "rec31", "chunk_text": "The universe is expanding, according to the Big Bang theory.", "category": "astronomy" },
          { "_id": "rec32", "chunk_text": "Elephants have excellent memory and strong social bonds.", "category": "biology" },
          { "_id": "rec33", "chunk_text": "The violin is a string instrument commonly used in orchestras.", "category": "music" },
          { "_id": "rec34", "chunk_text": "The heart pumps blood throughout the human body.", "category": "biology" },
          { "_id": "rec35", "chunk_text": "Ice cream melts when exposed to heat.", "category": "food science" },
          { "_id": "rec36", "chunk_text": "Solar panels convert sunlight into electricity.", "category": "technology" },
          { "_id": "rec37", "chunk_text": "The French Revolution began in 1789.", "category": "history" },
          { "_id": "rec38", "chunk_text": "The Taj Mahal is a mausoleum built by Emperor Shah Jahan.", "category": "history" },
          { "_id": "rec39", "chunk_text": "Rainbows are caused by light refracting through water droplets.", "category": "physics" },
          { "_id": "rec40", "chunk_text": "Mount Everest is the tallest mountain in the world.", "category": "geography" },
          { "_id": "rec41", "chunk_text": "Octopuses are highly intelligent marine creatures.", "category": "biology" },
          { "_id": "rec42", "chunk_text": "The speed of sound is around 343 meters per second in air.", "category": "physics" },
          { "_id": "rec43", "chunk_text": "Gravity keeps planets in orbit around the sun.", "category": "astronomy" },
          { "_id": "rec44", "chunk_text": "The Mediterranean diet is considered one of the healthiest in the world.", "category": "nutrition" },
          { "_id": "rec45", "chunk_text": "A haiku is a traditional Japanese poem with a 5-7-5 syllable structure.", "category": "literature" },
          { "_id": "rec46", "chunk_text": "The human body is made up of about 60% water.", "category": "biology" },
          { "_id": "rec47", "chunk_text": "The Industrial Revolution transformed manufacturing and transportation.", "category": "history" },
          { "_id": "rec48", "chunk_text": "Vincent van Gogh painted Starry Night.", "category": "art" },
          { "_id": "rec49", "chunk_text": "Airplanes fly due to the principles of lift and aerodynamics.", "category": "physics" },
          { "_id": "rec50", "chunk_text": "Renewable energy sources include wind, solar, and hydroelectric power.", "category": "energy" }
      ]
      ```

      ```javascript JavaScript [expandable] theme={null}
      const records = [
        { "_id": "rec1", "chunk_text": "The Eiffel Tower was completed in 1889 and stands in Paris, France.", "category": "history" },
        { "_id": "rec2", "chunk_text": "Photosynthesis allows plants to convert sunlight into energy.", "category": "science" },
        { "_id": "rec3", "chunk_text": "Albert Einstein developed the theory of relativity.", "category": "science" },
        { "_id": "rec4", "chunk_text": "The mitochondrion is often called the powerhouse of the cell.", "category": "biology" },
        { "_id": "rec5", "chunk_text": "Shakespeare wrote many famous plays, including Hamlet and Macbeth.", "category": "literature" },
        { "_id": "rec6", "chunk_text": "Water boils at 100°C under standard atmospheric pressure.", "category": "physics" },
        { "_id": "rec7", "chunk_text": "The Great Wall of China was built to protect against invasions.", "category": "history" },
        { "_id": "rec8", "chunk_text": "Honey never spoils due to its low moisture content and acidity.", "category": "food science" },
        { "_id": "rec9", "chunk_text": "The speed of light in a vacuum is approximately 299,792 km/s.", "category": "physics" },
        { "_id": "rec10", "chunk_text": "Newton's laws describe the motion of objects.", "category": "physics" },
        { "_id": "rec11", "chunk_text": "The human brain has approximately 86 billion neurons.", "category": "biology" },
        { "_id": "rec12", "chunk_text": "The Amazon Rainforest is one of the most biodiverse places on Earth.", "category": "geography" },
        { "_id": "rec13", "chunk_text": "Black holes have gravitational fields so strong that not even light can escape.", "category": "astronomy" },
        { "_id": "rec14", "chunk_text": "The periodic table organizes elements based on their atomic number.", "category": "chemistry" },
        { "_id": "rec15", "chunk_text": "Leonardo da Vinci painted the Mona Lisa.", "category": "art" },
        { "_id": "rec16", "chunk_text": "The internet revolutionized communication and information sharing.", "category": "technology" },
        { "_id": "rec17", "chunk_text": "The Pyramids of Giza are among the Seven Wonders of the Ancient World.", "category": "history" },
        { "_id": "rec18", "chunk_text": "Dogs have an incredible sense of smell, much stronger than humans.", "category": "biology" },
        { "_id": "rec19", "chunk_text": "The Pacific Ocean is the largest and deepest ocean on Earth.", "category": "geography" },
        { "_id": "rec20", "chunk_text": "Chess is a strategic game that originated in India.", "category": "games" },
        { "_id": "rec21", "chunk_text": "The Statue of Liberty was a gift from France to the United States.", "category": "history" },
        { "_id": "rec22", "chunk_text": "Coffee contains caffeine, a natural stimulant.", "category": "food science" },
        { "_id": "rec23", "chunk_text": "Thomas Edison invented the practical electric light bulb.", "category": "inventions" },
        { "_id": "rec24", "chunk_text": "The moon influences ocean tides due to gravitational pull.", "category": "astronomy" },
        { "_id": "rec25", "chunk_text": "DNA carries genetic information for all living organisms.", "category": "biology" },
        { "_id": "rec26", "chunk_text": "Rome was once the center of a vast empire.", "category": "history" },
        { "_id": "rec27", "chunk_text": "The Wright brothers pioneered human flight in 1903.", "category": "inventions" },
        { "_id": "rec28", "chunk_text": "Bananas are a good source of potassium.", "category": "nutrition" },
        { "_id": "rec29", "chunk_text": "The stock market fluctuates based on supply and demand.", "category": "economics" },
        { "_id": "rec30", "chunk_text": "A compass needle points toward the magnetic north pole.", "category": "navigation" },
        { "_id": "rec31", "chunk_text": "The universe is expanding, according to the Big Bang theory.", "category": "astronomy" },
        { "_id": "rec32", "chunk_text": "Elephants have excellent memory and strong social bonds.", "category": "biology" },
        { "_id": "rec33", "chunk_text": "The violin is a string instrument commonly used in orchestras.", "category": "music" },
        { "_id": "rec34", "chunk_text": "The heart pumps blood throughout the human body.", "category": "biology" },
        { "_id": "rec35", "chunk_text": "Ice cream melts when exposed to heat.", "category": "food science" },
        { "_id": "rec36", "chunk_text": "Solar panels convert sunlight into electricity.", "category": "technology" },
        { "_id": "rec37", "chunk_text": "The French Revolution began in 1789.", "category": "history" },
        { "_id": "rec38", "chunk_text": "The Taj Mahal is a mausoleum built by Emperor Shah Jahan.", "category": "history" },
        { "_id": "rec39", "chunk_text": "Rainbows are caused by light refracting through water droplets.", "category": "physics" },
        { "_id": "rec40", "chunk_text": "Mount Everest is the tallest mountain in the world.", "category": "geography" },
        { "_id": "rec41", "chunk_text": "Octopuses are highly intelligent marine creatures.", "category": "biology" },
        { "_id": "rec42", "chunk_text": "The speed of sound is around 343 meters per second in air.", "category": "physics" },
        { "_id": "rec43", "chunk_text": "Gravity keeps planets in orbit around the sun.", "category": "astronomy" },
        { "_id": "rec44", "chunk_text": "The Mediterranean diet is considered one of the healthiest in the world.", "category": "nutrition" },
        { "_id": "rec45", "chunk_text": "A haiku is a traditional Japanese poem with a 5-7-5 syllable structure.", "category": "literature" },
        { "_id": "rec46", "chunk_text": "The human body is made up of about 60% water.", "category": "biology" },
        { "_id": "rec47", "chunk_text": "The Industrial Revolution transformed manufacturing and transportation.", "category": "history" },
        { "_id": "rec48", "chunk_text": "Vincent van Gogh painted Starry Night.", "category": "art" },
        { "_id": "rec49", "chunk_text": "Airplanes fly due to the principles of lift and aerodynamics.", "category": "physics" },
        { "_id": "rec50", "chunk_text": "Renewable energy sources include wind, solar, and hydroelectric power.", "category": "energy" }
      ];
      ```

      ```java Java [expandable] theme={null}
      // Add to the Quickstart class:
      ArrayList<Map<String, String>> upsertRecords = new ArrayList<>();

      HashMap<String, String> record1 = new HashMap<>();
      record1.put("_id", "rec1");
      record1.put("chunk_text", "The Eiffel Tower was completed in 1889 and stands in Paris, France.");
      record1.put("category", "history");

      HashMap<String, String> record2 = new HashMap<>();
      record2.put("_id", "rec2");
      record2.put("chunk_text", "Photosynthesis allows plants to convert sunlight into energy.");
      record2.put("category", "science");

      HashMap<String, String> record3 = new HashMap<>();
      record3.put("_id", "rec3");
      record3.put("chunk_text", "Albert Einstein developed the theory of relativity.");
      record3.put("category", "science");

      HashMap<String, String> record4 = new HashMap<>();
      record4.put("_id", "rec4");
      record4.put("chunk_text", "The mitochondrion is often called the powerhouse of the cell.");
      record4.put("category", "biology");

      HashMap<String, String> record5 = new HashMap<>();
      record5.put("_id", "rec5");
      record5.put("chunk_text", "Shakespeare wrote many famous plays, including Hamlet and Macbeth.");
      record5.put("category", "literature");

      HashMap<String, String> record6 = new HashMap<>();
      record6.put("_id", "rec6");
      record6.put("chunk_text", "Water boils at 100°C under standard atmospheric pressure.");
      record6.put("category", "physics");

      HashMap<String, String> record7 = new HashMap<>();
      record7.put("_id", "rec7");
      record7.put("chunk_text", "The Great Wall of China was built to protect against invasions.");
      record7.put("category", "history");

      HashMap<String, String> record8 = new HashMap<>();
      record8.put("_id", "rec8");
      record8.put("chunk_text", "Honey never spoils due to its low moisture content and acidity.");
      record8.put("category", "food science");

      HashMap<String, String> record9 = new HashMap<>();
      record9.put("_id", "rec9");
      record9.put("chunk_text", "The speed of light in a vacuum is approximately 299,792 km/s.");
      record9.put("category", "physics");

      HashMap<String, String> record10 = new HashMap<>();
      record10.put("_id", "rec10");
      record10.put("chunk_text", "Newton's laws describe the motion of objects.");
      record10.put("category", "physics");

      HashMap<String, String> record11 = new HashMap<>();
      record11.put("_id", "rec11");
      record11.put("chunk_text", "The human brain has approximately 86 billion neurons.");
      record11.put("category", "biology");

      HashMap<String, String> record12 = new HashMap<>();
      record12.put("_id", "rec12");
      record12.put("chunk_text", "The Amazon Rainforest is one of the most biodiverse places on Earth.");
      record12.put("category", "geography");

      HashMap<String, String> record13 = new HashMap<>();
      record13.put("_id", "rec13");
      record13.put("chunk_text", "Black holes have gravitational fields so strong that not even light can escape.");
      record13.put("category", "astronomy");

      HashMap<String, String> record14 = new HashMap<>();
      record14.put("_id", "rec14");
      record14.put("chunk_text", "The periodic table organizes elements based on their atomic number.");
      record14.put("category", "chemistry");

      HashMap<String, String> record15 = new HashMap<>();
      record15.put("_id", "rec15");
      record15.put("chunk_text", "Leonardo da Vinci painted the Mona Lisa.");
      record15.put("category", "art");

      HashMap<String, String> record16 = new HashMap<>();
      record16.put("_id", "rec16");
      record16.put("chunk_text", "The internet revolutionized communication and information sharing.");
      record16.put("category", "technology");

      HashMap<String, String> record17 = new HashMap<>();
      record17.put("_id", "rec17");
      record17.put("chunk_text", "The Pyramids of Giza are among the Seven Wonders of the Ancient World.");
      record17.put("category", "history");

      HashMap<String, String> record18 = new HashMap<>();
      record18.put("_id", "rec18");
      record18.put("chunk_text", "Dogs have an incredible sense of smell, much stronger than humans.");
      record18.put("category", "biology");

      HashMap<String, String> record19 = new HashMap<>();
      record19.put("_id", "rec19");
      record19.put("chunk_text", "The Pacific Ocean is the largest and deepest ocean on Earth.");
      record19.put("category", "geography");

      HashMap<String, String> record20 = new HashMap<>();
      record20.put("_id", "rec20");
      record20.put("chunk_text", "Chess is a strategic game that originated in India.");
      record20.put("category", "games");

      HashMap<String, String> record21 = new HashMap<>();
      record21.put("_id", "rec21");
      record21.put("chunk_text", "The Statue of Liberty was a gift from France to the United States.");
      record21.put("category", "history");

      HashMap<String, String> record22 = new HashMap<>();
      record22.put("_id", "rec22");
      record22.put("chunk_text", "Coffee contains caffeine, a natural stimulant.");
      record22.put("category", "food science");

      HashMap<String, String> record23 = new HashMap<>();
      record23.put("_id", "rec23");
      record23.put("chunk_text", "Thomas Edison invented the practical electric light bulb.");
      record23.put("category", "inventions");

      HashMap<String, String> record24 = new HashMap<>();
      record24.put("_id", "rec24");
      record24.put("chunk_text", "The moon influences ocean tides due to gravitational pull.");
      record24.put("category", "astronomy");

      HashMap<String, String> record25 = new HashMap<>();
      record25.put("_id", "rec25");
      record25.put("chunk_text", "DNA carries genetic information for all living organisms.");
      record25.put("category", "biology");

      HashMap<String, String> record26 = new HashMap<>();
      record26.put("_id", "rec26");
      record26.put("chunk_text", "Rome was once the center of a vast empire.");
      record26.put("category", "history");

      HashMap<String, String> record27 = new HashMap<>();
      record27.put("_id", "rec27");
      record27.put("chunk_text", "The Wright brothers pioneered human flight in 1903.");
      record27.put("category", "inventions");

      HashMap<String, String> record28 = new HashMap<>();
      record28.put("_id", "rec28");
      record28.put("chunk_text", "Bananas are a good source of potassium.");
      record28.put("category", "nutrition");

      HashMap<String, String> record29 = new HashMap<>();
      record29.put("_id", "rec29");
      record29.put("chunk_text", "The stock market fluctuates based on supply and demand.");
      record29.put("category", "economics");

      HashMap<String, String> record30 = new HashMap<>();
      record30.put("_id", "rec30");
      record30.put("chunk_text", "A compass needle points toward the magnetic north pole.");
      record30.put("category", "navigation");

      HashMap<String, String> record31 = new HashMap<>();
      record31.put("_id", "rec31");
      record31.put("chunk_text", "The universe is expanding, according to the Big Bang theory.");
      record31.put("category", "astronomy");

      HashMap<String, String> record32 = new HashMap<>();
      record32.put("_id", "rec32");
      record32.put("chunk_text", "Elephants have excellent memory and strong social bonds.");
      record32.put("category", "biology");

      HashMap<String, String> record33 = new HashMap<>();
      record33.put("_id", "rec33");
      record33.put("chunk_text", "The violin is a string instrument commonly used in orchestras.");
      record33.put("category", "music");

      HashMap<String, String> record34 = new HashMap<>();
      record34.put("_id", "rec34");
      record34.put("chunk_text", "The heart pumps blood throughout the human body.");
      record34.put("category", "biology");

      HashMap<String, String> record35 = new HashMap<>();
      record35.put("_id", "rec35");
      record35.put("chunk_text", "Ice cream melts when exposed to heat.");
      record35.put("category", "food science");

      HashMap<String, String> record36 = new HashMap<>();
      record36.put("_id", "rec36");
      record36.put("chunk_text", "Solar panels convert sunlight into electricity.");
      record36.put("category", "technology");

      HashMap<String, String> record37 = new HashMap<>();
      record37.put("_id", "rec37");
      record37.put("chunk_text", "The French Revolution began in 1789.");
      record37.put("category", "history");

      HashMap<String, String> record38 = new HashMap<>();
      record38.put("_id", "rec38");
      record38.put("chunk_text", "The Taj Mahal is a mausoleum built by Emperor Shah Jahan.");
      record38.put("category", "history");

      HashMap<String, String> record39 = new HashMap<>();
      record39.put("_id", "rec39");
      record39.put("chunk_text", "Rainbows are caused by light refracting through water droplets.");
      record39.put("category", "physics");

      HashMap<String, String> record40 = new HashMap<>();
      record40.put("_id", "rec40");
      record40.put("chunk_text", "Mount Everest is the tallest mountain in the world.");
      record40.put("category", "geography");

      HashMap<String, String> record41 = new HashMap<>();
      record41.put("_id", "rec41");
      record41.put("chunk_text", "Octopuses are highly intelligent marine creatures.");
      record41.put("category", "biology");

      HashMap<String, String> record42 = new HashMap<>();
      record42.put("_id", "rec42");
      record42.put("chunk_text", "The speed of sound is around 343 meters per second in air.");
      record42.put("category", "physics");

      HashMap<String, String> record43 = new HashMap<>();
      record43.put("_id", "rec43");
      record43.put("chunk_text", "Gravity keeps planets in orbit around the sun.");
      record43.put("category", "astronomy");

      HashMap<String, String> record44 = new HashMap<>();
      record44.put("_id", "rec44");
      record44.put("chunk_text", "The Mediterranean diet is considered one of the healthiest in the world.");
      record44.put("category", "nutrition");

      HashMap<String, String> record45 = new HashMap<>();
      record45.put("_id", "rec45");
      record45.put("chunk_text", "A haiku is a traditional Japanese poem with a 5-7-5 syllable structure.");
      record45.put("category", "literature");

      HashMap<String, String> record46 = new HashMap<>();
      record46.put("_id", "rec46");
      record46.put("chunk_text", "The human body is made up of about 60% water.");
      record46.put("category", "biology");

      HashMap<String, String> record47 = new HashMap<>();
      record47.put("_id", "rec47");
      record47.put("chunk_text", "The Industrial Revolution transformed manufacturing and transportation.");
      record47.put("category", "history");

      HashMap<String, String> record48 = new HashMap<>();
      record48.put("_id", "rec48");
      record48.put("chunk_text", "Vincent van Gogh painted Starry Night.");
      record48.put("category", "art");

      HashMap<String, String> record49 = new HashMap<>();
      record49.put("_id", "rec49");
      record49.put("chunk_text", "Airplanes fly due to the principles of lift and aerodynamics.");
      record49.put("category", "physics");

      HashMap<String, String> record50 = new HashMap<>();
      record50.put("_id", "rec50");
      record50.put("chunk_text", "Renewable energy sources include wind, solar, and hydroelectric power.");
      record50.put("category", "energy");

      upsertRecords.add(record1);
      upsertRecords.add(record2);
      upsertRecords.add(record3);
      upsertRecords.add(record4);
      upsertRecords.add(record5);
      upsertRecords.add(record6);
      upsertRecords.add(record7);
      upsertRecords.add(record8);
      upsertRecords.add(record9);
      upsertRecords.add(record10);
      upsertRecords.add(record11);
      upsertRecords.add(record12);
      upsertRecords.add(record13);
      upsertRecords.add(record14);
      upsertRecords.add(record15);
      upsertRecords.add(record16);
      upsertRecords.add(record17);
      upsertRecords.add(record18);
      upsertRecords.add(record19);
      upsertRecords.add(record20);
      upsertRecords.add(record21);
      upsertRecords.add(record22);
      upsertRecords.add(record23);
      upsertRecords.add(record24);
      upsertRecords.add(record25);
      upsertRecords.add(record26);
      upsertRecords.add(record27);
      upsertRecords.add(record28);
      upsertRecords.add(record29);
      upsertRecords.add(record30);
      upsertRecords.add(record31);
      upsertRecords.add(record32);
      upsertRecords.add(record33);
      upsertRecords.add(record34);
      upsertRecords.add(record35);
      upsertRecords.add(record36);
      upsertRecords.add(record37);
      upsertRecords.add(record38);
      upsertRecords.add(record39);
      upsertRecords.add(record40);
      upsertRecords.add(record41);
      upsertRecords.add(record42);
      upsertRecords.add(record43);
      upsertRecords.add(record44);
      upsertRecords.add(record45);
      upsertRecords.add(record46);
      upsertRecords.add(record47);
      upsertRecords.add(record48);
      upsertRecords.add(record49);
      upsertRecords.add(record50);
      ```

      ```go Go [expandable] theme={null}
      // Add to the main function:
      records := []*pinecone.IntegratedRecord{
          { "_id": "rec1", "chunk_text": "The Eiffel Tower was completed in 1889 and stands in Paris, France.", "category": "history" },
          { "_id": "rec2", "chunk_text": "Photosynthesis allows plants to convert sunlight into energy.", "category": "science" },
          { "_id": "rec3", "chunk_text": "Albert Einstein developed the theory of relativity.", "category": "science" },
          { "_id": "rec4", "chunk_text": "The mitochondrion is often called the powerhouse of the cell.", "category": "biology" },
          { "_id": "rec5", "chunk_text": "Shakespeare wrote many famous plays, including Hamlet and Macbeth.", "category": "literature" },
          { "_id": "rec6", "chunk_text": "Water boils at 100°C under standard atmospheric pressure.", "category": "physics" },
          { "_id": "rec7", "chunk_text": "The Great Wall of China was built to protect against invasions.", "category": "history" },
          { "_id": "rec8", "chunk_text": "Honey never spoils due to its low moisture content and acidity.", "category": "food science" },
          { "_id": "rec9", "chunk_text": "The speed of light in a vacuum is approximately 299,792 km/s.", "category": "physics" },
          { "_id": "rec10", "chunk_text": "Newton's laws describe the motion of objects.", "category": "physics" },
          { "_id": "rec11", "chunk_text": "The human brain has approximately 86 billion neurons.", "category": "biology" },
          { "_id": "rec12", "chunk_text": "The Amazon Rainforest is one of the most biodiverse places on Earth.", "category": "geography" },
          { "_id": "rec13", "chunk_text": "Black holes have gravitational fields so strong that not even light can escape.", "category": "astronomy" },
          { "_id": "rec14", "chunk_text": "The periodic table organizes elements based on their atomic number.", "category": "chemistry" },
          { "_id": "rec15", "chunk_text": "Leonardo da Vinci painted the Mona Lisa.", "category": "art" },
          { "_id": "rec16", "chunk_text": "The internet revolutionized communication and information sharing.", "category": "technology" },
          { "_id": "rec17", "chunk_text": "The Pyramids of Giza are among the Seven Wonders of the Ancient World.", "category": "history" },
          { "_id": "rec18", "chunk_text": "Dogs have an incredible sense of smell, much stronger than humans.", "category": "biology" },
          { "_id": "rec19", "chunk_text": "The Pacific Ocean is the largest and deepest ocean on Earth.", "category": "geography" },
          { "_id": "rec20", "chunk_text": "Chess is a strategic game that originated in India.", "category": "games" },
          { "_id": "rec21", "chunk_text": "The Statue of Liberty was a gift from France to the United States.", "category": "history" },
          { "_id": "rec22", "chunk_text": "Coffee contains caffeine, a natural stimulant.", "category": "food science" },
          { "_id": "rec23", "chunk_text": "Thomas Edison invented the practical electric light bulb.", "category": "inventions" },
          { "_id": "rec24", "chunk_text": "The moon influences ocean tides due to gravitational pull.", "category": "astronomy" },
          { "_id": "rec25", "chunk_text": "DNA carries genetic information for all living organisms.", "category": "biology" },
          { "_id": "rec26", "chunk_text": "Rome was once the center of a vast empire.", "category": "history" },
          { "_id": "rec27", "chunk_text": "The Wright brothers pioneered human flight in 1903.", "category": "inventions" },
          { "_id": "rec28", "chunk_text": "Bananas are a good source of potassium.", "category": "nutrition" },
          { "_id": "rec29", "chunk_text": "The stock market fluctuates based on supply and demand.", "category": "economics" },
          { "_id": "rec30", "chunk_text": "A compass needle points toward the magnetic north pole.", "category": "navigation" },
          { "_id": "rec31", "chunk_text": "The universe is expanding, according to the Big Bang theory.", "category": "astronomy" },
          { "_id": "rec32", "chunk_text": "Elephants have excellent memory and strong social bonds.", "category": "biology" },
          { "_id": "rec33", "chunk_text": "The violin is a string instrument commonly used in orchestras.", "category": "music" },
          { "_id": "rec34", "chunk_text": "The heart pumps blood throughout the human body.", "category": "biology" },
          { "_id": "rec35", "chunk_text": "Ice cream melts when exposed to heat.", "category": "food science" },
          { "_id": "rec36", "chunk_text": "Solar panels convert sunlight into electricity.", "category": "technology" },
          { "_id": "rec37", "chunk_text": "The French Revolution began in 1789.", "category": "history" },
          { "_id": "rec38", "chunk_text": "The Taj Mahal is a mausoleum built by Emperor Shah Jahan.", "category": "history" },
          { "_id": "rec39", "chunk_text": "Rainbows are caused by light refracting through water droplets.", "category": "physics" },
          { "_id": "rec40", "chunk_text": "Mount Everest is the tallest mountain in the world.", "category": "geography" },
          { "_id": "rec41", "chunk_text": "Octopuses are highly intelligent marine creatures.", "category": "biology" },
          { "_id": "rec42", "chunk_text": "The speed of sound is around 343 meters per second in air.", "category": "physics" },
          { "_id": "rec43", "chunk_text": "Gravity keeps planets in orbit around the sun.", "category": "astronomy" },
          { "_id": "rec44", "chunk_text": "The Mediterranean diet is considered one of the healthiest in the world.", "category": "nutrition" },
          { "_id": "rec45", "chunk_text": "A haiku is a traditional Japanese poem with a 5-7-5 syllable structure.", "category": "literature" },
          { "_id": "rec46", "chunk_text": "The human body is made up of about 60% water.", "category": "biology" },
          { "_id": "rec47", "chunk_text": "The Industrial Revolution transformed manufacturing and transportation.", "category": "history" },
          { "_id": "rec48", "chunk_text": "Vincent van Gogh painted Starry Night.", "category": "art" },
          { "_id": "rec49", "chunk_text": "Airplanes fly due to the principles of lift and aerodynamics.", "category": "physics" },
          { "_id": "rec50", "chunk_text": "Renewable energy sources include wind, solar, and hydroelectric power.", "category": "energy" },
      }
      ```

      ```csharp C# [expandable] theme={null}
      var records = new List<UpsertRecord>
      {
          new UpsertRecord
          {
              Id = "rec1",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The Eiffel Tower was completed in 1889 and stands in Paris, France.",
                    ["category"] = "history",
                },
            },
            new UpsertRecord
            {
                Id = "rec2",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Photosynthesis allows plants to convert sunlight into energy.",
                    ["category"] = "science",
                },
            },  
            new UpsertRecord
            {
                Id = "rec3",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Albert Einstein developed the theory of relativity.",
                    ["category"] = "science",
                },
            },
            new UpsertRecord
            {
                Id = "rec4",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The mitochondrion is often called the powerhouse of the cell.",
                    ["category"] = "biology",
                },
            },
            new UpsertRecord
            {
                Id = "rec5",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Shakespeare wrote many famous plays, including Hamlet and Macbeth.",
                    ["category"] = "literature",
                },
            },
            new UpsertRecord
            {
                Id = "rec6",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Water boils at 100°C under standard atmospheric pressure.",
                    ["category"] = "physics",
                },
            },
            new UpsertRecord
            {
                Id = "rec7",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The Great Wall of China was built to protect against invasions.",
                    ["category"] = "history",
                },
            },
            new UpsertRecord
            {
                Id = "rec8",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Honey never spoils due to its low moisture content and acidity.",
                    ["category"] = "food science",
                },
            },
            new UpsertRecord
            {
                Id = "rec9",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The speed of light in a vacuum is approximately 299,792 km/s.",
                    ["category"] = "physics",
                },
            },
            new UpsertRecord
            {
                Id = "rec10",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Newton's laws describe the motion of objects.",
                    ["category"] = "physics",
                },
            },
            new UpsertRecord
            {
                Id = "rec11",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The human brain has approximately 86 billion neurons.",
                    ["category"] = "biology",
                },
            },
            new UpsertRecord
            {
                Id = "rec12",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The Amazon Rainforest is one of the most biodiverse places on Earth.",
                    ["category"] = "geography",
                },
            },
            new UpsertRecord
            {
                Id = "rec13",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Black holes have gravitational fields so strong that not even light can escape.",
                    ["category"] = "astronomy",
                },
            },
            new UpsertRecord
            {
                Id = "rec14",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The periodic table organizes elements based on their atomic number.",
                    ["category"] = "chemistry",
                },
            },
            new UpsertRecord
            {
                Id = "rec15",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Leonardo da Vinci painted the Mona Lisa.",
                    ["category"] = "art",
                },
            },
            new UpsertRecord
            {
                Id = "rec16",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The internet revolutionized communication and information sharing.",
                    ["category"] = "technology",
                },
            },
            new UpsertRecord
            {
                Id = "rec17",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The Pyramids of Giza are among the Seven Wonders of the Ancient World.",
                    ["category"] = "history",
                },
            },
            new UpsertRecord
            {
                Id = "rec18",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Dogs have an incredible sense of smell, much stronger than humans.",
                    ["category"] = "biology",
                },
            },
            new UpsertRecord
            {
                Id = "rec19",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The Pacific Ocean is the largest and deepest ocean on Earth.",
                    ["category"] = "geography",
                },
            },
            new UpsertRecord
            {
                Id = "rec20",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Chess is a strategic game that originated in India.",
                    ["category"] = "games",
                },
            },
            new UpsertRecord
            {
                Id = "rec21",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The Statue of Liberty was a gift from France to the United States.",
                    ["category"] = "history",
                },
            },
            new UpsertRecord
            {
                Id = "rec22",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Coffee contains caffeine, a natural stimulant.",
                    ["category"] = "food science",
                },
            },
            new UpsertRecord
            {
                Id = "rec23",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Thomas Edison invented the practical electric light bulb.",
                    ["category"] = "inventions",
                },
            },
            new UpsertRecord
            {
                Id = "rec24",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The moon influences ocean tides due to gravitational pull.",
                    ["category"] = "astronomy",
                },
            },
            new UpsertRecord
            {
                Id = "rec25",
                AdditionalProperties =
                {
                    ["chunk_text"] = "DNA carries genetic information for all living organisms.",
                    ["category"] = "biology",
                },
            },
            new UpsertRecord
            {
                Id = "rec26",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Rome was once the center of a vast empire.",
                    ["category"] = "history",
                },
            },
            new UpsertRecord
            {
                Id = "rec27",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The Wright brothers pioneered human flight in 1903.",
                    ["category"] = "inventions",
                },
            },
            new UpsertRecord
            {
                Id = "rec28",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Bananas are a good source of potassium.",
                    ["category"] = "nutrition",
                },
            },
            new UpsertRecord
            {
                Id = "rec29",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The stock market fluctuates based on supply and demand.",
                    ["category"] = "economics",
                },
            },
            new UpsertRecord
            {
                Id = "rec30",
                AdditionalProperties =
                {
                    ["chunk_text"] = "A compass needle points toward the magnetic north pole.",
                    ["category"] = "navigation",
                },
            },
            new UpsertRecord
            {
                Id = "rec31",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The universe is expanding, according to the Big Bang theory.",
                    ["category"] = "astronomy",
                },
            },
            new UpsertRecord
            {
                Id = "rec32",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Elephants have excellent memory and strong social bonds.",
                    ["category"] = "biology",
                },
            },
            new UpsertRecord
            {
                Id = "rec33",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The violin is a string instrument commonly used in orchestras.",
                    ["category"] = "music",
                },
            },
            new UpsertRecord
            {
                Id = "rec34",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The heart pumps blood throughout the human body.",
                    ["category"] = "biology",
                },
            },
            new UpsertRecord
            {
                Id = "rec35",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Ice cream melts when exposed to heat.",
                    ["category"] = "food science",
                },
            },
            new UpsertRecord
            {
                Id = "rec36",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Solar panels convert sunlight into electricity.",
                    ["category"] = "technology",
                },
            },
            new UpsertRecord
            {
                Id = "rec37",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The French Revolution began in 1789.",
                    ["category"] = "history",
                },
            },
            new UpsertRecord
            {
                Id = "rec38",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The Taj Mahal is a mausoleum built by Emperor Shah Jahan.",
                    ["category"] = "history",
                },
            },
            new UpsertRecord
            {
                Id = "rec39",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Rainbows are caused by light refracting through water droplets.",
                    ["category"] = "physics",
                },
            },
            new UpsertRecord
            {
                Id = "rec40",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Mount Everest is the tallest mountain in the world.",
                    ["category"] = "geography",
                },
            },
            new UpsertRecord
            {
                Id = "rec41",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Octopuses are highly intelligent marine creatures.",
                    ["category"] = "biology",
                },
            },
            new UpsertRecord
            {
                Id = "rec42",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The speed of sound is around 343 meters per second in air.",
                    ["category"] = "physics",
                },
            },
            new UpsertRecord
            {
                Id = "rec43",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Gravity keeps planets in orbit around the sun.",
                    ["category"] = "astronomy",
                },
            },
            new UpsertRecord
            {
                Id = "rec44",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The Mediterranean diet is considered one of the healthiest in the world.",
                    ["category"] = "nutrition",
                },
            },
            new UpsertRecord
            {
                Id = "rec45",
                AdditionalProperties =
                {
                    ["chunk_text"] = "A haiku is a traditional Japanese poem with a 5-7-5 syllable structure.",
                    ["category"] = "literature",
                },
            },
            new UpsertRecord
            {
                Id = "rec46",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The human body is made up of about 60% water.",
                    ["category"] = "biology",
                },
            },
            new UpsertRecord
            {
                Id = "rec47",
                AdditionalProperties =
                {
                    ["chunk_text"] = "The Industrial Revolution transformed manufacturing and transportation.",
                    ["category"] = "history",
                },
            },
            new UpsertRecord
            {
                Id = "rec48",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Vincent van Gogh painted Starry Night.",
                    ["category"] = "art",
                },
            },
            new UpsertRecord
            {
                Id = "rec49",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Airplanes fly due to the principles of lift and aerodynamics.",
                    ["category"] = "physics",
                },
            },
            new UpsertRecord
            {
                Id = "rec50",
                AdditionalProperties =
                {
                    ["chunk_text"] = "Renewable energy sources include wind, solar, and hydroelectric power.",
                    ["category"] = "energy",
                },
            },
      };
      ```
    </CodeGroup>

    [Upsert](/guides/index-data/upsert-data) the sample dataset into a new [namespace](/guides/index-data/indexing-overview#namespaces) in your index.

    Because your index is integrated with an embedding model, you provide the textual statements and Pinecone converts them to dense vectors automatically.

    <CodeGroup>
      ```python Python theme={null}
      # Target the index
      dense_index = pc.Index(index_name)

      # Upsert the records into a namespace
      dense_index.upsert_records("example-namespace", records)
      ```

      ```javascript JavaScript theme={null}
      // Target the index
      const index = pc.index(indexName).namespace("example-namespace");

      // Upsert the records into a namespace
      await index.upsertRecords(records);
      ```

      ```java Java theme={null}
      // Add to the Quickstart class:
      // Target the index
      Index index = new Index(config, connection, "quickstart-java");
      // Upsert the records into a namespace
      index.upsertRecords("example-namespace", upsertRecords);
      ```

      ```go Go theme={null}
      // Add to the main function:
      // Target the index
      idxModel, err := pc.DescribeIndex(ctx, indexName)
      if err != nil {
          log.Fatalf("Failed to describe index \"%v\": %v", indexName, err)
      }

      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: idxModel.Host, Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v: %v", idxModel.Host, err)
      }

      // Upsert the records into a namespace
      err = idxConnection.UpsertRecords(ctx, records)
      if err != nil {
          log.Fatalf("Failed to upsert vectors: %v", err)
      }
      ```

      ```csharp C# theme={null}
      // Upsert the records into a namespace
      await index.UpsertRecordsAsync(
          "example-namespace",
          records
      );
      ```
    </CodeGroup>

    <Tip>
      To control costs when ingesting large datasets (10,000,000+ records), use [import](/guides/index-data/import-data) instead of upsert.
    </Tip>

    Pinecone is eventually consistent, so there can be a slight delay before new or changed records are visible to queries. You can [view index stats](/guides/index-data/check-data-freshness) to check if the current vector count matches the number of vectors you upserted (50):

    <CodeGroup>
      ```python Python theme={null}
      # Wait for the upserted vectors to be indexed
      import time
      time.sleep(10)

      # View stats for the index
      stats = dense_index.describe_index_stats()
      print(stats)
      ```

      ```javascript JavaScript theme={null}
      // Wait for the upserted vectors to be indexed
      await new Promise(resolve => setTimeout(resolve, 10000));

      // View stats for the index
      const stats = await index.describeIndexStats();
      console.log(stats);
      ```

      ```java Java theme={null}
      // Add to the Quickstart class:
      // Wait for upserted vectors to be indexed
      Thread.sleep(5000);

      // View stats for the index
      DescribeIndexStatsResponse indexStatsResponse = index.describeIndexStats();
      System.out.println(indexStatsResponse);
      ```

      ```go Go theme={null}
      // Add to the main function:
      // View stats for the index
      stats, err := idxConnection.DescribeIndexStats(ctx)
      if err != nil {
          log.Fatalf("Failed to describe index \"%v\": %v", indexName, err)
      } else {
          fmt.Printf("%+v", prettifyStruct(*stats))
      }
      ```

      ```csharp C# theme={null}
      var indexStatsResponse = await index.DescribeIndexStatsAsync(new DescribeIndexStatsRequest());

      Console.WriteLine(indexStatsResponse);
      ```
    </CodeGroup>

    The response looks like this:

    <CodeGroup>
      ```python Python theme={null}
      {'dimension': 1024,
       'index_fullness': 0.0,
       'metric': 'cosine',
       'namespaces': {'example-namespace': {'vector_count': 50}},
       'total_vector_count': 50,
       'vector_type': 'dense'}
      ```

      ```javascript JavaScript theme={null}
      {
        namespaces: { 'example-namespace': { recordCount: 50 } },
        dimension: 1024,
        indexFullness: 0,
        totalRecordCount: 50
      }
      ```

      ```java Java theme={null}
      namespaces {
        key: "example-namespace"
        value {
          vector_count: 50
        }
      }
      dimension: 1024
      total_vector_count: 50
      metric: "cosine"
      vector_type: "dense"
      ```

      ```go Go theme={null}
      {
        "dimension": 1024,
        "index_fullness": 0,
        "total_vector_count": 50,
        "namespaces": {
          "example-namespace": {
            "vector_count": 50
          }
        }
      }
      ```

      ```csharp C# theme={null}
      {
        "namespaces": {
          "example-namespace": {
            "vectorCount": 50
          }
        },
        "dimension": 1024,
        "indexFullness": 0,
        "totalVectorCount": 50,
        "metric": "cosine",
        "vectorType": "dense"
      }
      ```
    </CodeGroup>

    ### 4. Semantic search

    [Search the dense index](/guides/search/semantic-search) for ten records that are most semantically similar to the query, "Famous historical structures and monuments".

    Again, because your index is integrated with an embedding model, you provide the query as text and Pinecone converts the text to a dense vector automatically.

    <CodeGroup>
      ```python Python theme={null}
      # Define the query
      query = "Famous historical structures and monuments"

      # Search the dense index
      results = dense_index.search(
          namespace="example-namespace",
          query={
              "top_k": 10,
              "inputs": {
                  'text': query
              }
          }
      )

      # Print the results
      for hit in results['result']['hits']:
              print(f"id: {hit['_id']:<5} | score: {round(hit['_score'], 2):<5} | category: {hit['fields']['category']:<10} | text: {hit['fields']['chunk_text']:<50}")
      ```

      ```javascript JavaScript theme={null}
      // Define the query
      const query = 'Famous historical structures and monuments';

      // Search the dense index
      const results = await index.searchRecords({
        query: {
          topK: 10,
          inputs: { text: query },
        },
      });

      // Print the results
      results.result.hits.forEach(hit => {
        console.log(`id: ${hit.id}, score: ${hit.score.toFixed(2)}, category: ${hit.fields.category}, text: ${hit.fields.chunk_text}`);
      });
      ```

      ```java Java theme={null}
      // Add to the Quickstart class:
      // Define the query
      String query = "Famous historical structures and monuments";
      List<String> fields = new ArrayList<>();
      fields.add("category");
      fields.add("chunk_text");

      // Search the dense index
      SearchRecordsResponse recordsResponse = index.searchRecordsByText(query,  "example-namespace", fields, 10, null, null);

      // Print the results
      System.out.println(recordsResponse);
      ```

      ```go Go theme={null}
      // Add to the main function:
      // Define the query
      query := "Famous historical structures and monuments"

      // Search the dense index
      res, err := idxConnection.SearchRecords(ctx, &pinecone.SearchRecordsRequest{
          Query: pinecone.SearchRecordsQuery{
              TopK: 10,
              Inputs: &map[string]interface{}{
                  "text": query,
              },
          },
      })
      if err != nil {
          log.Fatalf("Failed to search records: %v", err)
      }
      fmt.Printf(prettifyStruct(res))
      ```

      ```csharp C# theme={null}
      // Search the dense index
      var response = await index.SearchRecordsAsync(
          "example-namespace",
          new SearchRecordsRequest
          {
              Query = new SearchRecordsRequestQuery
              {
                  TopK = 10,
                  Inputs = new Dictionary<string, object?> { { "text", "Famous historical structures and monuments" } },
              },
              Fields = ["category", "chunk_text"],
          }
      );

      Console.WriteLine(response);
      ```
    </CodeGroup>

    Notice that most of the results are about historical structures and monuments. However, a few unrelated statements are included as well and are ranked high in the list, for example, a statement about Shakespeare.

    <CodeGroup>
      ```console Python theme={null}
      id: rec17 | score: 0.24  | category: history    | text: The Pyramids of Giza are among the Seven Wonders of the Ancient World.
      id: rec38 | score: 0.19  | category: history    | text: The Taj Mahal is a mausoleum built by Emperor Shah Jahan.
      id: rec5  | score: 0.19  | category: literature | text: Shakespeare wrote many famous plays, including Hamlet and Macbeth.
      id: rec15 | score: 0.11  | category: art        | text: Leonardo da Vinci painted the Mona Lisa.          
      id: rec50 | score: 0.1   | category: energy     | text: Renewable energy sources include wind, solar, and hydroelectric power.
      id: rec26 | score: 0.09  | category: history    | text: Rome was once the center of a vast empire.        
      id: rec47 | score: 0.08  | category: history    | text: The Industrial Revolution transformed manufacturing and transportation.
      id: rec7  | score: 0.07  | category: history    | text: The Great Wall of China was built to protect against invasions.
      id: rec1  | score: 0.07  | category: history    | text: The Eiffel Tower was completed in 1889 and stands in Paris, France.
      id: rec3  | score: 0.07  | category: science    | text: Albert Einstein developed the theory of relativity.
      ```

      ```console JavaScript theme={null}
      id: rec17, score: 0.24, text: The Pyramids of Giza are among the Seven Wonders of the Ancient World., category: history
      id: rec38, score: 0.19, text: The Taj Mahal is a mausoleum built by Emperor Shah Jahan., category: history
      id: rec5, score: 0.19, text: Shakespeare wrote many famous plays, including Hamlet and Macbeth., category: literature
      id: rec15, score: 0.11, text: Leonardo da Vinci painted the Mona Lisa., category: art
      id: rec50, score: 0.10, text: Renewable energy sources include wind, solar, and hydroelectric power., category: energy
      id: rec26, score: 0.09, text: Rome was once the center of a vast empire., category: history
      id: rec47, score: 0.08, text: The Industrial Revolution transformed manufacturing and transportation., category: history
      id: rec7, score: 0.07, text: The Great Wall of China was built to protect against invasions., category: history
      id: rec1, score: 0.07, text: The Eiffel Tower was completed in 1889 and stands in Paris, France., category: history
      id: rec3, score: 0.07, text: Albert Einstein developed the theory of relativity., category: science
      ```

      ```java Java [expandable] theme={null}
      class SearchRecordsResponse {
          result: class SearchRecordsResponseResult {
              hits: [class Hit {
                  id: rec17
                  score: 0.77387625
                  fields: {category=history, chunk_text=The Pyramids of Giza are among the Seven Wonders of the Ancient World.}
                  additionalProperties: null
              }, class Hit {
                  id: rec1
                  score: 0.77372295
                  fields: {category=history, chunk_text=The Eiffel Tower was completed in 1889 and stands in Paris, France.}
                  additionalProperties: null
              }, class Hit {
                  id: rec38
                  score: 0.75988203
                  fields: {category=history, chunk_text=The Taj Mahal is a mausoleum built by Emperor Shah Jahan.}
                  additionalProperties: null
              }, class Hit {
                  id: rec5
                  score: 0.75516135
                  fields: {category=literature, chunk_text=Shakespeare wrote many famous plays, including Hamlet and Macbeth.}
                  additionalProperties: null
              }, class Hit {
                  id: rec26
                  score: 0.7550185
                  fields: {category=history, chunk_text=Rome was once the center of a vast empire.}
                  additionalProperties: null
              }, class Hit {
                  id: rec45
                  score: 0.73588645
                  fields: {category=literature, chunk_text=A haiku is a traditional Japanese poem with a 5-7-5 syllable structure.}
                  additionalProperties: null
              }, class Hit {
                  id: rec4
                  score: 0.730563
                  fields: {category=biology, chunk_text=The mitochondrion is often called the powerhouse of the cell.}
                  additionalProperties: null
              }, class Hit {
                  id: rec7
                  score: 0.73037535
                  fields: {category=history, chunk_text=The Great Wall of China was built to protect against invasions.}
                  additionalProperties: null
              }, class Hit {
                  id: rec32
                  score: 0.72860974
                  fields: {category=biology, chunk_text=Elephants have excellent memory and strong social bonds.}
                  additionalProperties: null
              }, class Hit {
                  id: rec47
                  score: 0.7285921
                  fields: {category=history, chunk_text=The Industrial Revolution transformed manufacturing and transportation.}
                  additionalProperties: null
              }]
              additionalProperties: null
          }
          usage: class SearchUsage {
              readUnits: 6
              embedTotalTokens: 13
              rerankUnits: null
              additionalProperties: null
          }
          additionalProperties: null
      }
      ```

      ```json Go [expandable] theme={null}
      {
        "result": {
          "hits": [
            {
              "_id": "rec17",
              "_score": 0.24442708,
              "fields": {
                "category": "history",
                "chunk_text": "The Pyramids of Giza are among the Seven Wonders of the Ancient World."
              }
            },
            {
              "_id": "rec38",
              "_score": 0.1876694,
              "fields": {
                "category": "history",
                "chunk_text": "The Taj Mahal is a mausoleum built by Emperor Shah Jahan."
              }
            },
            {
              "_id": "rec5",
              "_score": 0.18504046,
              "fields": {
                "category": "literature",
                "chunk_text": "Shakespeare wrote many famous plays, including Hamlet and Macbeth."
              }
            },
            {
              "_id": "rec15",
              "_score": 0.109251045,
              "fields": {
                "category": "art",
                "chunk_text": "Leonardo da Vinci painted the Mona Lisa."
              }
            },
            {
              "_id": "rec50",
              "_score": 0.098952696,
              "fields": {
                "category": "energy",
                "chunk_text": "Renewable energy sources include wind, solar, and hydroelectric power."
              }
            },
            {
              "_id": "rec26",
              "_score": 0.085251465,
              "fields": {
                "category": "history",
                "chunk_text": "Rome was once the center of a vast empire."
              }
            },
            {
              "_id": "rec47",
              "_score": 0.07533597,
              "fields": {
                "category": "history",
                "chunk_text": "The Industrial Revolution transformed manufacturing and transportation."
              }
            },
            {
              "_id": "rec7",
              "_score": 0.06859385,
              "fields": {
                "category": "history",
                "chunk_text": "The Great Wall of China was built to protect against invasions."
              }
            },
            {
              "_id": "rec1",
              "_score": 0.06831257,
              "fields": {
                "category": "history",
                "chunk_text": "The Eiffel Tower was completed in 1889 and stands in Paris, France."
              }
            },
            {
              "_id": "rec3",
              "_score": 0.06689669,
              "fields": {
                "category": "science",
                "chunk_text": "Albert Einstein developed the theory of relativity."
              }
            }
          ]
        },
        "usage": {
          "read_units": 6,
          "embed_total_tokens": 8
        }
      }
      ```

      ```csharp C# [expandable] theme={null}
      {
        "result": {
          "hits": [
            {
              "_id": "rec17",
              "_score": 0.27985704,
              "fields": {
                "category": "history",
                "chunk_text": "The Pyramids of Giza are among the Seven Wonders of the Ancient World."
              }
            },
            {
              "_id": "rec38",
              "_score": 0.18836586,
              "fields": {
                "category": "history",
                "chunk_text": "The Taj Mahal is a mausoleum built by Emperor Shah Jahan."
              }
            },
            {
              "_id": "rec5",
              "_score": 0.18140909,
              "fields": {
                "category": "literature",
                "chunk_text": "Shakespeare wrote many famous plays, including Hamlet and Macbeth."
              }
            },
            {
              "_id": "rec15",
              "_score": 0.09603156,
              "fields": {
                "category": "art",
                "chunk_text": "Leonardo da Vinci painted the Mona Lisa."
              }
            },
            {
              "_id": "rec50",
              "_score": 0.091406636,
              "fields": {
                "category": "energy",
                "chunk_text": "Renewable energy sources include wind, solar, and hydroelectric power."
              }
            },
            {
              "_id": "rec1",
              "_score": 0.0828001,
              "fields": {
                "category": "history",
                "chunk_text": "The Eiffel Tower was completed in 1889 and stands in Paris, France."
              }
            },
            {
              "_id": "rec26",
              "_score": 0.081794746,
              "fields": {
                "category": "history",
                "chunk_text": "Rome was once the center of a vast empire."
              }
            },
            {
              "_id": "rec7",
              "_score": 0.078153394,
              "fields": {
                "category": "history",
                "chunk_text": "The Great Wall of China was built to protect against invasions."
              }
            },
            {
              "_id": "rec47",
              "_score": 0.06604649,
              "fields": {
                "category": "history",
                "chunk_text": "The Industrial Revolution transformed manufacturing and transportation."
              }
            },
            {
              "_id": "rec21",
              "_score": 0.056735568,
              "fields": {
                "category": "history",
                "chunk_text": "The Statue of Liberty was a gift from France to the United States."
              }
            }
          ]
        },
        "usage": {
          "read_units": 6,
          "embed_total_tokens": 8
        }
      }
      ```
    </CodeGroup>

    ### 5. Rerank results

    To get a more accurate ranking, search again but this time [rerank the initial results](/guides/search/rerank-results) based on their relevance to the query.

    <CodeGroup>
      ```python Python {10-14} theme={null}
      # Search the dense index and rerank results
      reranked_results = dense_index.search(
          namespace="example-namespace",
          query={
              "top_k": 10,
              "inputs": {
                  'text': query
              }
          },
          rerank={
              "model": "bge-reranker-v2-m3",
              "top_n": 10,
              "rank_fields": ["chunk_text"]
          }   
      )

      # Print the reranked results
      for hit in reranked_results['result']['hits']:
          print(f"id: {hit['_id']}, score: {round(hit['_score'], 2)}, text: {hit['fields']['chunk_text']}, category: {hit['fields']['category']}")
      ```

      ```javascript JavaScript {7-11} theme={null}
      // Search the dense index and rerank results
      const rerankedResults = await index.searchRecords({
        query: {
          topK: 10,
          inputs: { text: query },
        },
        rerank: {
          model: 'bge-reranker-v2-m3',
          topN: 10,
          rankFields: ['chunk_text'],
        },
      });

      // Print the reranked results
      rerankedResults.result.hits.forEach(hit => {
        console.log(`id: ${hit.id}, score: ${hit.score.toFixed(2)}, text: ${hit.fields.chunk_text}, category: ${hit.fields.category}`);
      });
      ```

      ```java Java {9} theme={null}
      // Add to the Quickstart class:
      // Define the rerank parameters
      List<String>rankFields = new ArrayList<>();
      rankFields.add("chunk_text");
      SearchRecordsRequestRerank rerank = new SearchRecordsRequestRerank()
              .query(query)
              .model("bge-reranker-v2-m3")
              .topN(10)
              .rankFields(rankFields);

      // Search the dense index and rerank results
      SearchRecordsResponse recordsResponseReranked = index.searchRecordsByText(query,  "example-namespace", fields, 10, null, rerank );

      // Print the reranked results
      System.out.println(recordsResponseReranked);
      ```

      ```go Go {11-15} theme={null}
      // Add to the main function:
      // Search the dense index and rerank results
      topN := int32(10)
      resReranked, err := idxConnection.SearchRecords(ctx, &pinecone.SearchRecordsRequest{
          Query: pinecone.SearchRecordsQuery{
              TopK: 10,
              Inputs: &map[string]interface{}{
                  "text": query,
              },
          },
          Rerank: &pinecone.SearchRecordsRerank{
              Model:      "bge-reranker-v2-m3",
              TopN:       &topN,
              RankFields: []string{"chunk_text"},
          },
      })
      if err != nil {
          log.Fatalf("Failed to search records: %v", err)
      }
      fmt.Printf(prettifyStruct(resReranked))
      ```

      ```csharp C# {12-17} theme={null}
      // Search the dense index and rerank results
      var responseReranked = await index.SearchRecordsAsync(
          "example-namespace",
          new SearchRecordsRequest
          {
              Query = new SearchRecordsRequestQuery
              {
                  TopK = 10,
                  Inputs = new Dictionary<string, object?> { { "text", "Famous historical structures and monuments" } },
              },
              Fields = ["category", "chunk_text"],
              Rerank = new SearchRecordsRequestRerank
              {
                  Model = "bge-reranker-v2-m3",
                  TopN = 10,
                  RankFields = ["chunk_text"],
              }
          }
      );

      Console.WriteLine(responseReranked);
      ```
    </CodeGroup>

    Notice that all of the most relevant results about historical structures and monuments are now ranked highest.

    <CodeGroup>
      ```console Python theme={null}
      id: rec1  | score: 0.11  | category: history    | text: The Eiffel Tower was completed in 1889 and stands in Paris, France.
      id: rec38 | score: 0.06  | category: history    | text: The Taj Mahal is a mausoleum built by Emperor Shah Jahan.
      id: rec7  | score: 0.06  | category: history    | text: The Great Wall of China was built to protect against invasions.
      id: rec17 | score: 0.02  | category: history    | text: The Pyramids of Giza are among the Seven Wonders of the Ancient World.
      id: rec26 | score: 0.01  | category: history    | text: Rome was once the center of a vast empire.        
      id: rec15 | score: 0.01  | category: art        | text: Leonardo da Vinci painted the Mona Lisa.          
      id: rec5  | score: 0.0   | category: literature | text: Shakespeare wrote many famous plays, including Hamlet and Macbeth.
      id: rec47 | score: 0.0   | category: history    | text: The Industrial Revolution transformed manufacturing and transportation.
      id: rec50 | score: 0.0   | category: energy     | text: Renewable energy sources include wind, solar, and hydroelectric power.
      id: rec3  | score: 0.0   | category: science    | text: Albert Einstein developed the theory of relativity.
      ```

      ```console JavaScript theme={null}
      id: rec1, score: 0.11, text: The Eiffel Tower was completed in 1889 and stands in Paris, France., category: history
      id: rec38, score: 0.06, text: The Taj Mahal is a mausoleum built by Emperor Shah Jahan., category: history
      id: rec7, score: 0.06, text: The Great Wall of China was built to protect against invasions., category: history
      id: rec17, score: 0.02, text: The Pyramids of Giza are among the Seven Wonders of the Ancient World., category: history
      id: rec26, score: 0.01, text: Rome was once the center of a vast empire., category: history
      id: rec15, score: 0.01, text: Leonardo da Vinci painted the Mona Lisa., category: art
      id: rec5, score: 0.00, text: Shakespeare wrote many famous plays, including Hamlet and Macbeth., category: literature
      id: rec47, score: 0.00, text: The Industrial Revolution transformed manufacturing and transportation., category: history
      id: rec50, score: 0.00, text: Renewable energy sources include wind, solar, and hydroelectric power., category: energy
      id: rec3, score: 0.00, text: Albert Einstein developed the theory of relativity., category: science
      ```

      ```java Java [expandable] theme={null}
      class SearchRecordsResponse {
          result: class SearchRecordsResponseResult {
              hits: [class Hit {
                  id: rec1
                  score: 0.10687689
                  fields: {category=history, chunk_text=The Eiffel Tower was completed in 1889 and stands in Paris, France.}
                  additionalProperties: null
              }, class Hit {
                  id: rec38
                  score: 0.06418265
                  fields: {category=history, chunk_text=The Taj Mahal is a mausoleum built by Emperor Shah Jahan.}
                  additionalProperties: null
              }, class Hit {
                  id: rec7
                  score: 0.062445287
                  fields: {category=history, chunk_text=The Great Wall of China was built to protect against invasions.}
                  additionalProperties: null
              }, class Hit {
                  id: rec17
                  score: 0.0153063545
                  fields: {category=history, chunk_text=The Pyramids of Giza are among the Seven Wonders of the Ancient World.}
                  additionalProperties: null
              }, class Hit {
                  id: rec26
                  score: 0.010652511
                  fields: {category=history, chunk_text=Rome was once the center of a vast empire.}
                  additionalProperties: null
              }, class Hit {
                  id: rec5
                  score: 3.194182E-5
                  fields: {category=literature, chunk_text=Shakespeare wrote many famous plays, including Hamlet and Macbeth.}
                  additionalProperties: null
              }, class Hit {
                  id: rec47
                  score: 1.7502925E-5
                  fields: {category=history, chunk_text=The Industrial Revolution transformed manufacturing and transportation.}
                  additionalProperties: null
              }, class Hit {
                  id: rec32
                  score: 1.631454E-5
                  fields: {category=biology, chunk_text=Elephants have excellent memory and strong social bonds.}
                  additionalProperties: null
              }, class Hit {
                  id: rec4
                  score: 1.6187581E-5
                  fields: {category=biology, chunk_text=The mitochondrion is often called the powerhouse of the cell.}
                  additionalProperties: null
              }, class Hit {
                  id: rec45
                  score: 1.6061611E-5
                  fields: {category=literature, chunk_text=A haiku is a traditional Japanese poem with a 5-7-5 syllable structure.}
                  additionalProperties: null
              }]
              additionalProperties: null
          }
          usage: class SearchUsage {
              readUnits: 6
              embedTotalTokens: 13
              rerankUnits: 1
              additionalProperties: null
          }
          additionalProperties: null
      }
      ```

      ```json Go [expandable] theme={null}
      {
        "result": {
          "hits": [
            {
              "_id": "rec1",
              "_score": 0.10743748,
              "fields": {
                "category": "history",
                "chunk_text": "The Eiffel Tower was completed in 1889 and stands in Paris, France."
              }
            },
            {
              "_id": "rec38",
              "_score": 0.064535476,
              "fields": {
                "category": "history",
                "chunk_text": "The Taj Mahal is a mausoleum built by Emperor Shah Jahan."
              }
            },
            {
              "_id": "rec7",
              "_score": 0.062445287,
              "fields": {
                "category": "history",
                "chunk_text": "The Great Wall of China was built to protect against invasions."
              }
            },
            {
              "_id": "rec17",
              "_score": 0.0153063545,
              "fields": {
                "category": "history",
                "chunk_text": "The Pyramids of Giza are among the Seven Wonders of the Ancient World."
              }
            },
            {
              "_id": "rec26",
              "_score": 0.010652511,
              "fields": {
                "category": "history",
                "chunk_text": "Rome was once the center of a vast empire."
              }
            },
            {
              "_id": "rec15",
              "_score": 0.007876706,
              "fields": {
                "category": "art",
                "chunk_text": "Leonardo da Vinci painted the Mona Lisa."
              }
            },
            {
              "_id": "rec5",
              "_score": 0.00003194182,
              "fields": {
                "category": "literature",
                "chunk_text": "Shakespeare wrote many famous plays, including Hamlet and Macbeth."
              }
            },
            {
              "_id": "rec47",
              "_score": 0.000017502925,
              "fields": {
                "category": "history",
                "chunk_text": "The Industrial Revolution transformed manufacturing and transportation."
              }
            },
            {
              "_id": "rec50",
              "_score": 0.00001631454,
              "fields": {
                "category": "energy",
                "chunk_text": "Renewable energy sources include wind, solar, and hydroelectric power."
              }
            },
            {
              "_id": "rec3",
              "_score": 0.000015936621,
              "fields": {
                "category": "science",
                "chunk_text": "Albert Einstein developed the theory of relativity."
              }
            }
          ]
        },
        "usage": {
          "read_units": 6,
          "embed_total_tokens": 8,
          "rerank_units": 1
        }
      }
      ```

      ```csharp C# [expandable] theme={null}
      {
        "result": {
          "hits": [
            {
              "_id": "rec1",
              "_score": 0.10687689,
              "fields": {
                "category": "history",
                "chunk_text": "The Eiffel Tower was completed in 1889 and stands in Paris, France."
              }
            },
            {
              "_id": "rec38",
              "_score": 0.064535476,
              "fields": {
                "category": "history",
                "chunk_text": "The Taj Mahal is a mausoleum built by Emperor Shah Jahan."
              }
            },
            {
              "_id": "rec7",
              "_score": 0.062445287,
              "fields": {
                "category": "history",
                "chunk_text": "The Great Wall of China was built to protect against invasions."
              }
            },
            {
              "_id": "rec21",
              "_score": 0.018511046,
              "fields": {
                "category": "history",
                "chunk_text": "The Statue of Liberty was a gift from France to the United States."
              }
            },
            {
              "_id": "rec17",
              "_score": 0.0153063545,
              "fields": {
                "category": "history",
                "chunk_text": "The Pyramids of Giza are among the Seven Wonders of the Ancient World."
              }
            },
            {
              "_id": "rec26",
              "_score": 0.010652511,
              "fields": {
                "category": "history",
                "chunk_text": "Rome was once the center of a vast empire."
              }
            },
            {
              "_id": "rec15",
              "_score": 0.007876706,
              "fields": {
                "category": "art",
                "chunk_text": "Leonardo da Vinci painted the Mona Lisa."
              }
            },
            {
              "_id": "rec5",
              "_score": 0.00003194182,
              "fields": {
                "category": "literature",
                "chunk_text": "Shakespeare wrote many famous plays, including Hamlet and Macbeth."
              }
            },
            {
              "_id": "rec47",
              "_score": 0.000017502925,
              "fields": {
                "category": "history",
                "chunk_text": "The Industrial Revolution transformed manufacturing and transportation."
              }
            },
            {
              "_id": "rec50",
              "_score": 0.00001631454,
              "fields": {
                "category": "energy",
                "chunk_text": "Renewable energy sources include wind, solar, and hydroelectric power."
              }
            }
          ]
        },
        "usage": {
          "read_units": 6,
          "embed_total_tokens": 8,
          "rerank_units": 1
        }
      }
      ```
    </CodeGroup>

    ### 6. Improve results

    [Reranking results](/guides/search/rerank-results) is one of the most effective ways to improve search accuracy and relevance, but there are many other techniques to consider. For example:

    * [Filtering by metadata](/guides/search/filter-by-metadata): When records contain additional metadata, you can limit the search to records matching a [filter expression](/guides/index-data/indexing-overview#metadata-filter-expressions).

    * [Hybrid search](/guides/search/hybrid-search): You can add [lexical search](/guides/search/lexical-search) to capture precise keyword matches (e.g., product SKUs, email addresses, domain-specific terms) in addition to semantic matches.

    * [Chunking strategies](https://www.pinecone.io/learn/chunking-strategies/): You can chunk your content in different ways to get better results. Consider factors like the length of the content, the complexity of queries, and how results will be used in your application.

    ### 7. Clean up

    When you no longer need your example index, delete it as follows:

    <CodeGroup>
      ```python Python theme={null}
      # Delete the index
      pc.delete_index(index_name)
      ```

      ```javascript JavaScript theme={null}
      // Delete the index
      await pc.deleteIndex(indexName);
      ```

      ```java Java theme={null}
      // Add to the Quickstart class:
      // Delete the index
      pc.deleteIndex(indexName);
      ```

      ```go Go theme={null}
      // Add to the main function:
      // Delete the index
      err = pc.DeleteIndex(ctx, indexName)
      if err != nil {
          log.Fatalf("Failed to delete index: %v", err)
      } else {
          fmt.Println("Index \"%v\" deleted successfully", indexName)
      }
      ```

      ```csharp C# theme={null}
      // Delete the index
      await pinecone.DeleteIndexAsync(indexName);
      ```
    </CodeGroup>

    <Tip>
      For production indexes, consider [enabling deletion protection](/guides/manage-data/manage-indexes#configure-deletion-protection).
    </Tip>

    ### Next steps

    <CardGroup cols={3}>
      <Card title="Index data" icon="book-open" href="/guides/index-data/indexing-overview">
        Learn more about storing data in Pinecone
      </Card>

      <Card title="Search" icon="magnifying-glass" href="/guides/search/search-overview">
        Explore different forms of vector search.
      </Card>

      <Card title="Optimize" icon="rocket" href="/guides/optimize">
        Find out how to improve performance
      </Card>
    </CardGroup>
  </Tab>

  <Tab title="Cursor">
    Use [Cursor](https://cursor.com/) to build sample apps with Pinecone. Work with an AI agent that understands Pinecone APIs and implements production-ready patterns automatically, without manually copying code snippets.

    To get started, click this button:

    <div className="not-prose">
      <a
        href="cursor://anysphere.cursor-deeplink/prompt?text=First%2C%20download%20and%20run%20this%20remote%20script%3A%0A%0Acurl%20-sSL%20https%3A%2F%2Fdocs.pinecone.io%2Finstall-agent-reference%20%7C%20sh%0A%0AThen%2C%20help%20me%20get%20started%20with%20Pinecone."
        style={{
    display: "inline-flex",
    alignItems: "center",
    justifyContent: "center",
    height: 36,
    paddingLeft: 16,
    paddingRight: 16,
    paddingTop: 6,
    paddingBottom: 6,
    background: "var(--brand-blue)",
    borderRadius: 4,
    gap: 8,
    textDecoration: "none",
    cursor: "pointer",
    transition: "background 0.2s ease",
  }}
        onMouseOver={(e) => (e.currentTarget.style.background = "var(--primary-dark)")}
        onMouseOut={(e) => (e.currentTarget.style.background = "var(--brand-blue)")}
      >
        <span
          style={{
      color: "white",
      fontSize: 14,
      fontWeight: "500",
      letterSpacing: 0.4,
    }}
        >
          Download AGENTS.md and launch in Cursor
        </span>
      </a>
    </div>

    This opens Cursor and enters this prompt:

    ```text  theme={null}
    First, download and run this remote script: 

    curl -sSL https://docs.pinecone.io/install-agent-reference | sh 

    Then, help me get started with Pinecone.
    ```

    <Note>
      To see the contents of the install script, use a web browser to visit this URL: [https://docs.pinecone.io/install-agent-reference](https://docs.pinecone.io/install-agent-reference).
    </Note>

    When you run this prompt:

    1. Cursor executes a remote script that downloads Pinecone agent reference files, extracts them into an `.agents` folder, and creates an `AGENTS.md` file. These files provide Cursor with context it can use to help you get started with Pinecone.

    2. Then, Cursor helps you build a sample app of your choice (quick test, semantic search, RAG, or recommendations) in the programming language of your choice.

    ### Next steps

    After building some sample apps, use the `AGENTS.md` file to continue building with Pinecone, planning and implementing code for your own use cases.

    To learn more about Pinecone, check out these resources:

    <CardGroup cols={3}>
      <Card title="Index data" icon="book-open" href="/guides/index-data/indexing-overview">
        Learn more about storing data in Pinecone
      </Card>

      <Card title="Search" icon="magnifying-glass" href="/guides/search/search-overview">
        Explore different forms of vector search.
      </Card>

      <Card title="Optimize" icon="rocket" href="/guides/optimize">
        Find out how to improve performance
      </Card>
    </CardGroup>
  </Tab>

  <Tab title="Claude Code">
    Use [Claude Code](https://www.claude.com/product/claude-code) to build a Pinecone application with current best practices. Instead of copying code snippets, work with an agent that understands Pinecone APIs and implements production-ready patterns automatically.

    <Note>
      Because this quickstart relies on AI, the exact implementation may vary each time.
    </Note>

    <Tip>
      If you don't have Claude Code installed, see the [Claude Code quickstart](https://docs.claude.com/en/docs/claude-code/quickstart).
    </Tip>

    ### 1. Set your API key

    Set the `PINECONE_API_KEY` environment variable before installing the plugin and activating Claude Code:

    ```shell  theme={null}
    export PINECONE_API_KEY="{{YOUR_API_KEY}}"
    ```

    ### 2. Install the Pinecone plugin for Claude Code

    The Pinecone plugin provides Claude Code with up-to-date Pinecone API references and best practices. Install it using one of these methods:

    1. From your terminal, run:

       ```shell  theme={null}
       claude plugin install pinecone
       ```

    2. Or, from within Claude Code, run:

       ```text  theme={null}
       /plugin install pinecone
       ```

    <Tip>
      For more information about Pinecone's Claude Code plugin, see its [GitHub repository](https://github.com/pinecone-io/pinecone-claude-code-plugin).
    </Tip>

    ### 3. Run the quickstart command

    To use the plugin to work through Pinecone's quickstart, start Claude Code and run this command:

    ```text  theme={null}
    /pinecone:quickstart
    ```

    This command does the following things:

    * Downloads and configures an [`AGENTS.md`](https://github.com/pinecone-io/pinecone-agents-ref) file that provides Claude Code with up-to-date Pinecone API information and best practices.
    * Guides you through setting up Pinecone using the [Pinecone CLI](https://docs.pinecone.io/reference/cli/overview) and [Pinecone MCP server](https://docs.pinecone.io/guides/operations/mcp-server).
    * Generates sample code to demonstrate various Pinecone use cases (quick start, semantic search, RAG, or recommendations).

    Once complete, you can use the configured `AGENTS.md` files, MCP server, and CLI to build your application. The plugin also includes other slash commands, such as `/pinecone:query`, for interactively querying your indexes.
  </Tab>

  <Tab title="n8n">
    Use [n8n](https://docs.n8n.io/choose-n8n/) to create a workflow that downloads files via HTTP and lets you chat with them using Pinecone Database and OpenAI.

    <Tip>
      If you're not interested in chunking and embedding your own data or figuring out which search method to use, [use n8n with Pinecone Assistant](/guides/assistant/quickstart/n8n-quickstart) instead.
    </Tip>

    ### 1. Get an OpenAI API key

    Create a new API key in the [OpenAI console](https://platform.openai.com/api-keys).

    ### 2. Create an index

    [Create an index](https://app.pinecone.io/organizations/-/projects/-/create-index/serverless) in the Pinecone console:

    * Name your index `n8n-dense-index`
    * Under **Configuration**, check **Custom settings** and set **Dimension** to 1536.
    * Leave everything else as default.

    ### 3. Set up n8n

    <Steps>
      <Step title="Create a new workflow">
        In your n8n account, [create a new workflow](https://docs.n8n.io/workflows/create/).
      </Step>

      <Step title="Import a workflow template">
        Copy this workflow template URL:

        ```shell  theme={null}
        https://raw.githubusercontent.com/pinecone-io/n8n-templates/refs/heads/main/database-quickstart/database-quickstart.json
        ```

        Paste the URL into the workflow editor and then click **Import** to add the workflow.
      </Step>

      <Step title="Add credentials to the workflow">
        * Add your Pinecone credentials:
          * In the **Pinecone Vector Store** node, select **Credential to connect with** > **Create new credential** and paste in your Pinecone API key.
          * Name the credential **Pinecone** so that other nodes reference it.

        * Add your OpenAI credentials:
          * In the **OpenAI Chat Model**, select **Credential to connect with** > **Create new credential** and paste in your OpenAI API key.
      </Step>

      <Step title="Activate the workflow">
        The workflow is configured to download recent Pinecone release notes and upload them to your Pinecone index. Click **Execute workflow** to start the workflow.

        <Tip>
          You can add your own files to the workflow by changing the URLs in the **Set file urls** node.
        </Tip>
      </Step>
    </Steps>

    ### 4. Chat with your docs

    Once the workflow is activated, ask it for the latest changes to Pinecone Database:

    ```
    What's new in Pinecone Database?
    ```

    ### Next steps

    * Use your own data:
      * Change the urls in **Set file urls** node to use your own files.
      * You may need to adjust the chunk sizes in the **Recursive Character Text Splitter** node or use a different chunking strategy. You want chunks that are big enough to contain meaningful information but not so big that the meaning is diluted or it can't fit within the context window of the embedding model. See [Chunking Strategies for LLM Applications](https://www.pinecone.io/learn/chunking-strategies/) for more info.
      * Customize the system message of the **AI Agent** node to reflect what the **Pinecone Vector Store Tool** will be used for. Be sure to include info on what data can be retrieved using that tool.
      * Customize the description of the **Pinecone Vector Store Tool** to reflect what data you are storing in the Pinecone index.
    * Use n8n, Pinecone Assistant, and OpenAI to [chat with your Google Drive documents](https://n8n.io/workflows/9942-rag-powered-document-chat-with-google-drive-openai-and-pinecone-assistant/).
    * Get help in the [Pinecone Discord community](https://discord.gg/tJ8V62S3sH).
  </Tab>
</Tabs>

---

# Source: https://docs.pinecone.io/integrations/redpanda.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Redpanda

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Redpanda Connect is a declarative data streaming service that solves a wide range of data engineering problems with simple, chained, stateless processing steps. It implements transaction based resiliency with back pressure, so when connecting to at-least-once sources and sinks, it's able to guarantee at-least-once delivery without needing to persist messages during transit.

It's simple to deploy, comes with a wide range of connectors, and is totally data agnostic, making it easy to drop into your existing infrastructure. Redpanda Connect has functionality that overlaps with integration frameworks, log aggregators and ETL workflow engines, and can therefore be used to complement these traditional data engineering tools or act as a simpler alternative.

The Pinecone connector for Redpanda provides a production-ready integration from many existing data sources, all in a few lines of YAML.

<PrimarySecondaryCTA primaryHref={"https://docs.redpanda.com/redpanda-connect/components/outputs/pinecone/"} primaryLabel={"Get started"} primaryTarget={"_blank"} />

---

# Source: https://docs.pinecone.io/examples/reference-architectures.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Reference architectures

<Card title="AWS Reference Architecture" icon="text-size" iconType="duotone">
  <br />

  The official AWS Reference Architecture for high-scale systems using Pinecone.

  <br />

  <a href="#" className="documentation-link">
    [Documentation](https://github.com/pinecone-io/aws-reference-architecture-pulumi/blob/main/README.md)
  </a>

  <a href="#" className="documentation-link">
    [Video tutorial](https://youtu.be/ySznARngHts)
  </a>

  <a href="#" className="documentation-link">
    [Source code](https://github.com/pinecone-io/aws-reference-architecture-pulumi)
  </a>
</Card>

---

# Pinecone Python SDK

[![License](https://img.shields.io/github/license/pinecone-io/pinecone-python-client?color=orange)](https://img.shields.io/github/license/pinecone-io/pinecone-python-client?color=orange)

The official Pinecone Python SDK.

## Documentation

*   [Conceptual docs and guides](https://docs.pinecone.io/)
*   [Github Source](https://github.com/pinecone-io/pinecone-python-client)

## Points of Interest

### DB Control Plane

*   [Pinecone](https://docs.pinecone.io/rest.html#pinecone.Pinecone)
*   [PineconeAsyncio](https://docs.pinecone.io/asyncio.html#pinecone.PineconeAsyncio)
*   [PineconeGRPC](https://docs.pinecone.io/grpc.html#pinecone.PineconeGRPC)

### DB Data Operations

*   [Index](https://docs.pinecone.io/rest.html#db-data-plane)
*   [IndexAsyncio](https://docs.pinecone.io/asyncio.html#db-data-plane)
*   [GRPCIndex](https://docs.pinecone.io/grpc.html#db-data-plane)

### Inference API

*   [Inference](https://docs.pinecone.io/rest.html#inference)
*   [InferenceAsyncio](https://docs.pinecone.io/asyncio.html#inference)

## Upgrading the SDK

> **Note**
>
> The official SDK package was renamed from `pinecone-client` to `pinecone` beginning in version `5.1.0`. Please remove `pinecone-client` from your project dependencies and add `pinecone` instead to get the latest updates.

For notes on changes between major versions, see [Upgrading](https://docs.pinecone.io/docs/upgrading.md).

## Prerequisites

*   The Pinecone Python SDK is compatible with Python 3.10 and greater. It has been tested with CPython versions from 3.10 to 3.13.
*   Before you can use the Pinecone SDK, you must sign up for an account and find your API key in the Pinecone console dashboard at [https://app.pinecone.io](https://app.pinecone.io/).

## Installation

The Pinecone Python SDK is distributed on PyPI using the package name `pinecone`. By default the `pinecone` has a minimal set of dependencies, but you can install some extras to unlock additional functionality.

### Available Extras

*   `pinecone[asyncio]` will add a dependency on `aiohttp` and enable usage of `PineconeAsyncio`, the asyncio-enabled version of the client for use with highly asynchronous modern web frameworks such as FastAPI.
*   `pinecone[grpc]` will add dependencies on `grpcio` and related libraries needed to make pinecone data calls such as `upsert` and `query` over [GRPC](https://grpc.io/) for a modest performance improvement. See the guide on [tuning performance](https://docs.pinecone.io/docs/performance-tuning).

#### Installing with pip

```bash
# Install the latest version
pip3 install pinecone

# Install the latest version, with optional dependencies
pip3 install "pinecone[asyncio,grpc]"
```

#### Installing with uv

[uv](https://docs.astral.sh/uv/) is a modern package manager that runs 10-100x faster than pip and supports most pip syntax.

```bash
# Install the latest version
uv add pinecone

# Install the latest version, optional dependencies
uv add "pinecone[asyncio,grpc]"
```

#### Installing with poetry

```bash
# Install the latest version
poetry add pinecone

# Install the latest version, with optional dependencies
poetry add pinecone --extras asyncio --extras grpc
```

## Quickstart

### Bringing Your Own Vectors to Pinecone

```python
from pinecone import (
   Pinecone,
   ServerlessSpec,
   CloudProvider,
   AwsRegion,
   VectorType
)

# 1. Instantiate the Pinecone client
pc = Pinecone(api_key='YOUR_API_KEY')

# 2. Create an index
index_config = pc.create_index(
   name="index-name",
   dimension=1536,
   spec=ServerlessSpec(
      cloud=CloudProvider.AWS,
      region=AwsRegion.US_EAST_1
   ),
   vector_type=VectorType.DENSE
)

# 3. Instantiate an Index client
idx = pc.Index(host=index_config.host)

# 4. Upsert embeddings
idx.upsert(
   vectors=[
      ("id1", [0.1, 0.2, 0.3, 0.4, ...], {"metadata_key": "value1"}),
      ("id2", [0.2, 0.3, 0.4, 0.5, ...], {"metadata_key": "value2"}),
   ],
   namespace="example-namespace"
)

# 5. Query your index using an embedding
query_embedding = [...] # list should have length == index dimension
idx.query(
   vector=query_embedding,
   top_k=10,
   include_metadata=True,
   filter={
      "metadata_key": { "$eq": "value1" }
   }
)
```

### Bring Your Own Data Using Pinecone Integrated Inference

```python
from pinecone import (
   Pinecone,
   CloudProvider,
   AwsRegion,
   EmbedModel,
)

# 1. Instantiate the Pinecone client
pc = Pinecone(api_key="<<<PINECONE_API_KEY>>>")


# 2. Create an index configured for use with a particular model
index_config = pc.create_index_for_model(
   name="my-model-index",
   cloud=CloudProvider.AWS,
   region=AwsRegion.US_EAST_1,
   embed=IndexEmbed(
      model=EmbedModel.Multilingual_E5_Large,
      field_map={"text": "my_text_field"}
   )
)


# 3. Instantiate an Index client
idx = pc.Index(host=index_config.host)

# 4. Upsert records
idx.upsert_records(
   namespace="my-namespace",
   records=[
      {
            "_id": "test1",
            "my_text_field": "Apple is a popular fruit known for its sweetness and crisp texture.",
      },
      {
            "_id": "test2",
            "my_text_field": "The tech company Apple is known for its innovative products like the iPhone.",
      },
      {
            "_id": "test3",
            "my_text_field": "Many people enjoy eating apples as a healthy snack.",
      },
      {
            "_id": "test4",
            "my_text_field": "Apple Inc. has revolutionized the tech industry with its sleek designs and user-friendly interfaces.",
      },
      {
            "_id": "test5",
            "my_text_field": "An apple a day keeps the doctor away, as the saying goes.",
      },
      {
            "_id": "test6",
            "my_text_field": "Apple Computer Company was founded on April 1, 1976, by Steve Jobs, Steve Wozniak, and Ronald Wayne as a partnership.",
      },
   ],
)


# 5. Search for similar records
from pinecone import SearchQuery, SearchRerank, RerankModel

response = index.search_records(
   namespace="my-namespace",
   query=SearchQuery(
      inputs={
            "text": "Apple corporation",
      },
      top_k=3
   ),
   rerank=SearchRerank(
      model=RerankModel.Bge_Reranker_V2_M3,
      rank_fields=["my_text_field"],
      top_n=3,
   ),
)
```

## Issues & Bugs

If you notice bugs or have feedback, please [file an issue](https://github.com/pinecone-io/pinecone-python-client/issues).

You can also get help in the [Pinecone Community Forum](https://community.pinecone.io/).

---

# Source: https://docs.pinecone.io/troubleshooting/remove-metadata-field.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Remove a metadata field from a record

You must perform an [`upsert`](/reference/api/2024-10/data-plane/upsert) operation to remove existing metadata fields from a record.

You will need to provide the existing ID and values of the vector. The metadata you provide in the upsert operation will replace any existing metadata, thus clearing the fields you seek to drop.

Metadata fields cannot be removed using the `update` operation.

---

# Source: https://docs.pinecone.io/guides/search/rerank-results.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Rerank results

> Improve the quality of results with reranking.

Reranking is used as part of a two-stage vector retrieval process to improve the quality of results. You first query an index for a given number of relevant results, and then you send the query and results to a reranking model. The reranking model scores the results based on their semantic relevance to the query and returns a new, more accurate ranking. This approach is one of the simplest methods for improving quality in retrieval augmented generation (RAG) pipelines.

Pinecone provides [hosted reranking models](#reranking-models) so it's easy to manage two-stage vector retrieval on a single platform. You can use a hosted model to rerank results as an integrated part of a query, or you can use a hosted model or external model to rerank results as a standalone operation.

## Integrated reranking

To rerank initial results as an integrated part of a query, without any extra steps, use the [`search`](/reference/api/latest/data-plane/search_records) operation with the `rerank` parameter, including the [hosted reranking model](#reranking-models) you want to use, the number of reranked results to return, and the fields to use for reranking, if different than the main query.

For example, the following code searches for the 3 records most semantically related to a query text and uses the `hosted bge-reranker-v2-m3` model to rerank the results and return only the 2 most relevant documents:

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  ranked_results = index.search(
      namespace="example-namespace", 
      query={
          "inputs": {"text": "Disease prevention"}, 
          "top_k": 4
      },
      rerank={
          "model": "bge-reranker-v2-m3",
          "top_n": 2,
          "rank_fields": ["chunk_text"]
      },
      fields=["category", "chunk_text"]
  )

  print(ranked_results)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const namespace = pc.index("INDEX_NAME", "INDEX_HOST").namespace("example-namespace");

  const response = await namespace.searchRecords({
    query: {
      topK: 2,
      inputs: { text: 'Disease prevention' },
    },
    fields: ['chunk_text', 'category'],
    rerank: {
      model: 'bge-reranker-v2-m3',
      rankFields: ['chunk_text'],
      topN: 2,
    },
  });

  console.log(response);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import org.openapitools.db_data.client.ApiException;
  import org.openapitools.db_data.client.model.SearchRecordsRequestRerank;
  import org.openapitools.db_data.client.model.SearchRecordsResponse;

  import java.util.*;

  public class SearchText {
      public static void main(String[] args) throws ApiException {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);

          Index index = new Index(config, connection, "integrated-dense-java");

          String query = "Disease prevention";
          List<String> fields = new ArrayList<>();
          fields.add("category");
          fields.add("chunk_text");

          List<String>rankFields = new ArrayList<>();
          rankFields.add("chunk_text");
          SearchRecordsRequestRerank rerank = new SearchRecordsRequestRerank()
                  .query(query)
                  .model("bge-reranker-v2-m3")
                  .topN(2)
                  .rankFields(rankFields);

          SearchRecordsResponse recordsResponseReranked = index.searchRecordsByText(query,  "example-namespace", fields,4, null, rerank);

          System.out.println(recordsResponseReranked);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
    	bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
      } 

      topN := int32(2)
      res, err := idxConnection.SearchRecords(ctx, &pinecone.SearchRecordsRequest{
          Query: pinecone.SearchRecordsQuery{
              TopK: 3,
              Inputs: &map[string]interface{}{
                  "text": "Disease prevention",
              },
          },
          Rerank: &pinecone.SearchRecordsRerank{
              Model:      "bge-reranker-v2-m3",
              TopN:       &topN,
              RankFields: []string{"chunk_text"},
          },
          Fields: &[]string{"chunk_text", "category"},
      })
      if err != nil {
          log.Fatalf("Failed to search records: %v", err)
      }
      fmt.Printf(prettifyStruct(res))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var response = await index.SearchRecordsAsync(
      "example-namespace",
      new SearchRecordsRequest
      {
          Query = new SearchRecordsRequestQuery
          {
              TopK = 4,
              Inputs = new Dictionary<string, object?> { { "text", "Disease prevention" } },
          },
          Fields = ["category", "chunk_text"],
          Rerank = new SearchRecordsRequestRerank
          {
              Model = "bge-reranker-v2-m3",
              TopN = 2,
              RankFields = ["chunk_text"],
          }
      }
  );

  Console.WriteLine(response);
  ```

  ```shell curl theme={null}
  INDEX_HOST="INDEX_HOST"
  NAMESPACE="YOUR_NAMESPACE"
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://$INDEX_HOST/records/namespaces/$NAMESPACE/search" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: unstable" \
    -d '{
          "query": {
              "inputs": {"text": "Disease prevention"},
              "top_k": 4
          },
          "rerank": {
              "model": "bge-reranker-v2-m3",
              "top_n": 2,
              "rank_fields": ["chunk_text"]
          },
          "fields": ["category", "chunk_text"]
       }'
  ```
</CodeGroup>

The response looks as follows. For each hit, the `_score` represents the relevance of a document to the query, normalized between 0 and 1, with scores closer to 1 indicating higher relevance.

<CodeGroup>
  ```python Python theme={null}
  {'result': {'hits': [{'_id': 'rec3',
                        '_score': 0.004399413242936134,
                        'fields': {'category': 'immune system',
                                   'chunk_text': 'Rich in vitamin C and other '
                                                  'antioxidants, apples '
                                                  'contribute to immune health '
                                                  'and may reduce the risk of '
                                                  'chronic diseases.'}},
                       {'_id': 'rec4',
                        '_score': 0.0029235430993139744,
                        'fields': {'category': 'endocrine system',
                                   'chunk_text': 'The high fiber content in '
                                                  'apples can also help regulate '
                                                  'blood sugar levels, making '
                                                  'them a favorable snack for '
                                                  'people with diabetes.'}}]},
   'usage': {'embed_total_tokens': 8, 'read_units': 6, 'rerank_units': 1}}
  ```

  ```javascript JavaScript theme={null}
  {
    result: { 
      hits: [ 
        {
          _id: 'rec3',
          _score: 0.004399413242936134,
          fields: {
            category: 'immune system',
            chunk_text: 'Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases.'
          }
        },
        {
          _id: 'rec4',
          _score: 0.0029235430993139744,
          fields: {
            category: 'endocrine system',
            chunk_text: 'The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes.'
          }
        }
      ]
    },
    usage: { 
      readUnits: 6, 
      embedTotalTokens: 8,
      rerankUnits: 1 
    }
  }
  ```

  ```java Java theme={null}
  class SearchRecordsResponse {
      result: class SearchRecordsResponseResult {
          hits: [class Hit {
              id: rec3
              score: 0.004399413242936134
              fields: {category=immune system, chunk_text=Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases.}
              additionalProperties: null
          }, class Hit {
              id: rec4
              score: 0.0029235430993139744
              fields: {category=endocrine system, chunk_text=The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes.}
              additionalProperties: null
          }]
          additionalProperties: null
      }
      usage: class SearchUsage {
          readUnits: 6
          embedTotalTokens: 13
          rerankUnits: 1
          additionalProperties: null
      }
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  {
    "result": {
      "hits": [
        {
          "_id": "rec3",
          "_score": 0.13683891,
          "fields": {
            "category": "immune system",
            "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases."
          }
        },
        {
          "_id": "rec4",
          "_score": 0.0029235430993139744,
          "fields": {
            "category": "endocrine system",
            "chunk_text": "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes."
          }
        }
      ]
    },
    "usage": {
      "read_units": 6,
      "embed_total_tokens": 8,
      "rerank_units": 1
    }
  }
  ```

  ```csharp C# theme={null}
  {
    "result": {
      "hits": [
        {
          "_id": "rec3",
          "_score": 0.004399413242936134,
          "fields": {
            "category": "immune system",
            "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases."
          }
        },
        {
          "_id": "rec4",
          "_score": 0.0029121784027665854,
          "fields": {
            "category": "endocrine system",
            "chunk_text": "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes."
          }
        }
      ]
    },
    "usage": {
      "read_units": 6,
      "embed_total_tokens": 8,
      "rerank_units": 1
    }
  }
  ```

  ```json curl theme={null}
  {
      "result": {
          "hits": [
              {
                  "_id": "rec3",
                  "_score": 0.004433765076100826,
                  "fields": {
                      "category": "immune system",
                      "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases."
                  }
              },
              {
                  "_id": "rec4",
                  "_score": 0.0029121784027665854,
                  "fields": {
                      "category": "endocrine system",
                      "chunk_text": "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes."
                  }
              }
          ]
      },
      "usage": {
          "embed_total_tokens": 8,
          "read_units": 6,
          "rerank_units": 1
      }
  }
  ```
</CodeGroup>

## Standalone reranking

To rerank initial results as a standalone operation, use the [`rerank`](/reference/api/latest/inference/rerank) operation with the [hosted reranking model](#reranking-models) you want to use, the query results and the query, the number of ranked results to return, the field to use for reranking, and any other model-specific parameters.

For example, the following code uses the hosted `bge-reranker-v2-m3` model to rerank the values of the `documents.chunk_text` fields based on their relevance to the query and return only the 2 most relevant documents, along with their score:

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  ranked_results = pc.inference.rerank(
      model="bge-reranker-v2-m3",
      query="What is AAPL's outlook, considering both product launches and market conditions?",
      documents=[
          {"id": "vec2", "chunk_text": "Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market."},
          {"id": "vec3", "chunk_text": "AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production."},
          {"id": "vec1", "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones."},
      ],
      top_n=2,
      rank_fields=["chunk_text"],
      return_documents=True,
      parameters={
          "truncate": "END"
      }
  )

  print(ranked_results)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const rerankingModel = 'bge-reranker-v2-m3';

  const query = "What is AAPL's outlook, considering both product launches and market conditions?";

  const documents = [
    { id: 'vec2', chunk_text: "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market." },
    { id: 'vec3', chunk_text: "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production." },
    { id: 'vec1', chunk_text: "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones." },
  ];

  const rerankOptions = {
    topN: 2,
    rankFields: ['chunk_text'],
    returnDocuments: true,
    parameters: {
      truncate: 'END'
    }, 
  };

  const rankedResults = await pc.inference.rerank(
    rerankingModel,
    query,
    documents,
    rerankOptions
  );

  console.log(rankedResults);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Inference;
  import io.pinecone.clients.Pinecone;
  import org.openapitools.inference.client.model.RerankResult;
  import org.openapitools.inference.client.ApiException;

  import java.util.*;

  public class RerankExample {
      public static void main(String[] args) throws ApiException {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          Inference inference = pc.getInferenceClient();

          // The model to use for reranking
          String model = "bge-reranker-v2-m3";

          // The query to rerank documents against
          String query = "What is AAPL's outlook, considering both product launches and market conditions?";

          // Add the documents to rerank
          List<Map<String, Object>> documents = new ArrayList<>();
          Map<String, Object> doc1 = new HashMap<>();
          doc1.put("id", "vec2");
          doc1.put("chunk_text", "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market.");
          documents.add(doc1);

          Map<String, Object> doc2 = new HashMap<>();
          doc2.put("id", "vec3");
          doc2.put("chunk_text", "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production");
          documents.add(doc2);

          Map<String, Object> doc3 = new HashMap<>();
          doc3.put("id", "vec1");
          doc3.put("chunk_text", "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.");
          documents.add(doc3);

          // The fields to rank the documents by. If not provided, the default is "text"
          List<String> rankFields = Arrays.asList("chunk_text");

          // The number of results to return sorted by relevance. Defaults to the number of inputs
          int topN = 2;

          // Whether to return the documents in the response
          boolean returnDocuments = true;

          // Additional model-specific parameters for the reranker
          Map<String, Object> parameters = new HashMap<>();
          parameters.put("truncate", "END");

          // Send ranking request
          RerankResult result = inference.rerank(model, query, documents, rankFields, topN, returnDocuments, parameters);

          // Get ranked data
          System.out.println(result.getData());
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
  	ctx := context.Background()

  	pc, err := pinecone.NewClient(pinecone.NewClientParams{
  		ApiKey: "YOUR_API_KEY",
  	})
  	if err != nil {
  		log.Fatalf("Failed to create Client: %v", err)
  	}

  	rerankModel := "bge-reranker-v2-m3"
  	topN := 2
  	returnDocuments := true
  	documents := []pinecone.Document{
  		{"id": "vec2", "chunk_text": "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market."},
  		{"id": "vec3", "chunk_text": "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production."},
  		{"id": "vec1", "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones."},
  	}

  	ranking, err := pc.Inference.Rerank(ctx, &pinecone.RerankRequest{
  		Model:           rerankModel,
  		Query:           "What is AAPL's outlook, considering both product launches and market conditions?",
  		ReturnDocuments: &returnDocuments,
  		TopN:            &topN,
  		RankFields:      &[]string{"chunk_text"},
  		Documents:       documents,
  	})
  	if err != nil {
  		log.Fatalf("Failed to rerank: %v", err)
  	}
  	fmt.Printf(prettifyStruct(ranking))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // Add the documents to rerank
  var documents = new List<Dictionary<string, object?>>
  {
      new()
      {
          ["id"] = "vec2",
          ["chunk_text"] = "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market."
      },
      new()
      {
          ["id"] = "vec3",
          ["chunk_text"] = "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production."
      },
      new()
      {
          ["id"] = "vec1",
          ["chunk_text"] = "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones."
      }
  };

  // The fields to rank the documents by. If not provided, the default is "text"
  var rankFields = new List<string> { "chunk_text" };

  // Additional model-specific parameters for the reranker
  var parameters = new Dictionary<string, object>
  {
      ["truncate"] = "END"
  };

  // Send ranking request
  var result = await pinecone.Inference.RerankAsync(
      new RerankRequest
      {
          Model = "bge-reranker-v2-m3",
          Query = "What is AAPL's outlook, considering both product launches and market conditions?",
          Documents = documents,
          RankFields = rankFields,
          TopN = 2,
          ReturnDocuments = true,
          Parameters = parameters
      });

  Console.WriteLine(result);
  ```

  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl https://api.pinecone.io/rerank \
    -H "Content-Type: application/json" \
    -H "Accept: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -d '{
    "model": "bge-reranker-v2-m3",
    "query": "What is AAPL'\''s outlook, considering both product launches and market conditions?",
    "documents": [
      {"id": "vec2", "chunk_text": "Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market."},
      {"id": "vec3", "chunk_text": "AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production."},
      {"id": "vec1", "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones."}
    ],
    "top_n": 2,
    "rank_fields": ["chunk_text"],
    "return_documents": true,
    "parameters": {
      "truncate": "END"
    }
  }'
  ```
</CodeGroup>

The response looks as follows. For each hit, the \_score represents the relevance of a document to the query, normalized between 0 and 1, with scores closer to 1 indicating higher relevance.

<CodeGroup>
  ```python Python theme={null}
  RerankResult(
    model='bge-reranker-v2-m3',
    data=[{
      index=0,
      score=0.004166256,
      document={
          id='vec2',
          chunk_text="Analysts suggest that AAPL'''s upcoming Q4 product launch event might solidify its position in the premium smartphone market."
      }
    },{
      index=2,
      score=0.0011513996,
      document={
          id='vec1',
          chunk_text='AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.'
      }
    }],
    usage={'rerank_units': 1}
  )
  ```

  ```javascript JavaScript theme={null}
  {
    model: 'bge-reranker-v2-m3',
    data: [
      { index: 0, score: 0.004166256, document: [id: 'vec2', chunk_text: "Analysts suggest that AAPL'''s upcoming Q4 product launch event might solidify its position in the premium smartphone market."] },
      { index: 2, score: 0.0011513996, document: [id: 'vec1', chunk_text: 'AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.'] }
    ],
    usage: { rerankUnits: 1 }
  }
  ```

  ```java Java theme={null}
  [class RankedDocument {
      index: 0
      score: 0.0063143647
      document: {id=vec2, chunk_text=Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market.}
      additionalProperties: null
  }, class RankedDocument {
      index: 2
      score: 0.0011513996
      document: {id=vec1, chunk_text=AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.}
      additionalProperties: null
  }]
  ```

  ```go Go theme={null}
  {
    "data": [
      {
        "document": {
          "id": "vec2",
          "chunk_text": "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market."
        },
        "index": 0,
        "score": 0.0063143647
      },
      {
        "document": {
          "id": "vec1",
          "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones."
        },
        "index": 2,
        "score": 0.0011513996
      }
    ],
    "model": "bge-reranker-v2-m3",
    "usage": {
      "rerank_units": 1
    }
  }
  ```

  ```csharp C# theme={null}
  {
    "model": "bge-reranker-v2-m3",
    "data": [
      {
        "index": 0,
        "score": 0.006289902,
        "document": {
          "chunk_text": "Analysts suggest that AAPL\u0027s upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
          "id": "vec2"
        }
      },
      {
        "index": 3,
        "score": 0.0011513996,
        "document": {
          "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
          "id": "vec1"
        }
      }
    ],
    "usage": {
      "rerank_units": 1
    }
  }
  ```

  ```json curl theme={null}
  {
      "model": "bge-reranker-v2-m3",
      "data": [
          {
              "index": 0,
              "document": {
                  "chunk_text": "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
                  "id": "vec2"
              },
              "score": 0.007606672
          },
          {
              "index": 3,
              "document": {
                  "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
                  "id": "vec1"
              },
              "score": 0.0013406205
          }
      ],
      "usage": {
          "rerank_units": 1
      }
  }
  ```
</CodeGroup>

## Reranking models

Pinecone hosts several reranking models so it's easy to manage two-stage vector retrieval on a single platform. You can use a hosted model to rerank results as an integrated part of a query, or you can use a hosted model to rerank results as a standalone operation.

The following reranking models are hosted by Pinecone.

<Note>
  To understand how cost is calculated for reranking, see [Reranking cost](/guides/manage-cost/understanding-cost#reranking). To get model details via the API, see [List models](/reference/api/latest/inference/list_models) and [Describe a model](/reference/api/latest/inference/describe_model).
</Note>

<AccordionGroup>
  <Accordion title="cohere-rerank-3.5">
    <PaidOnly />

    [`cohere-rerank-3.5`](/models/cohere-rerank-3.5) is Cohere's leading reranking model, balancing performance and latency for a wide range of enterprise search applications.

    **Details**

    * Modality: Text
    * Max tokens per query and document pair: 40,000
    * Max documents: 200

    For rate limits, see [Rerank requests per minute](/reference/api/database-limits#rerank-requests-per-minute-per-model) and [Rerank request per month](/reference/api/database-limits#rerank-requests-per-month-per-model).

    **Parameters**

    The `cohere-rerank-3.5` model supports the following parameters:

    | Parameter            | Type             | Required/Optional | Description                                                                                                                             |            |
    | :------------------- | :--------------- | :---------------- | :-------------------------------------------------------------------------------------------------------------------------------------- | ---------- |
    | `max_chunks_per_doc` | integer          | Optional          | Long documents will be automatically truncated to the specified number of chunks. Accepted range: `1 - 3072`.                           |            |
    | `rank_fields`        | array of strings | Optional          | The fields to use for reranking. The model reranks based on the order of the fields specified (e.g., `["field1", "field2", "field3"]`). | `["text"]` |
  </Accordion>

  <Accordion title="bge-reranker-v2-m3">
    [`bge-reranker-v2-m3`](/models/bge-reranker-v2-m3) is a high-performance, multilingual reranking model that works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs).

    **Details**

    * Modality: Text
    * Max tokens per query and document pair: 1024
    * Max documents: 100

    For rate limits, see [Rerank requests per minute](/reference/api/database-limits#rerank-requests-per-minute-per-model) and [Rerank request per month](/reference/api/database-limits#rerank-requests-per-month-per-model).

    **Parameters**

    The `bge-reranker-v2-m3` model supports the following parameters:

    | Parameter     | Type             | Required/Optional | Description                                                                                                                                                                                                                                    | Default    |
    | :------------ | :--------------- | :---------------- | :--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | :--------- |
    | `truncate`    | string           | Optional          | How to handle inputs longer than those supported by the model. Accepted values: `END` or `NONE`.<br /><br />`END` truncates the input sequence at the input token limit. `NONE` returns an error when the input exceeds the input token limit. | `NONE`     |
    | `rank_fields` | array of strings | Optional          | The field to use for reranking. The model supports only a single rerank field.                                                                                                                                                                 | `["text"]` |
  </Accordion>

  <Accordion title="pinecone-rerank-v0">
    <PP />

    [`pinecone-rerank-v0`](/models/pinecone-rerank-v0) is a state of the art reranking model that out-performs competitors on widely accepted benchmarks. It can handle chunks up to 512 tokens (1-2 paragraphs).

    **Details**

    * Modality: Text
    * Max tokens per query and document pair: 512
    * Max documents: 100

    For rate limits, see [Rerank requests per minute](/reference/api/database-limits#rerank-requests-per-minute-per-model) and [Rerank request per month](/reference/api/database-limits#rerank-requests-per-month-per-model).

    **Parameters**

    The `pinecone-rerank-v0` model supports the following parameters:

    | Parameter     | Type             | Required/Optional | Description                                                                                                                                                                                                                                    | Default    |
    | :------------ | :--------------- | :---------------- | :--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | :--------- |
    | `truncate`    | string           | Optional          | How to handle inputs longer than those supported by the model. Accepted values: `END` or `NONE`.<br /><br />`END` truncates the input sequence at the input token limit. `NONE` returns an error when the input exceeds the input token limit. | `END`      |
    | `rank_fields` | array of strings | Optional          | The field to use for reranking. The model supports only a single rerank field.                                                                                                                                                                 | `["text"]` |
  </Accordion>
</AccordionGroup>

---

# Source: https://docs.pinecone.io/reference/api/2025-10/inference/rerank.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Rerank results

> Rerank results according to their relevance to a query.

For guidance and examples, see [Rerank results](https://docs.pinecone.io/guides/search/rerank-results).

<RequestExample>
  ```shell curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl https://api.pinecone.io/rerank \
    -H "Content-Type: application/json" \
    -H "Accept: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -H "Api-Key: $PINECONE_API_KEY" \
  -d '{
    "model": "bge-reranker-v2-m3",
    "query": "The tech company Apple is known for its innovative products like the iPhone.",
    "return_documents": true,
    "top_n": 4,
    "documents": [
      {"id": "vec1", "text": "Apple is a popular fruit known for its sweetness and crisp texture."},
      {"id": "vec2", "text": "Many people enjoy eating apples as a healthy snack."},
      {"id": "vec3", "text": "Apple Inc. has revolutionized the tech industry with its sleek designs and user-friendly interfaces."},
      {"id": "vec4", "text": "An apple a day keeps the doctor away, as the saying goes."}
    ],
    "parameters": {
      "truncate": "END"
    }
  }'
  ```
</RequestExample>

<ResponseExample>
  ```JSON curl theme={null}
  {
    "data":[
      {
        "index":2,
        "document":{
          "id":"vec3",
          "text":"Apple Inc. has revolutionized the tech industry with its sleek designs and user-friendly interfaces."
        },
        "score":0.47654688
      },
      {
        "index":0,
        "document":{
          "id":"vec1",
          "text":"Apple is a popular fruit known for its sweetness and crisp texture."
        },
        "score":0.047963805
      },
      {
        "index":3,
        "document":{
          "id":"vec4",
          "text":"An apple a day keeps the doctor away, as the saying goes."
        },
        "score":0.007587992
      },
      {
        "index":1,
        "document":{
          "id":"vec2",
          "text":"Many people enjoy eating apples as a healthy snack."
        },
        "score":0.0006491712
      }
    ],
    "usage":{
      "rerank_units":1
    }
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/inference_2025-10.oas.yaml post /rerank
openapi: 3.0.3
info:
  title: Pinecone Inference API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Inference
    description: Model inference
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /rerank:
    post:
      tags:
        - Inference
      summary: Rerank results
      description: >-
        Rerank results according to their relevance to a query.


        For guidance and examples, see [Rerank
        results](https://docs.pinecone.io/guides/search/rerank-results).
      operationId: rerank
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        description: Rerank documents for the given query
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/RerankRequest'
      responses:
        '200':
          description: OK
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/RerankResult'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-metric-validation-error:
                  summary: Validation error
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Bad request. The request body included invalid request
                        parameters.
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    RerankRequest:
      type: object
      properties:
        model:
          example: bge-reranker-v2-m3
          description: >-
            The
            [model](https://docs.pinecone.io/guides/search/rerank-results#reranking-models)
            to use for reranking.
          type: string
        query:
          example: What is the capital of France?
          description: The query to rerank documents against.
          type: string
        top_n:
          example: 5
          description: >-
            The number of results to return sorted by relevance. Defaults to the
            number of inputs.
          type: integer
        return_documents:
          example: true
          description: Whether to return the documents in the response.
          default: true
          type: boolean
        rank_fields:
          description: >
            The field(s) to consider for reranking. If not provided, the default
            is `["text"]`.


            The number of fields supported is
            [model-specific](https://docs.pinecone.io/guides/search/rerank-results#reranking-models).
          default:
            - text
          type: array
          items:
            type: string
        documents:
          description: The documents to rerank.
          type: array
          items:
            $ref: '#/components/schemas/Document'
        parameters:
          example:
            truncate: END
          description: >-
            Additional model-specific parameters. Refer to the [model
            guide](https://docs.pinecone.io/guides/search/rerank-results#reranking-models)
            for available model parameters.
          type: object
          additionalProperties: true
      required:
        - model
        - documents
        - query
    RerankResult:
      description: The result of a reranking request.
      type: object
      properties:
        model:
          example: bge-reranker-v2-m3
          description: The model used to rerank documents.
          type: string
        data:
          description: The reranked documents.
          type: array
          items:
            $ref: '#/components/schemas/RankedDocument'
        usage:
          description: Usage statistics for the model inference.
          type: object
          properties:
            rerank_units:
              example: 1
              description: The number of rerank units consumed by this operation.
              type: integer
              format: int32
              minimum: 0
      required:
        - model
        - data
        - usage
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, or `FORBIDDEN`.
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              description: A human-readable error message describing the error.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
    Document:
      example:
        id: '1'
        text: Paris is the capital of France.
        title: France
        url: https://example.com
      description: Document for reranking
      type: object
      additionalProperties: true
    RankedDocument:
      description: A ranked document with a relevance score and an index position.
      type: object
      properties:
        index:
          description: The index position of the document from the original request.
          type: integer
        score:
          example: 0.5
          description: >-
            The relevance of the document to the query, normalized between 0 and
            1, with scores closer to 1 indicating higher relevance.
          type: number
        document:
          $ref: '#/components/schemas/Document'
      required:
        - index
        - score
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/guides/indexes/pods/restore-a-pod-based-index.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Restore a pod-based index

> Restore pod-based indexes from collections

<Warning>
  Customers who sign up for a Standard or Enterprise plan on or after August 18, 2025 cannot create pod-based indexes. Instead, create [serverless indexes](/guides/index-data/create-an-index), and consider using [dedicated read nodes](/guides/index-data/dedicated-read-nodes) for large workloads (millions of records or more, and moderate or high query rates).
</Warning>

You can restore a pod-based index by creating a new index from a [collection](/guides/indexes/pods/understanding-collections).

## Create a pod-based index from a collection

To create a pod-based index from a [collection](/guides/manage-data/back-up-an-index#pod-based-index-backups-using-collections), use the [`create_index`](/reference/api/latest/control-plane/create_index) endpoint and provide a [`source_collection`](/reference/api/latest/control-plane/create_index#!path=source%5Fcollection\&t=request) parameter containing the name of the collection from which you wish to create an index. The new index can differ from the original source index: the new index can have a different name, number of pods, or pod type. The new index is queryable and writable.

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone, PodSpec

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.create_index(
    name="docs-example",
    dimension=128,
    metric="cosine",
    spec=PodSpec(
      environment="us-west-1-gcp",
      pod_type="p1.x1",
      pods=1,
      source_collection="example-collection"
    )
  )
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({
    apiKey: 'YOUR_API_KEY'
  });

  await pc.createIndex({
    name: 'docs-example',
    dimension: 128,
    metric: 'cosine',
    spec: {
      pod: {
        environment: 'us-west-1-gcp',
        podType: 'p1.x1',
        pods: 1,
        sourceCollection: 'example-collection'
      }
    }
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.IndexModel;
  import org.openapitools.db_control.client.model.DeletionProtection;

  public class CreateIndexFromCollectionExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          pc.createPodsIndex("docs-example", 1536, "us-west1-gcp",
                  "p1.x1", "cosine", "example-collection", DeletionProtection.DISABLED);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      indexName := "docs-example"
      metric := pinecone.Dotproduct
      deletionProtection := pinecone.DeletionProtectionDisabled

      idx, err := pc.CreatePodIndex(ctx, &pinecone.CreatePodIndexRequest{
          Name:               indexName,
          Metric:             &metric,
          Dimension:          1536,
          Environment:        "us-east1-gcp",
          PodType:            "p1.x1",
          SourceCollection:   "example-collection",
          DeletionProtection: &deletionProtection,
      })
      if err != nil {
          log.Fatalf("Failed to create pod-based index: %v", idx.Name)
      } else {
          fmt.Printf("Successfully created pod-based index: %v", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var createIndexRequest = await pinecone.CreateIndexAsync(new CreateIndexRequest
  {
      Name = "docs-example",
      Dimension = 1538,
      Metric = MetricType.Cosine,
      Spec = new PodIndexSpec
      {
          Pod = new PodSpec
          {
              Environment = "us-east1-gcp",
              PodType = "p1.x1",
              Pods = 1,
              Replicas = 1,
              Shards = 1,
              SourceCollection = "example-collection",
          }
      },
      DeletionProtection = DeletionProtection.Enabled,
  });
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s "https://api.pinecone.io/indexes" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
           "name": "docs-example",
           "dimension": 128,
           "metric": "cosine",
           "spec": {
              "pod": {
                 "environment": "us-west-1-gcp",
                 "pod_type": "p1.x1",
                 "pods": 1,
                 "source_collection": "example-collection"
              }
           }
        }'
  ```
</CodeGroup>

---

# Source: https://docs.pinecone.io/guides/manage-data/restore-an-index.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Restore an index

> Restore serverless indexes from backup snapshots.

## Create a serverless index from a backup

When restoring a serverless index from backup, you can change the index name, tags, and deletion protection setting. All other properties of the restored index will remain identical to the source index, including cloud and region, dimension and similarity metric, and associated embedding model when restoring an index with [integrated embedding](/guides/index-data/indexing-overview#integrated-embedding).

To [create a serverless index from a backup](/reference/api/latest/control-plane/create_index_from_backup), provide the ID of the backup, the name of the new index, and, optionally, changes to the index tags and deletion protection settings:

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.create_index_from_backup(
      backup_id="a65ff585-d987-4da5-a622-72e19a6ed5f4",
      name="restored-index",
      tags={
          "tag0": "val0", 
          "tag1": "val1"
      },
      deletion_protection="enabled"
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  const response = await pc.createIndexFromBackup({
    backupId: 'a65ff585-d987-4da5-a622-72e19a6ed5f4',
    name: 'restored-index',
    tags: {
      tag0: 'val0',
      tag1: 'val1'
    },
    deletionProtection: 'enabled'
  });

  console.log(response);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.ApiException;
  import org.openapitools.db_control.client.model.*;

  public class CreateIndexFromBackup {
      public static void main(String[] args) throws ApiException {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();

          String backupID = "a65ff585-d987-4da5-a622-72e19a6ed5f4";
          String indexName = "restored-index";

          CreateIndexFromBackupResponse backupResponse = pc.createIndexFromBackup(backupID, indexName);
          System.out.println(backupResponse);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"
  	"time"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
  	ctx := context.Background()

  	pc, err := pinecone.NewClient(pinecone.NewClientParams{
  		ApiKey: "YOUR_API_KEY",
  	})
  	if err != nil {
  		log.Fatalf("Failed to create Client: %v", err)
  	}

  	indexName := "restored-index"
  	restoredIndexTags := pinecone.IndexTags{"restored_on": time.Now().Format("2006-01-02 15:04")}
  	createIndexFromBackupResp, err := pc.CreateIndexFromBackup(ctx, &pinecone.CreateIndexFromBackupParams{
  		BackupId: "e12269b0-a29b-4af0-9729-c7771dec03e3",
  		Name:     indexName,
  		Tags:     &restoredIndexTags,
  	})

  	fmt.Printf(prettifyStruct(createIndexFromBackupResp))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var response = await pinecone.Backups.CreateIndexFromBackupAsync(
      "a65ff585-d987-4da5-a622-72e19a6ed5f4", 
      new CreateIndexFromBackupRequest
      {
          Name = "restored-index",
          Tags = new Dictionary<string, string> 
          { 
              { "tag0", "val0" },
              { "tag1", "val1" }
          },
          DeletionProtection = DeletionProtection.Enabled
      }
  );

  Console.WriteLine(response);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  BACKUP_ID="a65ff585-d987-4da5-a622-72e19a6ed5f4"

  curl "https://api.pinecone.io/backups/$BACKUP_ID/create-index" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -H 'Content-Type: application/json' \
    -d '{
          "name": "restored-index",
          "tags": {
            "tag0": "val0",
            "tag1": "val1"
          },
          "deletion_protection": "enabled"
        }'
  ```
</CodeGroup>

The example returns a response like the following:

<CodeGroup>
  ```python Python theme={null}
  {'deletion_protection': 'enabled',
   'dimension': 1024,
   'embed': {'dimension': 1024,
             'field_map': {'text': 'chunk_text'},
             'metric': 'cosine',
             'model': 'multilingual-e5-large',
             'read_parameters': {'input_type': 'query', 'truncate': 'END'},
             'vector_type': 'dense',
             'write_parameters': {'input_type': 'passage', 'truncate': 'END'}},
   'host': 'example-dense-index-python3-govk0nt.svc.aped-4627-b74a.pinecone.io',
   'metric': 'cosine',
   'name': 'example-dense-index-python3',
   'spec': {'serverless': {'cloud': 'aws', 'region': 'us-east-1'}},
   'status': {'ready': True, 'state': 'Ready'},
   'tags': {'tag0': 'val0', 'tag1': 'val1'},
   'vector_type': 'dense'}
  ```

  ```javascript JavaScript theme={null}
  {
    restoreJobId: 'e9ba8ff8-7948-4cfa-ba43-34227f6d30d4',
    indexId: '025117b3-e683-423c-b2d1-6d30fbe5027f'
  }
  ```

  ```java Java theme={null}
  class CreateIndexFromBackupResponse {
      restoreJobId: e9ba8ff8-7948-4cfa-ba43-34227f6d30d4
      indexId: 025117b3-e683-423c-b2d1-6d30fbe5027f
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  {
    "index_id": "025117b3-e683-423c-b2d1-6d30fbe5027f",
    "restore_job_id": "e9ba8ff8-7948-4cfa-ba43-34227f6d30d4"
  }
  ```

  ```csharp C# theme={null}
  {
      "restore_job_id":"e9ba8ff8-7948-4cfa-ba43-34227f6d30d4",
      "index_id":"025117b3-e683-423c-b2d1-6d30fbe5027f"
  }
  ```

  ```json curl theme={null}
  {
      "restore_job_id":"e9ba8ff8-7948-4cfa-ba43-34227f6d30d4",
      "index_id":"025117b3-e683-423c-b2d1-6d30fbe5027f"
  }
  ```
</CodeGroup>

<Tip>
  You can create a serverless index from a backup using the [Pinecone console](https://app.pinecone.io/organizations/-/projects).
</Tip>

## List restore jobs

You can [list all restore jobs](/reference/api/latest/control-plane/list_restore_jobs) as follows.

<Note>
  Up to 100 restore jobs are returned at a time by default, in sorted order (bitwise “C” collation). If the `limit` parameter is set, up to that number of restore jobs are returned instead. Whenever there are additional restore jobs to return, the response also includes a `pagination_token` that you can use to get the next batch of jobs. When the response does not include a `pagination_token`, there are no more restore jobs to return.
</Note>

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  restore_jobs = pc.list_restore_jobs()

  print(restore_jobs)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  const restoreJobs = await pc.listRestoreJobs();

  console.log(restoreJobs);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.ApiException;
  import org.openapitools.db_control.client.model.*;

  public class CreateIndexFromBackup {
      public static void main(String[] args) throws ApiException {
          Pinecone pc = new Pinecone.Builder("YOUR_API-KEY").build();

          // List all restore jobs with default pagination limit
          RestoreJobList restoreJobList = pc.listRestoreJobs(null, null);

          // List all restore jobs with pagination limit of 5
          RestoreJobList restoreJobListWithLimit = pc.listRestoreJobs(5);

          // List all restore jobs with pagination limit and token
          RestoreJobList restoreJobListPaginated = pc.listRestoreJobs(5, "eyJza2lwX3Bhc3QiOiIxMDEwMy0=");

          System.out.println(restoreJobList);
          System.out.println(restoreJobListWithLimit);
          System.out.println(restoreJobListPaginated);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
  	ctx := context.Background()

  	pc, err := pinecone.NewClient(pinecone.NewClientParams{
  		ApiKey: "YOUR_API_KEY",
  	})
  	if err != nil {
  		log.Fatalf("Failed to create Client: %v", err)
  	}

  	limit := 2
  	restoreJobs, err := pc.ListRestoreJobs(ctx, &pinecone.ListRestoreJobsParams{
  		Limit: &limit,
  	})
  	if err != nil {
  		log.Fatalf("Failed to list restore jobs: %v", err)
  	}

  	fmt.Printf(prettifyStruct(restoreJobs))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var jobs = await pinecone.RestoreJobs.ListAsync(new ListRestoreJobsRequest());

  Console.WriteLine(jobs);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://api.pinecone.io/restore-jobs" \
  	-H "X-Pinecone-Api-Version: 2025-10" \
  	-H "Api-Key: $PINECONE_API_KEY"
  ```
</CodeGroup>

The example returns a response like the following:

<CodeGroup>
  ```python Python theme={null}
  [{
      "restore_job_id": "06b08366-a0a9-404d-96c2-e791c71743e5",
      "backup_id": "95707edb-e482-49cf-b5a5-312219a51a97",
      "target_index_name": "restored-index",
      "target_index_id": "027aff93-de40-4f48-a573-6dbcd654f961",
      "status": "Completed",
      "created_at": "2025-05-15T13:59:51.439479+00:00",
      "completed_at": "2025-05-15T14:00:09.222998+00:00",
      "percent_complete": 100.0
  }, {
      "restore_job_id": "4902f735-b876-4e53-a05c-bc01d99251cb",
      "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
      "target_index_name": "restored-index2",
      "target_index_id": "027aff93-de40-4f48-a573-6dbcd654f961",
      "status": "Completed",
      "created_at": "2025-05-15T21:06:19.906074+00:00",
      "completed_at": "2025-05-15T21:06:39.360509+00:00",
      "percent_complete": 100.0
  }]
  ```

  ```javascript JavaScript theme={null}
  {
    data: [
      {
        restoreJobId: '69acc1d0-9105-4fcb-b1db-ebf97b285c5e',
        backupId: '8c85e612-ed1c-4f97-9f8c-8194e07bcf71',
        targetIndexName: 'restored-index2',
        targetIndexId: 'e6c0387f-33db-4227-9e91-32181106e56b',
        status: 'Completed',
        createdAt: 2025-05-14T17:25:59.378Z,
        completedAt: 2025-05-14T17:26:23.997Z,
        percentComplete: 100
      },
      {
        restoreJobId: '9857add2-99d4-4399-870e-aa7f15d8d326',
        backupId: '94a63aeb-efae-4f7a-b059-75d32c27ca57',
        targetIndexName: 'restored-index',
        targetIndexId: '0d8aed24-adf8-4b77-8e10-fd674309dc85',
        status: 'Completed',
        createdAt: 2025-04-25T18:14:05.227Z,
        completedAt: 2025-04-25T18:14:11.074Z,
        percentComplete: 100
      }
    ],
    pagination: undefined
  }
  ```

  ```java Java theme={null}
  class RestoreJobList {
      data: [class RestoreJobModel {
          restoreJobId: cf597d76-4484-4b6c-b07c-2bfcac3388aa
          backupId: 0d75b99f-be61-4a93-905e-77201286c02e
          targetIndexName: restored-index
          targetIndexId: 8a810881-1505-46c0-b906-947c048b15f5
          status: Completed
          createdAt: 2025-05-16T20:09:18.700631Z
          completedAt: 2025-05-16T20:11:30.673296Z
          percentComplete: 100.0
          additionalProperties: null
      }, class RestoreJobModel {
          restoreJobId: 4902f735-b876-4e53-a05c-bc01d99251cb
          backupId: 8c85e612-ed1c-4f97-9f8c-8194e07bcf71
          targetIndexName: restored-index2
          targetIndexId: 710cb6e6-bfb4-4bf5-a425-9754e5bbc832
          status: Completed
          createdAt: 2025-05-15T21:06:19.906074Z
          completedAt: 2025-05-15T21:06:39.360509Z
          percentComplete: 100.0
          additionalProperties: null
      }]
      pagination: class PaginationResponse {
          next: eyJsaW1pdCI6Miwib2Zmc2V0IjoyfQ==
          additionalProperties: null
      }
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  {
    "data": [
      {
        "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
        "completed_at": "2025-05-16T20:11:30.673296Z",
        "created_at": "2025-05-16T20:09:18.700631Z",
        "percent_complete": 100,
        "restore_job_id": "e9ba8ff8-7948-4cfa-ba43-34227f6d30d4",
        "status": "Completed",
        "target_index_id": "025117b3-e683-423c-b2d1-6d30fbe5027f",
        "target_index_name": "restored-index"
      },
      {
        "backup_id": "95707edb-e482-49cf-b5a5-312219a51a97",
        "completed_at": "2025-05-15T21:04:34.2463Z",
        "created_at": "2025-05-15T21:04:15.949067Z",
        "percent_complete": 100,
        "restore_job_id": "eee4f8b8-cd3e-45fe-9ed5-93c28e237f24",
        "status": "Completed",
        "target_index_id": "5a0d555f-7ccd-422a-a3a6-78f7b73350c0",
        "target_index_name": "restored-index2"
      }
    ],
    "pagination": {
      "next": "eyJsaW1pdCI6MTAsIm9mZnNldCI6MTB9"
    }
  }
  ```

  ```csharp C# theme={null}
  {
    "data": [
      {
        "restore_job_id": "9857add2-99d4-4399-870e-aa7f15d8d326",
        "backup_id": "94a63aeb-efae-4f7a-b059-75d32c27ca57",
        "target_index_name": "restored-index",
        "target_index_id": "0d8aed24-adf8-4b77-8e10-fd674309dc85",
        "status": "Completed",
        "created_at": "2025-04-25T18:14:05.227526Z",
        "completed_at": "2025-04-25T18:14:11.074618Z",
        "percent_complete": 100
      },
      {
        "restore_job_id": "69acc1d0-9105-4fcb-b1db-ebf97b285c5e",
        "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
        "target_index_name": "restored-index2",
        "target_index_id": "e6c0387f-33db-4227-9e91-32181106e56b",
        "status": "Completed",
        "created_at": "2025-05-14T17:25:59.378989Z",
        "completed_at": "2025-05-14T17:26:23.997284Z",
        "percent_complete": 100
      }
    ]
  }
  ```

  ```json curl theme={null}
  {
    "data": [
      {
        "restore_job_id": "9857add2-99d4-4399-870e-aa7f15d8d326",
        "backup_id": "94a63aeb-efae-4f7a-b059-75d32c27ca57",
        "target_index_name": "restored-index",
        "target_index_id": "0d8aed24-adf8-4b77-8e10-fd674309dc85",
        "status": "Completed",
        "created_at": "2025-04-25T18:14:05.227526Z",
        "completed_at": "2025-04-25T18:14:11.074618Z",
        "percent_complete": 100
      },
      {
        "restore_job_id": "69acc1d0-9105-4fcb-b1db-ebf97b285c5e",
        "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
        "target_index_name": "restored-index2",
        "target_index_id": "e6c0387f-33db-4227-9e91-32181106e56b",
        "status": "Completed",
        "created_at": "2025-05-14T17:25:59.378989Z",
        "completed_at": "2025-05-14T17:26:23.997284Z",
        "percent_complete": 100
      }
    ],
    "pagination": null
  }
  ```
</CodeGroup>

## View restore job details

You can [view the details of a specific restore job](/reference/api/latest/control-plane/describe_restore_job), as in the following example:

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  restore_job = pc.describe_restore_job(job_id="9857add2-99d4-4399-870e-aa7f15d8d326")

  print(restore_job)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' })

  const restoreJob = await pc.describeRestoreJob('9857add2-99d4-4399-870e-aa7f15d8d326');

  console.log(restoreJob);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.ApiException;
  import org.openapitools.db_control.client.model.*;

  public class CreateIndexFromBackup {
      public static void main(String[] args) throws ApiException {
          Pinecone pc = new Pinecone.Builder("YOUR_API-KEY").build();

          RestoreJobModel restoreJob = pc.describeRestoreJob("9857add2-99d4-4399-870e-aa7f15d8d326");

          System.out.println(restoreJob);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
  	"context"
  	"encoding/json"
  	"fmt"
  	"log"

  	"github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
  	ctx := context.Background()

  	pc, err := pinecone.NewClient(pinecone.NewClientParams{
  		ApiKey: "YOUR_API_KEY",
  	})
  	if err != nil {
  		log.Fatalf("Failed to create Client: %v", err)
  	}

  	restoreJob, err := pc.DescribeRestoreJob(ctx, "e9ba8ff8-7948-4cfa-ba43-34227f6d30d4")
  	if err != nil {
  		log.Fatalf("Failed to describe restore job: %v", err)
  	}

  	fmt.Printf(prettifyStruct(restoreJob))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var job = await pinecone.RestoreJobs.GetAsync("9857add2-99d4-4399-870e-aa7f15d8d326");

  Console.WriteLine(job);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  JOB_ID="9857add2-99d4-4399-870e-aa7f15d8d326"

  curl "https://api.pinecone.io/restore-jobs/$JOB_ID" \
      -H "X-Pinecone-Api-Version: 2025-10" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H 'accept: application/json'
  ```
</CodeGroup>

The example returns a response like the following:

<CodeGroup>
  ```python Python theme={null}
  {'backup_id': '94a63aeb-efae-4f7a-b059-75d32c27ca57',
   'completed_at': datetime.datetime(2025, 4, 25, 18, 14, 11, 74618, tzinfo=tzutc()),
   'created_at': datetime.datetime(2025, 4, 25, 18, 14, 5, 227526, tzinfo=tzutc()),
   'percent_complete': 100.0,
   'restore_job_id': '9857add2-99d4-4399-870e-aa7f15d8d326',
   'status': 'Completed',
   'target_index_id': '0d8aed24-adf8-4b77-8e10-fd674309dc85',
   'target_index_name': 'restored-index'}
  ```

  ```javascript JavaScript theme={null}
  {
    restoreJobId: '9857add2-99d4-4399-870e-aa7f15d8d326',
    backupId: '94a63aeb-efae-4f7a-b059-75d32c27ca57',
    targetIndexName: 'restored-index',
    targetIndexId: '0d8aed24-adf8-4b77-8e10-fd674309dc85',
    status: 'Completed',
    createdAt: 2025-04-25T18:14:05.227Z,
    completedAt: 2025-04-25T18:14:11.074Z,
    percentComplete: 100
  }
  ```

  ```java Java theme={null}
  class RestoreJobModel {
      restoreJobId: cf597d76-4484-4b6c-b07c-2bfcac3388aa
      backupId: 0d75b99f-be61-4a93-905e-77201286c02e
      targetIndexName: restored-index
      targetIndexId: 0d8aed24-adf8-4b77-8e10-fd674309dc85
      status: Completed
      createdAt: 2025-05-16T20:09:18.700631Z
      completedAt: 2025-05-16T20:11:30.673296Z
      percentComplete: 100.0
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  {
    "backup_id": "8c85e612-ed1c-4f97-9f8c-8194e07bcf71",
    "completed_at": "2025-05-16T20:11:30.673296Z",
    "created_at": "2025-05-16T20:09:18.700631Z",
    "percent_complete": 100,
    "restore_job_id": "e9ba8ff8-7948-4cfa-ba43-34227f6d30d4",
    "status": "Completed",
    "target_index_id": "025117b3-e683-423c-b2d1-6d30fbe5027f",
    "target_index_name": "restored-index"
  }
  ```

  ```csharp C# theme={null}
  {
    "restore_job_id": "9857add2-99d4-4399-870e-aa7f15d8d326",
    "backup_id": "94a63aeb-efae-4f7a-b059-75d32c27ca57",
    "target_index_name": "restored-index",
    "target_index_id": "0d8aed24-adf8-4b77-8e10-fd674309dc85",
    "status": "Completed",
    "created_at": "2025-04-25T18:14:05.227526Z",
    "completed_at": "2025-04-25T18:14:11.074618Z",
    "percent_complete": 100
  }
  ```

  ```json curl theme={null}
  {
    "restore_job_id": "9857add2-99d4-4399-870e-aa7f15d8d326",
    "backup_id": "94a63aeb-efae-4f7a-b059-75d32c27ca57",
    "target_index_name": "restored-index",
    "target_index_id": "0d8aed24-adf8-4b77-8e10-fd674309dc85",
    "status": "Completed",
    "created_at": "2025-04-25T18:14:05.227526Z",
    "completed_at": "2025-04-25T18:14:11.074618Z",
    "percent_complete": 100
  }
  ```
</CodeGroup>

---

# Source: https://docs.pinecone.io/troubleshooting/restrictions-on-index-names.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Restrictions on index names

There are two main restrictions on index names in Pinecone: **character restrictions** and a **maximum length**.

## Character restrictions

Index names can only use UTF-8 lowercase alphanumeric Latin characters and dashes. Non-Latin characters (such as Chinese or Cyrillic) and emojis are not supported. Additionally, they cannot contain dots, as these are used to separate hosts and subnets in DNS, which Pinecone uses to route requests and queries.

## Maximum length

The maximum length of your index name is a factor of limits imposed by the infrastructure Pinecone uses behind the scenes. The combination of your index name and project ID (normally a seven-character, alphanumeric string) cannot exceed 52 characters, plus a dash to separate them. Your project ID is different from your project name, which is often longer than seven characters. You can identify your project ID by the hostname used to connect to your index; it's the last set of characters after the final `-`. For example, if your index is `foo` and your project ID is `abc1234` in the `us-east1-gcp` environment, your index's hostname would be `foo-abc1234.svc.us-east1-gcp.pinecone.io`, and its length would be 11 characters (3 for the index name, 1 for the dash, 7 for the project ID).

---

# Source: https://docs.pinecone.io/guides/assistant/retrieve-context-snippets.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Retrieve context snippets

> Access relevant context and citations from Pinecone Assistant.

This page shows you how to [retrieve context snippets](/guides/assistant/context-snippets-overview).

<Tip>
  To try this in your browser, use the [Pinecone Assistant - Context colab notebook](https://colab.research.google.com/drive/1AD4QWsXBG1FQRwq-ModlaggR7Cx7NJCz).
</Tip>

## Retrieve context snippets from an assistant

You can [retrieve context snippets](/reference/api/latest/assistant/context_assistant) from an assistant, as in the following example:

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")
  assistant = pc.assistant.Assistant(assistant_name="example-assistant")

  response = assistant.context(query="Who is the CFO of Netflix?")

  for snippet in response.snippets:
      print(snippet)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  const response = await assistant.context({
    query: 'Who is the CFO of Netflix?',
  });
  console.log(response);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME/context" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "accept: application/json" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "query": "Who is the CFO of Netflix?"
  }'
  ```
</CodeGroup>

The example above returns a JSON object like the following:

```json JSON theme={null}
{
    "snippets":
    [
        {
            "type":"text",
            "content":"EXHIBIT 31.3\nCERTIFICATION OF CHIEF FINANCIAL OFFICER\nPURSUANT TO SECTION 302 OF THE SARBANES-OXLEY ACT OF 2002\nI, Spencer Neumann, certify that: ..."
            "score":0.9960699,
            "reference":
            {
                "type":"pdf",
                "file":
                {
                    "status":"Available",
                    "id":"e6034e51-0bb9-4926-84c6-70597dbd07a7",
                    "name":"Netflix-10-K-01262024.pdf",
                    "size":1073470,
                    "metadata":null,
                    "updated_on":"2024-11-21T22:59:10.426001030Z",
                    "created_on":"2024-11-21T22:58:35.879120257Z", 
                    "percent_done":1.0,
                    "signed_url":"https://storage.googleapis.com...",
                    "error_message":null
                    },
                "pages":[78]
            }
        },
{
    "type":"text",
    "content":"EXHIBIT 32.1\n..."
...
```

<Warning>
  [`signed_url`](https://cloud.google.com/storage/docs/access-control/signed-urls) provides temporary, read-only access to the relevant file. Anyone with the link can access the file, so treat it as sensitive data. Expires in one hour.
</Warning>

## Control the snippets retrieved

<Note>
  This is available in API versions `2025-04` and later.
</Note>

You can limit [token usage](/guides/assistant/pricing-and-limits#token-usage) by tuning `top_k * snippet_size`:

* `snippet_size`: Controls the max size of a snippet (default is 2048 tokens). Note that snippet size can vary and, in rare cases, may be bigger than the set `snippet_size`. Snippet size controls the amount of context given for each chunk of text.
* `top_k`: Controls the max number of context snippets retrieved (default is 16). `top_k` controls the diversity of information received in the returned snippets.

While additional tokens will be used for other parameters, adjusting the `top_k` and `snippet_size` can help manage token consumption.

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")
  assistant = pc.assistant.Assistant(assistant_name="example-assistant")

  response = assistant.context(query="Who is the CFO of Netflix?", top_k=10, snippet_size=2500)

  for snippet in response.snippets:
      print(snippet)
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl "https://prod-1-data.ke.pinecone.io/assistant/chat/$ASSISTANT_NAME/context" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "accept: application/json" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "query": "Who is the CFO of Netflix?",
      "top_k": 10,
      "snippet_size": 2500
  }'
  ```
</CodeGroup>

---

# Source: https://docs.pinecone.io/troubleshooting/return-all-vectors-in-an-index.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Return all vectors in an index

Pinecone is designed to find vectors that are similar to a given set of conditions, either by comparing a new vector to the ones in the index or by comparing a vector in the index to all of the others using the [query by ID feature](/reference/api/2024-10/data-plane/query). Because the Pinecone query function relies on performing this similarity search, there isn't a way to return all of the vectors currently stored in the index with a single query.

There isn't a guaranteed workaround for this type of query today but providing the ability to query all or export the entire index is on our roadmap for the future.

---

# Source: https://docs.pinecone.io/reference/rust-sdk.md

# Rust SDK

<Warning>
  The Rust SDK is in "alpha" and is under active development. The SDK should be considered unstable and should not be used in production. Before a 1.0 release, there are no guarantees of backward compatibility between minor versions. See the [Rust SDK README](https://github.com/pinecone-io/pinecone-rust-client/blob/main/README.md) for full installation instructions and usage examples.

  To make a feature request or report an issue, please [file an issue](https://github.com/pinecone-io/pinecone-rust-client/issues).
</Warning>

## Install

To install the latest version of the [Rust SDK](https://github.com/pinecone-io/pinecone-rust-client), add a dependency to the current project:

```shell  theme={null}
cargo add pinecone-sdk
```

## Initialize

Once installed, you can import the SDK and then use an [API key](/guides/production/security-overview#api-keys) to initialize a client instance:

```rust Rust theme={null}
use pinecone_sdk::pinecone::PineconeClientConfig;   
use pinecone_sdk::utils::errors::PineconeError;
	
#[tokio::main]
async fn main() -> Result<(), PineconeError> {
    let config = PineconeClientConfig {
        api_key: Some("YOUR_API_KEY".to_string()),
        ..Default::default()
    };

    let pinecone = config.client()?;
    let indexes = pinecone.list_indexes().await?;

    println!("Indexes: {:?}", indexes);

    Ok(())
}
```

---

# Source: https://docs.pinecone.io/examples/sample-apps.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Sample apps

export const Tag = ({text, icon}) => {
  return <span className="card-tag">
            {icon && <img src={icon} className={`w-4 h-4 object-contain ${icon.includes("openai") ? "dark-inverted" : ""}`} />}
            {text}
        </span>;
};

export const UtilityExampleCard = ({title, text, link}) => {
  return <a href={link} className="example-card group">
            <h2 className="font-bold" style={{
    fontSize: "0.875rem"
  }}>{title}</h2>
            <p style={{
    fontSize: "0.875rem",
    marginTop: "0"
  }}>{text}</p>
        </a>;
};

export const ExampleCard = ({title, text, link, children, arrow, vectors, namespaces}) => {
  return <a href={link} className="example-card group">
            <h2 className="font-semibold text-base">{title}</h2>
            <p>{text}</p>

            {children && <div className="tags">{children}</div>}

            {arrow && <svg xmlns="http://www.w3.org/2000/svg" className="arrow" width="24" height="24" viewBox="0 0 24 24" fill="none">
                <path d="M5.30739 20L4 18.6926L16.8249 5.8677H5.05837V4H20V18.9416H18.1323V7.1751L5.30739 20Z" fill="var(--text-secondary)" />
            </svg>}

            {(vectors || namespaces) && <div className="vectors">
                {vectors && <span>{vectors} vectors</span>}
                {namespaces && <span>{namespaces} namespaces</span>}
            </div>}
        </a>;
};

<div>
  <div className="card-grid not-prose">
    <ExampleCard title="Semantic search" text="The Legal Semantic Search app shows you how to perform semantic search over PDF documents." link="/examples/sample-apps/legal-semantic-search">
      <Tag text="NextJS" icon="/images/examples/nextjs.svg" />

      <Tag text="OpenAI" icon="/images/examples/openai-icon.svg" />
    </ExampleCard>

    <ExampleCard title="Multi-tenant RAG" text="Namespace Notes is a simple multi-tenant RAG app. Upload documents that feed workspaces with context isolated by namespace." link="/examples/sample-apps/namespace-notes">
      <Tag text="NextJS" icon="/images/examples/nextjs.svg" />

      <Tag text="OpenAI" icon="/images/examples/openai-icon.svg" />
    </ExampleCard>

    <ExampleCard title="Multimodal search" text="The Shop the Look app shows you how to build multimodal search across text, images, and videos." link="/examples/sample-apps/shop-the-look">
      <Tag text="NextJS" icon="/images/examples/nextjs.svg" />

      <Tag text="Google Vertex AI" icon="/images/google-cloud-vertex-ai.svg" />
    </ExampleCard>

    <ExampleCard title="Pinecone Assistant" text="The Pinecone Assistant app connects a chat interface to your Pinecone Assistant to answer complex questions on your data." link="/examples/sample-apps/pinecone-assistant">
      <Tag text="NextJS" icon="/images/examples/nextjs.svg" />
    </ExampleCard>
  </div>

  <h2 className="examples-h2">More code examples</h2>

  <div className="card-grid not-prose">
    <UtilityExampleCard title="Use the Python SDK with FastAPI" text="A FastAPI app to demonstrate how to use the Pinecone Python SDK with asyncio support." link="https://github.com/pinecone-io/fastapi-pinecone-async-example" />

    <UtilityExampleCard title="Implement semantic search with TypeScript" text="A simple semantic search app written in TypeScript." link="https://github.com/pinecone-io/semantic-search-example" />

    <UtilityExampleCard title="Build an article recommender with TypeScript" text="A simple article recommender app written in TypeScript." link="https://github.com/pinecone-io/recommender-example-typescript" />

    <UtilityExampleCard title="Create a chatbot agent with LangChain" text="A conversational agent built with LangChain and TypeScript." link="https://github.com/pinecone-io/langchain-retrieval-agent-example" />

    <UtilityExampleCard title="Implement image search with TypeScript" text="An image search app written in TypeScript." link="https://github.com/pinecone-io/image-search-example" />
  </div>
</div>

---

# Source: https://docs.pinecone.io/guides/indexes/pods/scale-pod-based-indexes.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Scale pod-based indexes

> Scale indexes vertically or horizontally as needed

<Warning>
  Customers who sign up for a Standard or Enterprise plan on or after August 18, 2025 cannot create pod-based indexes. Instead, create [serverless indexes](/guides/index-data/create-an-index), and consider using [dedicated read nodes](/guides/index-data/dedicated-read-nodes) for large workloads (millions of records or more, and moderate or high query rates).
</Warning>

While your index can still serve queries, new upserts may fail as the capacity becomes exhausted. If you need to scale your environment to accommodate more vectors, you can modify your existing index and scale it vertically or create a new index and scale horizontally.

This page explains how you can scale your [pod-based indexes](/guides/index-data/indexing-overview#pod-based-indexes) horizontally and vertically.

## Vertical vs. horizontal scaling

If you need to scale your environment to accommodate more vectors, you can modify your existing index to scale it vertically or create a new index and scale horizontally. This article will describe both methods and how to scale your index effectively.

## Vertical scaling

[Vertical scaling](https://www.pinecone.io/learn/testing-p2-collections-scaling/#vertical-scaling-on-p1-and-s1) is fast and involves no downtime. This is a good choice when you can't pause upserts and must continue serving traffic. It also allows you to double your capacity instantly. However, there are some factors to consider.

### Increase pod size

The default [pod size](/guides/index-data/indexing-overview#pod-size-and-performance) is `x1`. You can increase the size to `x2`, `x4`, or `x8`. Moving up to the next size effectively doubles the capacity of the index. If you need to scale by smaller increments, then consider horizontal scaling.

Increasing the pod size of your index does not result in downtime. Reads and writes continue uninterrupted during the scaling process, which completes in about 10 minutes. You cannot reduce the pod size of your indexes.

The number of base pods you specify when you initially create the index is static and cannot be changed. For example, if you start with 10 pods of `p1.x1` and vertically scale to `p1.x2`, this equates to 20 pods worth of usage. Pod types (performance versus storage pods) also cannot be changed with vertical scaling. If you want to change your pod type while scaling, then horizontal scaling is the better option.

#### When to increase pod size

If your index is at around 90% fullness, we recommend increasing its size. This helps ensure optimal performance and prevents upserts from failing due to capacity constraints.

#### How to increase pod size

You can increase the pod size in the Pinecone console or using the API.

<Tabs>
  <Tab title="Console">
    1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).
    2. Select the project containing the index you want to configure.
    3. Go to **Database > Indexes**.
    4. Select the index.
    5. Click the **...** button.
    6. Select **Configure**.
    7. In the dropdown, choose the pod size to use.
    8. Click **Confirm**.
  </Tab>

  <Tab title="API">
    Use the [`configure_index`](/reference/api/latest/control-plane/configure_index) operation and append the new size to the `pod_type` parameter, separated by a period (.).

    **Example**

    The following example assumes that `docs-example` has size `x1` and increases the size to `x2`.

    <CodeGroup>
      ```Python Python theme={null}
      from pinecone.grpc import PineconeGRPC as Pinecone

      pc = Pinecone(api_key="YOUR_API_KEY")

      pc.configure_index("docs-example", pod_type="s1.x2")
      ```

      ```JavaScript JavaScript theme={null}
      import { Pinecone } from '@pinecone-database/pinecone'

      const pinecone = new Pinecone({
        apiKey: 'YOUR_API_KEY'
      });

      await pc.configureIndex('docs-example', {
        spec: {
          pod: {
            podType: 's1.x2',
          },
        },
      });
      ```

      ```java Java theme={null}
      import io.pinecone.clients.Pinecone;

      public class ConfigureIndexExample {
          public static void main(String[] args) {
              Pinecone pc = new Pinecone.Builder("PINECONE_API_KEY").build();
              pc.configurePodsIndex("docs-example", "s1.x2");
          }
      }
      ```

      ```go Go theme={null}
      package main

      import (
          "context"
          "fmt"
          "log"

          "github.com/pinecone-io/go-pinecone/v4/pinecone"
      )

      func main() {
          ctx := context.Background()

          pc, err := pinecone.NewClient(pinecone.NewClientParams{
              ApiKey: "YOUR_API_KEY",
          })
          if err != nil {
              log.Fatalf("Failed to create Client: %v", err)
          }

          idx, err := pc.ConfigureIndex(ctx, "docs-example", pinecone.ConfigureIndexParams{PodType: "s1.x2"})
        	if err != nil {
              log.Fatalf("Failed to configure index \"%v\": %v", idx.Name, err)
          } else {
              fmt.Printf("Successfully configured index \"%v\"", idx.Name)
          }
      }
      ```

      ```csharp C# theme={null}
      using Pinecone;

      var pinecone = new PineconeClient("YOUR_API_KEY");

      var indexMetadata = await pinecone.ConfigureIndexAsync("docs-example", new ConfigureIndexRequest
      {
          Spec = new ConfigureIndexRequestSpec
          {
              Pod = new ConfigureIndexRequestSpecPod {
                  PodType = "s1.x2",
              }
          }
      });
      ```

      ```bash curl theme={null}
      PINECONE_API_KEY="YOUR_API_KEY"

      curl -s -X PATCH "https://api.pinecone.io/indexes/docs-example-curl" \
        -H "Content-Type: application/json" \
        -H "Api-Key: $PINECONE_API_KEY" \
        -H "X-Pinecone-Api-Version: 2025-10" \
        -d '{
               "pod_type": "s1.x2"
            }'
      ```
    </CodeGroup>
  </Tab>
</Tabs>

<Note>
  The size change can take up to 15 minutes to complete.
</Note>

### Decrease pod size

After creating an index, you cannot vertically downscale the index/pod size. Instead, you must [create a collection](/guides/indexes/pods/back-up-a-pod-based-index) and then [create a new index from your collection](/guides/indexes/pods/restore-a-pod-based-index) and specify your desired pod size.

### Check the status of a pod size change

To check the status of a pod size change, use the [`describe_index`](/reference/api/latest/control-plane/describe_index/) endpoint. The `status` field in the results contains the key-value pair `"state":"ScalingUp"` or `"state":"ScalingDown"` during the resizing process and the key-value pair `"state":"Ready"` after the process is complete.

The index fullness metric provided by [`describe_index_stats`](/reference/api/latest/data-plane/describeindexstats) may be inaccurate until the resizing process is complete.

**Example**

The following example uses `describe_index` to get the index status of the index `docs-example`. The `status` field contains the key-value pair `"state":"ScalingUp"`, indicating that the resizing process is still ongoing.

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.describe_index(name="docs-example")
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({
    apiKey: 'YOUR_API_KEY'
  });

  await pc.describeIndex({
    name: "docs-example",
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.*;

  public class DescribeIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          IndexModel indexModel = pc.describeIndex("docs-example");
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idx, err := pc.DescribeIndex(ctx, "docs-example")
    	if err != nil {
          log.Fatalf("Failed to describe index %v: %v", idx.Name, err)
      } else {
          fmt.Printf("Successfully found index: %v", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var indexModel = await pinecone.DescribeIndexAsync("docs-example");

  Console.WriteLine(indexModel);
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s -X GET "https://api.pinecone.io/indexes/docs-example-curl" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
  ```
</CodeGroup>

## Horizontal scaling

There are two approaches to horizontal scaling in Pinecone: adding pods and adding replicas. Adding pods increases all resources but requires a pause in upserts; adding replicas only increases throughput and requires no pause in upserts.

### Add pods

Adding additional pods to a running index is not supported directly. However, you can increase the number of pods by using our [collections](/guides/indexes/pods/understanding-collections) feature to create a new index with more pods.

A collection is an immutable snapshot of your index in time: a collection stores the data but not the original index configuration. When you [create an index from a collection](/guides/indexes/pods/create-a-pod-based-index#create-a-pod-index-from-a-collection), you define the new index configuration. This allows you to scale the base pod count horizontally without scaling vertically.

The main advantage of this approach is that you can scale incrementally instead of doubling capacity as with vertical scaling. Also, you can redefine pod types if you are experimenting or if you need to use a different pod type, such as performance-optimized pods or storage-optimized pods. Another advantage of this method is that you can change your [metadata configuration](/guides/indexes/pods/manage-pod-based-indexes#selective-metadata-indexing) to redefine metadata fields as indexed or stored-only. This is important when tuning your index for the best throughput.

Here are the general steps to make a copy of your index and create a new index while changing the pod type, pod count, metadata configuration, replicas, and all typical parameters when creating a new collection:

1. Pause upserts.
2. Create a collection from the current index.
3. Create an index from the collection with new parameters.
4. Continue upserts to the newly created index. Note: the URL has likely changed.
5. Delete the old index if desired.

For detailed steps on creating the collection, see [backup indexes](/guides/manage-data/back-up-an-index#create-a-backup-using-a-collection). For steps on creating an index from a collection, see [Create an index from a collection](/guides/indexes/pods/create-a-pod-based-index#create-a-pod-index-from-a-collection).

### Add replicas

Each replica duplicates the resources and data in an index. This means that adding additional replicas increases the throughput of the index but not its capacity. However, adding replicas does not require downtime.

Throughput in terms of queries per second (QPS) scales linearly with the number of replicas per index.

#### When to add replicas

There are two primary scenarios where adding replicas is beneficial:

**Increase QPS**: The primary reason to add replicas is to increase your index's queries per second (QPS). Each new replica adds another pod for reading from your index and, generally speaking, will increase your QPS by an equal amount as a single pod. For example, if you consistently get 25 QPS for a single pod, each replica will result in 25 more QPS.

If you don't see an increase in QPS after adding replicas, add multiprocessing to your application to ensure you are running parallel operations. You can use the [Pinecone gRPC SDK](/guides/index-data/upsert-data#grpc-python-sdk), or your multiprocessing library of choice.

**Provide data redundancy**: When you add a replica to your index, the Pinecone controller will choose a zone in the same region that does not currently have a replica, up to a maximum of three zones (your fourth and subsequent replicas will be hosted in zones with existing replicas). If your application requires multizone redundancy, this is our recommended approach to achieve that.

#### How to add replicas

To add replicas, use the `configure_index` endpoint to increase the number of replicas for your index:

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.configure_index("docs-example", replicas=4)
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({
    apiKey: 'YOUR_API_KEY'
  });

  await pc.configureIndex('docs-example', {
    spec: {
      pod: {
        replicas: 4,
      },
    },
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;

  public class ConfigureIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("PINECONE_API_KEY").build();
          pc.configurePodsIndex("docs-example", 4, DeletionProtection.DISABLED);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idx, err := pc.ConfigureIndex(ctx, "docs-example", pinecone.ConfigureIndexParams{Replicas: 4})
    	if err != nil {
          log.Fatalf("Failed to configure index \"%v\": %v", idx.Name, err)
      } else {
          fmt.Printf("Successfully configured index \"%v\"", idx.Name)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var configureIndexRequest = await pinecone.ConfigureIndexAsync("docs-example", new ConfigureIndexRequest
  {
      Spec = new ConfigureIndexRequestSpec
      {
          Pod = new ConfigureIndexRequestSpecPod {
              Replicas = 4,
          }
      }
  });
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -s -X PATCH "https://api.pinecone.io/indexes/docs-example-curl" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
           "replicas": 4
        }'
  ```
</CodeGroup>

## Next steps

* See our learning center for more information on [vertical scaling](https://www.pinecone.io/learn/testing-p2-collections-scaling/#vertical-scaling-on-p1-and-s1).
* Learn more about [collections](/guides/indexes/pods/understanding-collections).

---

# Source: https://docs.pinecone.io/guides/assistant/quickstart/sdk-quickstart.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Pinecone Assistant: SDK quickstart

> Use a Pinecone SDK to create an assistant, upload documents, and chat with the assistant.

Use a Pinecone SDK to create an assistant, upload documents, and chat with the assistant.

<Tip>
  To get started in your browser, use the [Assistant Quickstart colab notebook](https://colab.research.google.com/github/pinecone-io/examples/blob/master/docs/assistant-quickstart.ipynb).
</Tip>

## 1. Install an SDK

The Pinecone [Python SDK](/reference/sdks/python/overview) and [Node.js SDK](/reference/sdks/node/overview) provide convenient programmatic access to the [Assistant API](/reference/api/latest/assistant/).

<CodeGroup>
  ```shell Python theme={null}
  pip install pinecone
  pip install pinecone-plugin-assistant
  ```

  ```shell JavaScript theme={null}
  npm install @pinecone-database/pinecone
  ```
</CodeGroup>

## 2. Get an API key

You need an API key to make calls to your assistant.

Create a new API key in the [Pinecone console](https://app.pinecone.io/organizations/-/keys), or use the widget below to generate a key. If you don't have a Pinecone account, the widget will sign you up for the free [Starter plan](https://www.pinecone.io/pricing/).

<div style={{minWidth: '450px', minHeight:'152px'}}>
  <div id="pinecone-connect-widget">
    <div class="connect-widget-skeleton">
      <div class="skeleton-content" />
    </div>
  </div>
</div>

Your generated API key:

```shell  theme={null}
"{{YOUR_API_KEY}}"
```

## 3. Create an assistant

[Create an assistant](/reference/api/latest/assistant/create_assistant), as in the following example:

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="{{YOUR_API_KEY}}")

  assistant = pc.assistant.create_assistant(
      assistant_name="example-assistant", 
      instructions="Use American English for spelling and grammar.", # Description or directive for the assistant to apply to all responses.
      region="us", # Region to deploy assistant. Options: "us" (default) or "eu".
      timeout=30 # Maximum seconds to wait for assistant status to become "Ready" before timing out.
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "{{YOUR_API_KEY}}" });

  const assistant = await pc.createAssistant({
    name: 'example-assistant',
    instructions: 'Use American English for spelling and grammar.', // Description or directive for the assistant to apply to all responses.
    region: 'us'
  });
  ```
</CodeGroup>

## 4. Upload a file to the assistant

With Pinecone Assistant, you can upload documents, ask questions, and receive responses that reference your documents. This is known as retrieval-augmented generation (RAG).

For this quickstart, [download a sample 10-k filing file](https://s22.q4cdn.com/959853165/files/doc_financials/2023/ar/Netflix-10-K-01262024.pdf) to your local device.

Next, [upload the file](/reference/api/latest/assistant/upload_file) to your assistant:

<CodeGroup>
  ```python Python theme={null}
  # Get the assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant", 
  )

  # Upload a file.
  response = assistant.upload_file(
      file_path="/path/to/file/Netflix-10-K-01262024.pdf",
      metadata={"company": "netflix", "document_type": "form 10k"},
      timeout=None
  )
  ```

  ```javascript JavaScript theme={null}
  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);

  await assistant.uploadFile({
    path: '/Users/jdoe/Downloads/example_file.txt'
  });
  ```
</CodeGroup>

## 5. Chat with the assistant

With the sample file uploaded, you can now [chat with the assistant](/reference/api/latest/assistant/chat_assistant). Ask the assistant questions about your document. It returns either a JSON object or a text stream.

<Tip>
  For faster chat responses, use GPT models (`gpt-4o`, `gpt-4.1`, `gpt-5`, or `o4-mini`). You can also enable streaming to improve perceived latency by showing content as it's generated.
</Tip>

The following example requests a default response to the message, "Who is the CFO of Netflix?":

<CodeGroup>
  ```python Python theme={null}
  from pinecone_plugins.assistant.models.chat import Message

  msg = Message(role="user", content="Who is the CFO of Netflix?")
  resp = assistant.chat(messages=[msg])

  print(resp)
  ```

  ```javascript JavaScript theme={null}
  const chatResp = await assistant.chat({
    messages: [{ role: 'user', content: 'Who is the CFO of Netflix?' }],
    model: 'gpt-4o'
  });

  console.log(chatResp);
  ```
</CodeGroup>

The example above returns a response like the following:

```
{
    'id': '0000000000000000163008a05b317b7b', 
    'model': 'gpt-4o-2024-05-13', 
    'usage': {
        'prompt_tokens': 9259, 
        'completion_tokens': 30, 
        'total_tokens': 9289
        }, 
        'message': {
            'content': 'The Chief Financial Officer (CFO) of Netflix is Spencer Neumann.', 
            'role': '"assistant"'
            }, 
            'finish_reason': 'stop', 
            'citations': [
                {
                    'position': 63, 
                    'references': [
                        {
                            'pages': [78, 72, 79], 
                            'file': {
                                'name': 'Netflix-10-K-01262024.pdf', 
                                'id': '76a11dd1...', 
                                'metadata': {
                                    'company': 'netflix', 
                                    'document_type': 'form 10k'
                                    }, 
                                    'created_on': '2024-12-06T01:29:07.369208590Z', 
                                    'updated_on': '2024-12-06T01:29:50.923493799Z', 
                                    'status': 'Available', 
                                    'percent_done': 1.0, 
                                    'signed_url': 'https://storage.googleapis.com/...', 
                                    "error_message": null, 
                                    'size': 1073470.0
                                }
                            }
                        ]
                    }
                ]
            }
```

<Warning>
  [`signed_url`](https://cloud.google.com/storage/docs/access-control/signed-urls) provides temporary, read-only access to the relevant file. Anyone with the link can access the file, so treat it as sensitive data. Expires in one hour.
</Warning>

## 6. Clean up

When you no longer need the `example-assistant`, [delete the assistant](/reference/api/latest/assistant/delete_assistant):

<Warning>
  Deleting an assistant also deletes all files uploaded to the assistant.
</Warning>

<CodeGroup>
  ```python Python theme={null}
  pc.assistant.delete_assistant(
      assistant_name="example-assistant", 
  )
  ```

  ```javascript JavaScript theme={null}
  await pc.deleteAssistant('example-assistant');
  ```
</CodeGroup>

## Next steps

* Learn more about [Pinecone Assistant](/guides/assistant/overview)
* Learn about [additional assistant features](https://www.pinecone.io/learn/assistant-api-deep-dive/)
* [Evaluate](/guides/assistant/evaluate-answers) the assistant's responses
* View a [sample app](/examples/sample-apps/pinecone-assistant) that uses Pinecone Assistant

---

# Source: https://docs.pinecone.io/guides/search/search-overview.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Search overview

> Explore semantic, lexical, and hybrid search options.

## Search types

* [Semantic search](/guides/search/semantic-search)
* [Lexical search](/guides/search/lexical-search)
* [Hybrid search](/guides/search/hybrid-search)

## Optimization

* [Filter by metadata](/guides/search/filter-by-metadata)
* [Rerank results](/guides/search/rerank-results)
* [Parallel queries](/guides/search/semantic-search#parallel-queries)

## Limits

| Metric            | Limit  |
| :---------------- | :----- |
| Max `top_k` value | 10,000 |
| Max result size   | 4MB    |

The query result size is affected by the dimension of the dense vectors and whether or not dense vector values and metadata are included in the result.

<Tip>
  If a query fails due to exceeding the 4MB result size limit, choose a lower `top_k` value, or use `include_metadata=False` or `include_values=False` to exclude metadata or values from the result. For better performance, especially with higher `top_k` values, avoid including vector values unless you need them.
</Tip>

## Cost

* To understand how cost is calculated for queries, see [Understanding cost](/guides/manage-cost/understanding-cost#query).
* For up-to-date pricing information, see [Pricing](https://www.pinecone.io/pricing/).

## Data freshness

Pinecone is eventually consistent, so there can be a slight delay before new or changed records are visible to queries. You can view index stats to [check data freshness](/guides/index-data/check-data-freshness).

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/search_records.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Search with text

> Search a namespace with a query text, query vector, or record ID and return the most similar records, along with their similarity scores. Optionally, rerank the initial results based on their relevance to the query. 

Searching with text is supported only for indexes with [integrated embedding](https://docs.pinecone.io/guides/index-data/indexing-overview#vector-embedding). Searching with a query vector or record ID is supported for all indexes. 

For guidance and examples, see [Search](https://docs.pinecone.io/guides/search/search-overview).

<RequestExample>
  ```shell curl theme={null}
  INDEX_HOST="INDEX_HOST"
  NAMESPACE="YOUR_NAMESPACE"
  PINECONE_API_KEY="YOUR_API_KEY"

  # Search with a query text and rerank the results
  # Supported only for indexes with integrated embedding
  curl "https://$INDEX_HOST/records/namespaces/$NAMESPACE/search" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "query": {
              "inputs": {"text": "Disease prevention"},
              "top_k": 4,
          },
          "fields": ["category", "chunk_text"]
          "rerank": {
              "model": "bge-reranker-v2-m3",
              "top_n": 2,
              "rank_fields": ["chunk_text"] # Specified field must also be included in 'fields'
          }
       }'

  # Search with a query vector and rerank the results
  curl "https://$INDEX_HOST/records/namespaces/$NAMESPACE/search" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "query": {
              "vector": {
                  "values": [0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3]
              },
              "top_k": 4,
          },
          "fields": ["category", "chunk_text"]
          "rerank": {
              "query": "Disease prevention",
              "model": "bge-reranker-v2-m3",
              "top_n": 2,
              "rank_fields": ["chunk_text"] # Specified field must also be included in 'fields'
          }
       }'

  # Search with a record ID and rerank the results
  # Supported only for indexes with integrated embedding
  curl "https://$INDEX_HOST/records/namespaces/$NAMESPACE/search" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "query": {
              "id": "rec1",
              "top_k": 4,
          },
          "fields": ["category", "chunk_text"]
          "rerank": {
              "query": "Disease prevention",
              "model": "bge-reranker-v2-m3",
              "top_n": 2,
              "rank_fields": ["chunk_text"]
          }
       }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
      "result": {
          "hits": [
              {
                  "_id": "rec3",
                  "_score": 0.004433765076100826,
                  "fields": {
                      "category": "immune system",
                      "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases."
                  }
              },
              {
                  "_id": "rec4",
                  "_score": 0.0029121784027665854,
                  "fields": {
                      "category": "endocrine system",
                      "chunk_text": "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes."
                  }
              }
          ]
      },
      "usage": {
          "embed_total_tokens": 8,
          "read_units": 6,
          "rerank_units": 1
      }
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml post /records/namespaces/{namespace}/search
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /records/namespaces/{namespace}/search:
    post:
      tags:
        - Vector Operations
      summary: Search with text
      description: >-
        Search a namespace with a query text, query vector, or record ID and
        return the most similar records, along with their similarity scores.
        Optionally, rerank the initial results based on their relevance to the
        query. 


        Searching with text is supported only for indexes with [integrated
        embedding](https://docs.pinecone.io/guides/index-data/indexing-overview#vector-embedding).
        Searching with a query vector or record ID is supported for all
        indexes. 


        For guidance and examples, see
        [Search](https://docs.pinecone.io/guides/search/search-overview).
      operationId: searchRecordsNamespace
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: namespace
          description: The namespace to search.
          required: true
          schema:
            type: string
          style: simple
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/SearchRecordsRequest'
        required: true
      responses:
        '200':
          description: A successful search namespace response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/SearchRecordsResponse'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    SearchRecordsRequest:
      example:
        fields:
          - chunk_text
        query:
          inputs:
            text: your query text
          top_k: 10
      description: A search request for records in a specific namespace.
      type: object
      properties:
        query:
          description: .
          type: object
          properties:
            top_k:
              example: 10
              description: The number of similar records to return.
              type: integer
              format: int32
            filter:
              description: >-
                The filter to apply. You can use vector metadata to limit your
                search. See [Understanding
                metadata](https://docs.pinecone.io/guides/index-data/indexing-overview#metadata).
              type: object
            inputs:
              $ref: '#/components/schemas/EmbedInputs'
            vector:
              $ref: '#/components/schemas/SearchRecordsVector'
            id:
              example: example-vector-1
              description: The unique ID of the vector to be used as a query vector.
              type: string
              maxLength: 512
            match_terms:
              $ref: '#/components/schemas/SearchMatchTerms'
          required:
            - top_k
        fields:
          example:
            - chunk_text
          description: >-
            The fields to return in the search results. If not specified, the
            response will include all fields.
          type: array
          items:
            type: string
          maxLength: 100
        rerank:
          description: Parameters for reranking the initial search results.
          type: object
          properties:
            model:
              example: bge-reranker-v2-m3
              description: >-
                The name of the [reranking
                model](https://docs.pinecone.io/guides/search/rerank-results#reranking-models)
                to use.
              type: string
            rank_fields:
              example:
                - chunk_text
                - title
              description: >
                The field(s) to consider for reranking. If not provided, the
                default is `["text"]`.


                The number of fields supported is
                [model-specific](https://docs.pinecone.io/guides/search/rerank-results#reranking-models).
              type: array
              items:
                type: string
            top_n:
              example: 5
              description: >-
                The number of top results to return after reranking. Defaults to
                top_k.
              type: integer
              format: int32
            parameters:
              example:
                truncate: END
              description: >-
                Additional model-specific parameters. Refer to the [model
                guide](https://docs.pinecone.io/guides/search/rerank-results#reranking-models)
                for available model parameters.
              type: object
              additionalProperties: true
            query:
              example: What is the capital of France?
              description: >-
                The query to rerank documents against. If a specific rerank
                query is specified,  it overwrites the query input that was
                provided at the top level.
              type: string
          required:
            - model
            - rank_fields
      required:
        - query
    SearchRecordsResponse:
      example:
        result:
          hits:
            - _id: example-record-1
              _score: 0.9281134605407715
              fields:
                data: your example text
        usage:
          embed_total_tokens: 10
          read_units: 5
      description: The records search response.
      type: object
      properties:
        result:
          type: object
          properties:
            hits:
              description: The hits for the search document request.
              type: array
              items:
                $ref: '#/components/schemas/Hit'
          required:
            - hits
        usage:
          $ref: '#/components/schemas/SearchUsage'
      required:
        - usage
        - result
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    EmbedInputs:
      example:
        text: chunk_text
      type: object
    SearchRecordsVector:
      type: object
      properties:
        values:
          $ref: '#/components/schemas/VectorValues'
        sparse_values:
          example:
            - 0.1
            - 0.2
            - 0.3
          description: The sparse embedding values.
          type: array
          items:
            type: number
            format: float
        sparse_indices:
          example:
            - 10
            - 3
            - 156
          description: The sparse embedding indices.
          type: array
          items:
            type: integer
            format: int32
            minimum: 0
    SearchMatchTerms:
      example:
        strategy: all
        terms:
          - animal
          - CHARACTER
          - donald Duck
      description: >-
        Specifies which terms must be present in the text of each search hit
        based on the specified strategy. The match is performed

        against the text field specified in the integrated index `field_map`
        configuration.


        Terms are normalized and tokenized into single tokens before matching,
        and order does not matter.


        Example:

          `"match_terms": {"terms": ["animal", "CHARACTER", "donald Duck"], "strategy": "all"}` will tokenize
          to `["animal", "character", "donald", "duck"]`, and would match
          `"Donald F. Duck is a funny animal character"` but would not match `"A duck is a funny animal"`.

        Match terms filtering is supported only for sparse indexes with
        [integrated
        embedding](https://docs.pinecone.io/guides/index-data/indexing-overview#vector-embedding)

        configured to use the
        [pinecone-sparse-english-v0](https://docs.pinecone.io/models/pinecone-sparse-english-v0)
        model.
      type: object
      properties:
        strategy:
          description: >-
            The strategy for matching terms in the text. Currently, only `all`
            is supported, which means all specified terms must be present.
          x-enum:
            - all
          type: string
        terms:
          description: >-
            A list of terms that must be present in the text of each search hit
            based on the specified strategy.
          type: array
          items:
            type: string
    Hit:
      example:
        _id: example-record-1
        _score: 0.9281134605407715
        fields:
          data: your example text
          more_data:
            text: your example text
      description: A record whose vector values are similar to the provided search query.
      type: object
      properties:
        _id:
          description: The record id of the search hit.
          type: string
        _score:
          description: The similarity score of the returned record.
          type: number
          format: float
        fields:
          description: The selected record fields associated with the search hit.
          type: object
      required:
        - _id
        - _score
        - fields
    SearchUsage:
      type: object
      properties:
        read_units:
          example: 5
          description: The number of read units consumed by this operation.
          type: integer
          format: int32
          minimum: 0
        embed_total_tokens:
          example: 2
          description: The number of embedding tokens consumed by this operation.
          type: integer
          format: int32
          minimum: 0
        rerank_units:
          example: 1
          description: The number of rerank units consumed by this operation.
          type: integer
          format: int32
          minimum: 0
      required:
        - read_units
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
    VectorValues:
      description: This is the vector data included in the request.
      type: array
      items:
        type: number
        format: float
      minLength: 1
      maxLength: 20000
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/guides/production/security-overview.md

# Source: https://docs.pinecone.io/guides/assistant/admin/security-overview.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Security overview

> Understand Pinecone's security features, including authentication, encryption, and audit logs.

This page describes Pinecone's security protocols, practices, and features.

## Access management

### API keys

Each Pinecone [project](/guides/assistant/admin/projects-overview) has one or more [API keys](/guides/assistant/admin/manage-api-keys). In order to make calls to the Pinecone API, a user must provide a valid API key for the relevant Pinecone project.

You can [manage API key permissions](/guides/assistant/admin/manage-api-keys) in the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/keys). The available permission roles are as follows:

#### General permissions

<Tabs>
  <Tab title="Pinecone console">
    | Role | Permissions                                     |
    | :--- | :---------------------------------------------- |
    | All  | Permissions to read and write all project data. |
  </Tab>

  <Tab title="API">
    | Role            | Permissions                                      |
    | :-------------- | :----------------------------------------------- |
    | `ProjectEditor` | Permissions to read  and write all project data. |
    | `ProjectViewer` | Permissions to read all project data.            |
  </Tab>
</Tabs>

#### Control plane permissions

<Tabs>
  <Tab title="Pinecone console">
    | Role      | Permissions                                                                                                 |
    | :-------- | :---------------------------------------------------------------------------------------------------------- |
    | ReadWrite | Permissions to list, describe, create, delete, and configure indexes, backups, collections, and assistants. |
    | ReadOnly  | Permissions to list and describe indexes, backups, collections, and assistants.                             |
    | None      | No control plane permissions.                                                                               |
  </Tab>

  <Tab title="API">
    | Role                 | Permissions                                                                                                 |
    | :------------------- | :---------------------------------------------------------------------------------------------------------- |
    | `ControlPlaneEditor` | Permissions to list, describe, create, delete, and configure indexes, backups, collections, and assistants. |
    | `ControlPlaneViewer` | Permissions to list and describe indexes, backups, collections, and assistants.                             |
    | None                 | No control plane permissions.                                                                               |
  </Tab>
</Tabs>

#### Data plane permissions

<Tabs>
  <Tab title="Pinecone console">
    | Role      | Permissions                                                                                                                                                                                                                                                                                                            |
    | :-------- | :--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
    | ReadWrite | <ul><li>Indexes: Permissions to query, import, fetch, add, update, and delete index data.</li><li>Pinecone Assistant: Permissions to add, list, view, and delete files; chat with an assistant, and evaluate responses.</li><li>Pinecone Inference: Permissions to generate embeddings and rerank documents.</li></ul> |
    | ReadOnly  | <ul><li>Indexes: Permissions to query, fetch, list ID, and view stats.</li><li>Pinecone Assistant: Permissions to list and view files, chat with an assistant, and evaluate responses.</li><li>Pinecone Inference: Permissions to generate embeddings and rerank documents.</li></ul>                                  |
    | None      | No data plane permissions.                                                                                                                                                                                                                                                                                             |
  </Tab>

  <Tab title="API">
    | Role              | Permissions                                                                                                                                                                                                                                                                                                            |
    | :---------------- | :--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
    | `DataPlaneEditor` | <ul><li>Indexes: Permissions to query, import, fetch, add, update, and delete index data.</li><li>Pinecone Assistant: Permissions to add, list, view, and delete files; chat with an assistant, and evaluate responses.</li><li>Pinecone Inference: Permissions to generate embeddings and rerank documents.</li></ul> |
    | `DataPlaneViewer` | <ul><li>Indexes: Permissions to query, fetch, list ID, and view stats.</li><li>Pinecone Assistant: Permissions to list and view files, chat with an assistant, and evaluate responses.</li><li>Pinecone Inference: Permissions to generate embeddings and rerank documents.</li></ul>                                  |
    | None              | No data plane permissions.                                                                                                                                                                                                                                                                                             |
  </Tab>
</Tabs>

### Organization single sign-on (SSO)

SSO allows organizations to manage their teams' access to Pinecone through their identity management solution. Once your integration is configured, you can require that users from your domain sign in through SSO, and you can specify a default role for teammates when they sign up. SSO is available on Standard and Enterprise plans.

For more information, see [configure single sign on](/guides/assistant/admin/configure-sso-with-okta).

### Role-based access controls (RBAC)

Pinecone uses role-based access controls (RBAC) to manage access to resources.

Service accounts, API keys, and users are all *principals*. A principal's access is determined by the *roles* assigned to it. Roles are assigned to a principal for a *resource*, either a project or an organization. The roles available to be assigned depend on the type of principal and resource.

#### Service account roles

A service account can be assigned roles for the organization it belongs to, and any projects within that organization. A user can be assigned roles for each organization they belong to, and any projects within that organization. For more information, see [Organization roles](/guides/assistant/admin/organizations-overview#organization-roles) and [Project roles](/guides/assistant/admin/projects-overview#project-roles).

#### API key roles

An API key can only be assigned permissions for the projects it belongs to. For more information, see [API keys](#api-keys).

#### User roles

A user can be assigned roles for each organization they belong to, and any projects within that organization. For more information, see [Organization roles](/guides/assistant/admin/organizations-overview#organization-roles) and [Project roles](/guides/assistant/admin/projects-overview#project-roles).

## Compliance

<Note>
  To learn more about data privacy and compliance at Pinecone, visit the [Pinecone Trust and Security Center](https://security.pinecone.io/).
</Note>

### Audit logs

<publicPreviewEnt />

[Audit logs](/guides/assistant/admin/configure-audit-logs) provide a detailed record of user and API actions that occur within Pinecone.

Events are captured every 30 minutes and each log batch will be saved into its own file as a JSON blob, keyed by the time of the log to be written. Only logs since the integration was created and enabled will be saved.

Audit log events adhere to a standard JSON schema and include the following fields:

```json JSON theme={null}
{
    "id": "00000000-0000-0000-0000-000000000000",
    "organization_id": "AA1bbbbCCdd2EEEe3FF",
    "organization_name": "example-org",
    "client": {
        "userAgent": "rawUserAgent"
    },
    "actor": {
        "principal_id": "00000000-0000-0000-0000-000000000000",
        "principal_name": "example@pinecone.io",
        "principal_type": "user", // user, api_key, service_account
        "display_name": "Example Person" // Only in case of user
    },
	"event": {
        "time": "2024-10-21T20:51:53.697Z",
        "action": "create",
        "resource_type": "index",
        "resource_id": "uuid",
        "resource_name": "docs-example",
        "outcome": {
            "result": "success",
            "reason": "", // Only displays for "result": "failure"
            "error_code": "", // Only displays for "result": "failure"
        },
        "parameters": { // Varies based on event
        }
	}
}
```

The following events are captured in the audit logs:

* [Organization events](#organization-events)
* [Project events](#project-events)
* [Index events](#index-events)
* [User and API key events](#user-and-api-key-events)
* [Security and governance events](#security-and-governance-events)

#### Organization events

| Action            | Query parameters                                                                                               |
| ----------------- | -------------------------------------------------------------------------------------------------------------- |
| Rename org        | `event.action: update`, `event.resource_type: organization`, `event.resource_id: NEW_ORG_NAME`                 |
| Delete org        | `event.action: delete`, `event.resource_type: organization`, `event.resource_id: DELETED_ORG_NAME`             |
| Create org member | `event.action: create`, `event.resource_type: user`, `event.resource_id: [ARRAY_OF_USER_EMAILS]`               |
| Update org member | `event.action: update`, `event.resource_type: user`, `event.resource_id: { user: USER_EMAIL, role: NEW_ROLE }` |
| Delete org member | `event.action: delete`, `event.resource_type: user`, `event.resource_id: USER_EMAIL`                           |

#### Project events

| Action                     | Query parameters                                                                                                   |
| -------------------------- | ------------------------------------------------------------------------------------------------------------------ |
| Create project             | `event.action: create`, `event.resource_type: project`, `event.resouce_id: PROJ_NAME`                              |
| Update project             | `event.action: update`, `event.resource_type: project`, `event.resource_id: PROJECT_NAME`                          |
| Delete project             | `event.action: delete`, `event.resource_type: project`, `event.resource_id: PROJECT_NAME`                          |
| Invite project member      | `event.action: create`, `event.resource_type: user`, `event.resource_id: [ARRAY_OF_USER_EMAILS]`                   |
| Update project member role | `event.action: update`, `event.resource_type: user`, `event.resource_id: { user: USER_EMAIL, role: NEW_ROLE }`     |
| Delete project member      | `event.action: delete`, `event.resource_type: user`, `event.resource_id: { user: USER_EMAIL, project: PROJ_NAME }` |

#### Index events

| Action        | Query parameters                                                                        |
| ------------- | --------------------------------------------------------------------------------------- |
| Create index  | `event.action: create`, `event.resource_type: index`, `event.resouce_id: INDEX_NAME`    |
| Update index  | `event.action: update`, `event.resource_type: index`, `event.resource_id: INDEX_NAME`   |
| Delete index  | `event.action: delete`, `event.resource_type: index`, `event.resource_id: INDEX_NAME`   |
| Create backup | `event.action: create`, `event.resource_type: backup`, `event.resource_id: BACKUP_NAME` |
| Delete backup | `event.action: delete`, `event.resource_type: backup`, `event.resource_id: BACKUP_NAME` |

#### User and API key events

| Action         | Query parameters                                                                        |
| -------------- | --------------------------------------------------------------------------------------- |
| User login     | `event.action: login`, `event.resource_type: user`, `event.resouce_id: USERNAME`        |
| Create API key | `event.action: create`, `event.resource_type: api-key`, `event.resource_id: API_KEY_ID` |
| Delete API key | `event.action: delete`, `event.resource_type: api-key`, `event.resource_id: API_KEY_ID` |

#### Security and governance events

| Action                  | Query parameters                                                                                           |
| ----------------------- | ---------------------------------------------------------------------------------------------------------- |
| Create Private Endpoint | `event.action: create`, `event.resource_type: private-endpoints`, `event.resource_id: PRIVATE_ENDPOINT_ID` |
| Delete Private Endpoint | `event.action: delete`, `event.resource_type: private-endpoints`, `event.resource_id: PRIVATE_ENDPOINT_ID` |

## Data protection

### Encryption at rest

Pinecone encrypts stored data using the 256-bit Advanced Encryption Standard (AES-256) encryption algorithm.

### Encryption in transit

Pinecone uses standard protocols to encrypt user data in transit. Clients open HTTPS or gRPC connections to the Pinecone API; the Pinecone API gateway uses gRPC connections to user deployments in the cloud. These HTTPS and gRPC connections use the TLS 1.2 protocol with 256-bit Advanced Encryption Standard (AES-256) encryption.

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/encryption-in-transit-2.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6da22e2916d66cf6ff03a4cfd7623705" alt="Diagram showing encryption protocols for user data in transit" data-og-width="4134" width="4134" data-og-height="2570" height="2570" data-path="images/encryption-in-transit-2.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/encryption-in-transit-2.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=25e7ed0c6201c3a1f706cedf0ed823a5 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/encryption-in-transit-2.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=ec43f21a1237d675e4c382352b673457 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/encryption-in-transit-2.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=3bda7ada10d9343161aa691602ca5a5b 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/encryption-in-transit-2.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=39f360b1bea6b4865d7508dc30741987 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/encryption-in-transit-2.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=633cf84c092a6a9997b0b8e7daff1311 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/encryption-in-transit-2.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=0035c658e3af1f534e4281e92391716b 2500w" />

Traffic is also encrypted in transit between the Pinecone backend and cloud infrastructure services, such as S3 and GCS. For more information, see [Google Cloud Platform](https://cloud.google.com/docs/security/encryption-in-transit) and [AWS security documentation](https://docs.aws.amazon.com/AmazonS3/userguide/UsingEncryption.html).

## Network security

### Proxies

The following Pinecone SDKs support the use of proxies:

* [Python SDK](/reference/sdks/python/overview#proxy-configuration)
* [Node.js SDK](/reference/sdks/node/overview#proxy-configuration)

---

# Source: https://docs.pinecone.io/guides/search/semantic-search.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Semantic search

> Find semantically similar records using dense vectors.

This page shows you how to search a [dense index](/guides/index-data/indexing-overview#dense-indexes) for records that are most similar in meaning and context to a query. This is often called semantic search, nearest neighbor search, similarity search, or just vector search.

Semantic search uses [dense vectors](https://www.pinecone.io/learn/vector-embeddings/). Each number in a dense vector corresponds to a point in a multidimensional space. Vectors that are closer together in that space are semantically similar.

## Search with text

<Note>
  Searching with text is supported only for [indexes with integrated embedding](/guides/index-data/indexing-overview#integrated-embedding).
</Note>

To search a dense index with a query text, use the [`search_records`](/reference/api/latest/data-plane/search_records) operation with the following parameters:

* The `namespace` to query. To use the default namespace, set the namespace to `"__default__"`.
* The `query.inputs.text` parameter with the query text. Pinecone uses the embedding model integrated with the index to convert the text to a dense vector automatically.
* The `query.top_k` parameter with the number of similar records to return.
* Optionally, you can specify the `fields` to return in the response. If not specified, the response will include all fields.

For example, the following code searches for the 2 records most semantically related to a query text:

<CodeGroup>
  ```python Python theme={null}
  from pinecone import Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  results = index.search(
      namespace="example-namespace", 
      query={
          "inputs": {"text": "Disease prevention"}, 
          "top_k": 2
      },
      fields=["category", "chunk_text"]
  )

  print(results)
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const namespace = pc.index("INDEX_NAME", "INDEX_HOST").namespace("example-namespace");

  const response = await namespace.searchRecords({
    query: {
      topK: 2,
      inputs: { text: 'Disease prevention' },
    },
    fields: ['chunk_text', 'category'],
  });

  console.log(response);
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import org.openapitools.db_data.client.ApiException;
  import org.openapitools.db_data.client.model.SearchRecordsResponse;

  import java.util.*;

  public class SearchText {
      public static void main(String[] args) throws ApiException {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);

          Index index = new Index(config, connection, "integrated-dense-java");

          String query = "Disease prevention";
          List<String> fields = new ArrayList<>();
          fields.add("category");
          fields.add("chunk_text");

          // Search the dense index
          SearchRecordsResponse recordsResponse = index.searchRecordsByText(query,  "example-namespace", fields, 2, null, null);

          // Print the results
          System.out.println(recordsResponse);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
    	bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
      } 

      res, err := idxConnection.SearchRecords(ctx, &pinecone.SearchRecordsRequest{
          Query: pinecone.SearchRecordsQuery{
              TopK: 2,
              Inputs: &map[string]interface{}{
                  "text": "Disease prevention",
              },
          },
          Fields: &[]string{"chunk_text", "category"},
      })
      if err != nil {
          log.Fatalf("Failed to search records: %v", err)
      }
      fmt.Printf(prettifyStruct(res))
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var response = await index.SearchRecordsAsync(
      "example-namespace",
      new SearchRecordsRequest
      {
          Query = new SearchRecordsRequestQuery
          {
              TopK = 4,
              Inputs = new Dictionary<string, object?> { { "text", "Disease prevention" } },
          },
          Fields = ["category", "chunk_text"],
      }
  );

  Console.WriteLine(response);
  ```

  ```shell curl theme={null}
  INDEX_HOST="INDEX_HOST"
  NAMESPACE="YOUR_NAMESPACE"
  PINECONE_API_KEY="YOUR_API_KEY"

  curl "https://$INDEX_HOST/records/namespaces/$NAMESPACE/search" \
    -H "Accept: application/json" \
    -H "Content-Type: application/json" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: unstable" \
    -d '{
          "query": {
              "inputs": {"text": "Disease prevention"},
              "top_k": 2
          },
          "fields": ["category", "chunk_text"]
       }'
  ```
</CodeGroup>

The response will look as follows. Each record is returned with a similarity score that represents its distance to the query vector, calculated according to the [similarity metric](/guides/index-data/create-an-index#similarity-metrics) for the index.

<CodeGroup>
  ```python Python theme={null}
  {'result': {'hits': [{'_id': 'rec3',
                        '_score': 0.8204272389411926,
                        'fields': {'category': 'immune system',
                                   'chunk_text': 'Rich in vitamin C and other '
                                                 'antioxidants, apples '
                                                 'contribute to immune health '
                                                 'and may reduce the risk of '
                                                 'chronic diseases.'}},
                       {'_id': 'rec1',
                        '_score': 0.7931625843048096,
                        'fields': {'category': 'digestive system',
                                   'chunk_text': 'Apples are a great source of '
                                                 'dietary fiber, which supports '
                                                 'digestion and helps maintain a '
                                                 'healthy gut.'}}]},
   'usage': {'embed_total_tokens': 8, 'read_units': 6}}
  ```

  ```javascript JavaScript theme={null}
  {
    result: { 
      hits: [ 
        {
          _id: 'rec3',
          _score: 0.82042724,
          fields: {
            category: 'immune system',
            chunk_text: 'Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases.'
          }
        },
        {
          _id: 'rec1',
          _score: 0.7931626,
          fields: {
            category: 'digestive system',
            chunk_text: 'Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut.'
          }
        }
      ]
    },
    usage: { 
      readUnits: 6, 
      embedTotalTokens: 8 
    }
  }
  ```

  ```java Java theme={null}
  class SearchRecordsResponse {
      result: class SearchRecordsResponseResult {
          hits: [class Hit {
              id: rec3
              score: 0.8204272389411926
              fields: {category=immune system, chunk_text=Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases.}
              additionalProperties: null
          }, class Hit {
              id: rec1
              score: 0.7931625843048096
              fields: {category=endocrine system, chunk_text=Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut.}
              additionalProperties: null
          }]
          additionalProperties: null
      }
      usage: class SearchUsage {
          readUnits: 6
          embedTotalTokens: 13
      }
      additionalProperties: null
  }
  ```

  ```go Go theme={null}
  {
    "result": {
      "hits": [
        {
          "_id": "rec3",
          "_score": 0.82042724,
          "fields": {
            "category": "immune system",
            "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases."
          }
        },
        {
          "_id": "rec1",
          "_score": 0.7931626,
          "fields": {
            "category": "digestive system",
            "chunk_text": "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut."
          }
        }
      ]
    },
    "usage": {
      "read_units": 6,
      "embed_total_tokens": 8
    }
  }
  ```

  ```csharp C# theme={null}
  {
      "result": {
          "hits": [
              {
                  "_id": "rec3",
                  "_score": 0.13741668,
                  "fields": {
                      "category": "immune system",
                      "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases."
                  }
              },
              {
                  "_id": "rec1",
                  "_score": 0.0023413408,
                  "fields": {
                      "category": "digestive system",
                      "chunk_text": "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut."
                  }
              }
          ]
      },
      "usage": {
          "read_units": 6,
          "embed_total_tokens": 5,
          "rerank_units": 1
      }
  }
  ```

  ```json curl theme={null}
   {
      "result": {
          "hits": [
              {
                  "_id": "rec3",
                  "_score": 0.82042724,
                  "fields": {
                      "category": "immune system",
                      "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases."
                  }
              },
              {
                  "_id": "rec1",
                  "_score": 0.7931626,
                  "fields": {
                      "category": "digestive system",
                      "chunk_text": "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut."
                  }
              }
          ]
      },
      "usage": {
          "embed_total_tokens": 8,
          "read_units": 6
      }
  }
  ```
</CodeGroup>

## Search with a dense vector

To search a dense index with a dense vector representation of a query, use the [`query`](/reference/api/latest/data-plane/query) operation with the following parameters:

* The `namespace` to query. To use the default namespace, set the namespace to `"__default__"`.
* The `vector` parameter with the dense vector values representing your query.
* The `top_k` parameter with the number of results to return.
* Optionally, you can set `include_values` and/or `include_metadata` to `true` to include the vector values and/or metadata of the matching records in the response. For better performance, especially with higher `top_k` values, avoid including vector values unless you need them. See [Decrease latency](/guides/optimize/decrease-latency#avoid-including-vector-values-when-not-needed) for more details.

For example, the following code uses a dense vector representation of the query “Disease prevention” to search for the 3 most semantically similar records in the `example-namespaces` namespace:

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  index.query(
      namespace="example-namespace",
      vector=[0.0236663818359375,-0.032989501953125, ..., -0.01041412353515625,0.0086669921875], 
      top_k=3,
      include_metadata=True,
      include_values=False
  )
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  const queryResponse = await index.namespace('example-namespace').query({
      vector: [0.0236663818359375,-0.032989501953125,...,-0.01041412353515625,0.0086669921875],
      topK: 3,
      includeValues: false,
      includeMetadata: true,
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import io.pinecone.unsigned_indices_model.QueryResponseWithUnsignedIndices;

  import java.util.Arrays;
  import java.util.List;

  public class QueryExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");
          List<Float> query = Arrays.asList(0.0236663818359375f, -0.032989501953125f, ..., -0.01041412353515625f, 0.0086669921875f);
          QueryResponseWithUnsignedIndices queryResponse = index.query(3, query, null, null, null, "example-namespace", null, false, true);
          System.out.println(queryResponse);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      queryVector := []float32{0.0236663818359375,-0.032989501953125,...,-0.01041412353515625,0.0086669921875}

      res, err := idxConnection.QueryByVectorValues(ctx, &pinecone.QueryByVectorValuesRequest{
          Vector:          queryVector,
          TopK:            3,
          IncludeValues:   false,
          includeMetadata: true,
      })
      if err != nil {
          log.Fatalf("Error encountered when querying by vector: %v", err)
      } else {
          fmt.Printf(prettifyStruct(res))
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var queryResponse = await index.QueryAsync(new QueryRequest {
      Vector = new[] { 0.0236663818359375f ,-0.032989501953125f, ..., -0.01041412353515625f, 0.0086669921875f },
      Namespace = "example-namespace",
      TopK = 3,
      IncludeMetadata = true,
  });

  Console.WriteLine(queryResponse);
  ```

  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/query" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "vector": [0.0236663818359375,-0.032989501953125,...,-0.01041412353515625,0.0086669921875],
          "namespace": "example-namespace",
          "topK": 3,
          "includeMetadata": true,
          "includeValues": false
      }'
  ```
</CodeGroup>

The response will look as follows. Each record is returned with a similarity score that represents its distance to the query vector, calculated according to the [similarity metric](/guides/index-data/create-an-index#similarity-metrics) for the index.

<CodeGroup>
  ```python Python theme={null}
  {'matches': [{'id': 'rec3',
                'metadata': {'category': 'immune system',
                             'chunk_text': 'Rich in vitamin C and other '
                                            'antioxidants, apples contribute to '
                                            'immune health and may reduce the '
                                            'risk of chronic diseases.'},
                'score': 0.82026422,
                'values': []},
               {'id': 'rec1',
                'metadata': {'category': 'digestive system',
                             'chunk_text': 'Apples are a great source of '
                                            'dietary fiber, which supports '
                                            'digestion and helps maintain a '
                                            'healthy gut.'},
                'score': 0.793068111,
                'values': []},
               {'id': 'rec4',
                'metadata': {'category': 'endocrine system',
                             'chunk_text': 'The high fiber content in apples '
                                            'can also help regulate blood sugar '
                                            'levels, making them a favorable '
                                            'snack for people with diabetes.'},
                'score': 0.780169606,
                'values': []}],
   'namespace': 'example-namespace',
   'usage': {'read_units': 6}}
  ```

  ```JavaScript JavaScript theme={null}
  {
    matches: [
      {
        id: 'rec3',
        score: 0.819709897,
        values: [],
        sparseValues: undefined,
        metadata: [Object]
      },
      {
        id: 'rec1',
        score: 0.792900264,
        values: [],
        sparseValues: undefined,
        metadata: [Object]
      },
      {
        id: 'rec4',
        score: 0.780068815,
        values: [],
        sparseValues: undefined,
        metadata: [Object]
      }
    ],
    namespace: 'example-namespace',
    usage: { readUnits: 6 }
  }
  ```

  ```java Java theme={null}
  class QueryResponseWithUnsignedIndices {
      matches: [ScoredVectorWithUnsignedIndices {
          score: 0.8197099
          id: rec3
          values: []
          metadata: fields {
            key: "category"
            value {
              string_value: "immune system"
            }
          }
          fields {
            key: "chunk_text"
            value {
              string_value: "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases."
            }
          }
          
          sparseValuesWithUnsignedIndices: SparseValuesWithUnsignedIndices {
              indicesWithUnsigned32Int: []
              values: []
          }
      }, ScoredVectorWithUnsignedIndices {
          score: 0.79290026
          id: rec1
          values: []
          metadata: fields {
            key: "category"
            value {
              string_value: "digestive system"
            }
          }
          fields {
            key: "chunk_text"
            value {
              string_value: "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut."
            }
          }
          
          sparseValuesWithUnsignedIndices: SparseValuesWithUnsignedIndices {
              indicesWithUnsigned32Int: []
              values: []
          }
      }, ScoredVectorWithUnsignedIndices {
          score: 0.7800688
          id: rec4
          values: []
          metadata: fields {
            key: "category"
            value {
              string_value: "endocrine system"
            }
          }
          fields {
            key: "chunk_text"
            value {
              string_value: "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes."
            }
          }
          
          sparseValuesWithUnsignedIndices: SparseValuesWithUnsignedIndices {
              indicesWithUnsigned32Int: []
              values: []
          }
      }]
      namespace: example-namespace
      usage: read_units: 6

  }
  ```

  ```go Go theme={null}
  {
    "matches": [
      {
        "vector": {
          "id": "rec3",
          "metadata": {
            "category": "immune system",
            "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases."
          }
        },
        "score": 0.8197099
      },
      {
        "vector": {
          "id": "rec1",
          "metadata": {
            "category": "digestive system",
            "chunk_text": "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut."
          }
        },
        "score": 0.79290026
      },
      {
        "vector": {
          "id": "rec4",
          "metadata": {
            "category": "endocrine system",
            "chunk_text": "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes."
          }
        },
        "score": 0.7800688
      }
    ],
    "usage": {
      "read_units": 6
    },
    "namespace": "example-namespace"
  }
  ```

  ```csharp C# theme={null}
  {
    "results": [],
    "matches": [
      {
        "id": "rec3",
        "score": 0.8197099,
        "values": [],
        "metadata": {
          "category": "immune system",
          "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases."
        }
      },
      {
        "id": "rec1",
        "score": 0.79290026,
        "values": [],
        "metadata": {
          "category": "digestive system",
          "chunk_text": "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut."
        }
      },
      {
        "id": "rec4",
        "score": 0.7800688,
        "values": [],
        "metadata": {
          "category": "endocrine system",
          "chunk_text": "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes."
        }
      }
    ],
    "namespace": "example-namespace",
    "usage": {
      "readUnits": 6
    }
  }
  ```

  ```json curl theme={null}
  {
      "results": [],
      "matches": [
          {
              "id": "rec3",
              "score": 0.820593238,
              "values": [],
              "metadata": {
                  "category": "immune system",
                  "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases."
              }
          },
          {
              "id": "rec1",
              "score": 0.792266726,
              "values": [],
              "metadata": {
                  "category": "digestive system",
                  "chunk_text": "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut."
              }
          },
          {
              "id": "rec4",
              "score": 0.780045748,
              "values": [],
              "metadata": {
                  "category": "endocrine system",
                  "chunk_text": "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes."
              }
          }
      ],
      "namespace": "example-namespace",
      "usage": {
          "readUnits": 6
      }
  }
  ```
</CodeGroup>

## Search with a record ID

When you search with a record ID, Pinecone uses the dense vector associated with the record as the query. To search a dense index with a record ID, use the [`query`](/reference/api/latest/data-plane/query) operation with the following parameters:

* The `namespace` to query. To use the default namespace, set the namespace to `"__default__"`.
* The `id` parameter with the unique record ID containing the vector to use as the query.
* The `top_k` parameter with the number of results to return.
* Optionally, you can set `include_values` and/or `include_metadata` to `true` to include the vector values and/or metadata of the matching records in the response. For better performance, especially with higher `top_k` values, avoid including vector values unless you need them. See [Decrease latency](/guides/optimize/decrease-latency#avoid-including-vector-values-when-not-needed) for more details.

For example, the following code uses an ID to search for the 3 records in the `example-namespace` namespace that are most semantically similar to the dense vector in the record:

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  index.query(
      namespace="example-namespace",
      id="rec2", 
      top_k=3,
      include_metadata=True,
      include_values=False
  )
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  const queryResponse = await index.namespace('example-namespace').query({
      id: 'rec2',
      topK: 3,
      includeValues: false,
      includeMetadata: true,
  });
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import io.pinecone.unsigned_indices_model.QueryResponseWithUnsignedIndices;

  public class QueryExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");
          QueryResponseWithUnsignedIndices queryRespone = index.queryByVectorId(3, "rec2", "example-namespace", null, false, true);
          System.out.println(queryResponse);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
  	bytes, _ := json.MarshalIndent(obj, "", "  ")
  	return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      vectorId := "rec2"
      res, err := idxConnection.QueryByVectorId(ctx, &pinecone.QueryByVectorIdRequest{
          VectorId:      vectorId,
          TopK:          3,
          IncludeValues: false,
          IncludeMetadata: true,
      })
      if err != nil {
          log.Fatalf("Error encountered when querying by vector ID `%v`: %v", vectorId, err)
      } else {
          fmt.Printf(prettifyStruct(res.Matches))
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var queryResponse = await index.QueryAsync(new QueryRequest {
      Id = "rec2",
      Namespace = "example-namespace",
      TopK = 3,
      IncludeValues = false,
      IncludeMetadata = true
  });

  Console.WriteLine(queryResponse);
  ```

  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/query" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "id": "rec2",
          "namespace": "example-namespace",
          "topK": 3,
          "includeMetadata": true,
          "includeValues": false
      }'
  ```
</CodeGroup>

## Parallel queries

Python SDK v6.0.0 and later provide `async` methods for use with [asyncio](https://docs.python.org/3/library/asyncio.html). Async support makes it possible to use Pinecone with modern async web frameworks such as FastAPI, Quart, and Sanic, and can significantly increase the efficiency of running queries in parallel. For more details, see the [Async requests](/reference/sdks/python/overview#async-requests).

---

# Source: https://docs.pinecone.io/troubleshooting/serverless-index-connection-errors.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Serverless index connection errors

## Problem

To connect to a serverless index, you must use an updated Pinecone client. Trying to connect to a serverless index with an outdated client will raise errors similar to one of the following:

```console console theme={null}
Failed to resolve 'controller.us-west-2.pinecone.io'

controller.us-west-2-aws.pinecone.io not found

Request failed to reach Pinecone while calling https://controller.us-west-2.pinecone.io/actions/whoami
```

## Solution

Upgrade to the latest [Python](https://github.com/pinecone-io/pinecone-python-client) or [Node.js](https://sdk.pinecone.io/typescript/) client and try again:

<CodeGroup>
  ```python Python theme={null}
  pip install "pinecone[grpc]" --upgrade
  ```

  ```js JavaScript theme={null}
  npm install @pinecone-database/pinecone@latest
  ```
</CodeGroup>

---

# Source: https://docs.pinecone.io/integrations/snowflake.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Snowflake

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Deploy and run Pinecone with Snowpark Container Services. Snowpark Container Services is a fully managed container offering designed to facilitate the deployment, management, and scaling of containerized applications within the Snowflake ecosystem. This service enables users to run containerized workloads directly within Snowflake, ensuring that data does't need to be moved out of the Snowflake environment for processing. Unlike traditional container orchestration platforms like Docker or Kubernetes, Snowpark Container Services offers an OCI runtime execution environment specifically optimized for Snowflake. This integration allows for the seamless execution of OCI images, leveraging Snowflak's robust data platform.

<PrimarySecondaryCTA primaryLabel={"Get started"} primaryHref={"https://www.snowflake.com/blog/snowpark-container-services-deploy-genai-full-stack-apps/"} />

## Related articles

* [Snowpark Container Services: Securely Deploy and run Sophisticated Generative AI and full-stack apps in Snowflake](https://www.snowflake.com/blog/snowpark-container-services-deploy-genai-full-stack-apps/)

---

# Source: https://docs.pinecone.io/guides/organizations/manage-billing/standard-trial.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Standard trial

> Get $300 credits for 21 days with the Standard plan trial.

The Standard trial lets you evaluate Pinecone without requiring any up-front payment. You get \$300 in credits over 21 days with access to Standard plan [features](https://www.pinecone.io/pricing/) and [limits](/reference/api/database-limits) that are suitable for testing Pinecone at scale.

<Note>
  If you're building a small or personal project, consider the free [Starter plan](https://www.pinecone.io/pricing/) instead.
</Note>

## Key features

* \$300 in credits
* 21 days of access to Standard plan [features](https://www.pinecone.io/pricing/), including:
  * [Bulk import](/guides/index-data/import-data)
  * [Backup and restore](/guides/manage-data/backups-overview)
  * [RBAC (role-based access control)](/guides/production/security-overview#role-based-access-control)
* [Higher limits](/reference/api/database-limits) for testing at scale
* Access to all [cloud regions](/guides/index-data/create-an-index#cloud-regions)
* Access to [Developer Support](https://www.pinecone.io/pricing/?plans=support)

## Expiration

At the end of a Standard trial, or when you've used all of your credits, you can take one of the following actions:

* Add a payment method and continue on with the Standard plan.
* Upgrade to the Enterprise plan.
* [Downgrade to a Starter plan](#downgrading-to-the-starter-plan) (you can also do this before your trial expires, if you choose).

<Note>
  Learn more about [pricing](https://www.pinecone.io/pricing/).
</Note>

## Downgrading to the Starter plan

To downgrade from a Standard trial to the Starter plan, you'll need to bring your usage within Starter plan limits.

* No more than 5 indexes, all serverless and in the `us-east-1` region of AWS
  * If you have serverless indexes in a region other than `us-east-1`, [create a new serverless index](/guides/index-data/create-an-index#create-a-serverless-index) in `us-east-1`, [re-upsert your data](/guides/index-data/upsert-data) into the new index, and [delete the old index](/guides/manage-data/manage-indexes#delete-an-index).
  * If you have more than 5 serverless indexes, [delete indexes](/guides/manage-data/manage-indexes#delete-an-index) until you have 5 or fewer.
  * If you have pod-based indexes, [delete them](/guides/manage-data/manage-indexes#delete-an-index).
* No more than 1 project
  * If you have more than 1 project, [delete all but 1 project](/guides/projects/manage-projects#delete-a-project).
  * Before you can delete a project, you must [delete all indexes](/guides/manage-data/manage-indexes#delete-an-index) and [delete all collections](/guides/manage-data/back-up-an-index#delete-a-collection) in the project.
* No more than 2 GB of data across all of your serverless indexes
  * If you are storing more than 2 GB of data, [delete records](/guides/manage-data/delete-data) until you're storing less than 2 GB.
* No more than 100 namespaces per serverless index
  * If any serverless index has more than 100 namespaces, [delete namespaces](/guides/manage-data/delete-data#delete-all-records-from-a-namespace) until it has 100 or fewer remaining.
* No more than 3 [assistants](/guides/assistant/overview)
  * If you have more than 3 assistants, [delete assistants](/guides/assistant/manage-assistants#delete-an-assistant) until you have 3 or fewer.
* No more than 10 files per assistant
  * If you have more than 10 files uploaded to an assistant, [delete files](/guides/assistant/manage-files#delete-a-file) until the assistant has 10 or fewer.
* No more than 1 GB of assistant storage
  * If you have more than 1 GB of assistant storage, [delete files](https://docs.pinecone.io/guides/assistant/manage-files#delete-a-file) until you're storing less than 1 GB.
* No more than 2 users
* No collections or backups (these are automatically deleted as part of the downgrade process)

If you have questions, [contact Support](https://www.pinecone.io/contact/support/).

## Limits

* Each organization is allowed only one trial.
* Organizations already on a Standard or Enterprise plan cannot activate a Standard plan trial.
* Organizations that initially subscribed to Pinecone through marketplace partners cannot activate a Standard plan trial.

If you have any questions, [contact Support](https://www.pinecone.io/contact/support/).

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/start_import.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Start import

> Start an asynchronous import of vectors from object storage into an index.

For guidance and examples, see [Import data](https://docs.pinecone.io/guides/index-data/import-data).

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Standard and Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

<RequestExample>
  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/bulk/imports" \
    -H 'Api-Key: $PINECONE_API_KEY' \
    -H 'Content-Type: application/json' \
    -H 'X-Pinecone-Api-Version: 2025-10' \
    -d '{
          "integrationId": "a12b3d4c-47d2-492c-a97a-dd98c8dbefde",
          "uri": "s3://BUCKET_NAME/PATH/TO/DIR",
          "errorMode": {
              "onError": "continue"
              }
          }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
     "id": "101"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml post /bulk/imports
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /bulk/imports:
    post:
      tags:
        - Bulk Operations
      summary: Start import
      description: >-
        Start an asynchronous import of vectors from object storage into an
        index.


        For guidance and examples, see [Import
        data](https://docs.pinecone.io/guides/index-data/import-data).
      operationId: startBulkImport
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/StartImportRequest'
        required: true
      responses:
        '200':
          description: Successful import operation
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/StartImportResponse'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    StartImportRequest:
      description: The request for the `start_import` operation.
      type: object
      properties:
        integrationId:
          description: >-
            The id of the [storage
            integration](https://docs.pinecone.io/guides/operations/integrations/manage-storage-integrations)
            that should be used to access the data.
          type: string
          maxLength: 1000
        uri:
          description: >-
            The URI of the bucket (or container) and import directory containing
            the namespaces and Parquet files you want to import. For example,
            `s3://BUCKET_NAME/IMPORT_DIR` for Amazon S3,
            `gs://BUCKET_NAME/IMPORT_DIR` for Google Cloud Storage, or
            `https://STORAGE_ACCOUNT.blob.core.windows.net/CONTAINER_NAME/IMPORT_DIR`
            for Azure Blob Storage. For more information, see [Import
            records](https://docs.pinecone.io/guides/index-data/import-data#prepare-your-data).
          type: string
          minLength: 1
          maxLength: 1500
        errorMode:
          $ref: '#/components/schemas/ImportErrorMode'
      required:
        - uri
    StartImportResponse:
      description: The response for the `start_import` operation.
      type: object
      properties:
        id:
          example: '101'
          description: Unique identifier for the import operation.
          type: string
          minLength: 1
          maxLength: 1000
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    ImportErrorMode:
      description: Indicates how to respond to errors during the import process.
      type: object
      properties:
        onError:
          description: |-
            Indicates how to respond to errors during the import process.
            Possible values: `abort` or `continue`.
          x-enum:
            - abort
            - continue
          type: string
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/integrations/streamnative.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# StreamNative

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Founded by the original developers of Apache Pulsar and Apache BookKeeper, [StreamNative](https://streamnative.io) provides StreamNative Cloud, offering Apache Pulsar as a Service. The company also supports on-premise Pulsar deployments and related commercial support. StreamNative Cloud provides a scalable, resilient, and secure messaging and event streaming platform for enterprises. Additionally, StreamNative offers Kafka compatibility, enabling seamless integration with existing Kafka-based systems.

The Pinecone integration with StreamNative allows access to pinecone.io with a Pulsar topic. The sink connector takes in messages and writes them if they are in a proper format to a Pinecone index.

<PrimarySecondaryCTA primaryHref={"https://docs.streamnative.io/hub/connector-pinecone-sink-v3.3#pinecone-sink-connector"} primaryLabel={"Get started"} />

---

# Source: https://docs.pinecone.io/guides/manage-data/target-an-index.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Target an index

> Target an index by host or name for data operations such as upsert and query.

<Warning>
  **Do not target an index by name in production.**

  When you target an index by name for data operations such as `upsert` and `query`, the SDK gets the unique DNS host for the index using the `describe_index` operation. This is convenient for testing but should be avoided in production because `describe_index` uses a different API than data operations and therefore adds an additional network call and point of failure. Instead, you should get an index host once and cache it for reuse or specify the host directly.
</Warning>

## Target by index host (recommended)

This method is recommended for production:

<Note>
  When using Private Endpoints for private connectivity between your application and Pinecone, you must target the index using the [Private Endpoint URL](/guides/production/connect-to-aws-privatelink#run-data-plane-commands) for the host.
</Note>

<CodeGroup>
  ```Python Python {5} theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  index = pc.Index(host="INDEX_HOST")
  ```

  ```javascript JavaScript {6} theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  // For the Node.js SDK, you must specify both the index host and name.
  const index = pc.index("INDEX_NAME", "INDEX_HOST");
  ```

  ```java Java {11} theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;

  public class TargetIndexByHostExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          // For the Java SDK, you must specify both the index host and name.
          Index index = new Index(connection, "INDEX_NAME");
      }
  }
  ```

  ```go Go {21} theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // This creates a new gRPC index connection, targeting the namespace "example-namespace"
      idxConnectionNs1, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host %v: %v", idx.Host, err)
      }

      // This reuses the gRPC index connection, targeting a different namespace
  	idxConnectionNs2 := idxConnectionNs1.WithNamespace("example-namespace2")
  }
  ```

  ```csharp C# {5} theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var index = pinecone.Index(host: "INDEX_HOST");
  ```
</CodeGroup>

### Get an index host

You can get the unique DNS host for an index from the Pinecone console or the Pinecone API.

To get an index host from the Pinecone console:

1. Open the [Pinecone console](https://app.pinecone.io/organizations/-/projects).
2. Select the project containing the index.
3. Select the index.
4. Copy the URL under **HOST**.

To get an index host from the Pinecone API, use the [`describe_index`](/reference/api/latest/control-plane/describe_index) operation, which returns the index host as the `host` value:

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  pc.describe_index(name="docs-example")

  # Response:
  # {'deletion_protection': 'disabled',
  #  'dimension': 1536,
  #  'host': 'docs-example-4zo0ijk.svc.us-east1-aws.pinecone.io',
  #  'metric': 'cosine',
  #  'name': 'docs-example',
  #  'spec': {'serverless': {'cloud': 'aws', 'region': 'us-east-1'}},
  #  'status': {'ready': True, 'state': 'Ready'}}
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  await pc.describeIndex('docs-example');

  // Response:
  // {
  //    "name": "docs-example",
  //    "dimension": 1536,
  //    "metric": "cosine",
  //    "host": "docs-example-4zo0ijk.svc.us-east1-aws.pinecone.io",
  //    "deletionProtection": "disabled",
  //    "spec": {
  //       "serverless": {
  //          "cloud": "aws",
  //          "region": "us-east-1"
  //       }
  //    },
  //    "status": {
  //       "ready": true,
  //       "state": "Ready"
  //    }
  // }
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Pinecone;
  import org.openapitools.db_control.client.model.*;

  public class DescribeIndexExample {
      public static void main(String[] args) {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          IndexModel indexModel = pc.describeIndex("docs-example");
          System.out.println(indexModel);
      }
  }

  // Response:
  // class IndexModel {
  //     name: docs-example-java
  //     dimension: 1536
  //     metric: cosine
  //     host: docs-example-4zo0ijk.svc.us-west2-aws.pinecone.io
  //     deletionProtection: enabled
  //     spec: class IndexModelSpec {
  //         pod: null
  //         serverless: class ServerlessSpec {
  //             cloud: aws
  //             region: us-east-1
  //         }
  //     }
  //     status: class IndexModelStatus {
  //         ready: true
  //         state: Ready
  //     }
  // }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "encoding/json"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func prettifyStruct(obj interface{}) string {
      bytes, _ := json.MarshalIndent(obj, "", "  ")
      return string(bytes)
  }

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      idx, err := pc.DescribeIndex(ctx, "docs-example")
      if err != nil {
          log.Fatalf("Failed to describe index \"%v\": %v", idx.Name, err)
      } else {
          fmt.Printf("index: %v\n", prettifyStruct(idx))
      }
  }

  // Response:
  // index: {
  // 	"name": "docs-example",
  // 	"dimension": 1536,
  // 	"host": "docs-example-govk0nt.svc.apw5-4e34-81fa.pinecone.io",
  // 	"metric": "cosine",
  //  "deletion_protection": "disabled",
  // 	"spec": {
  // 	  "serverless": {
  // 		"cloud": "aws",
  // 		"region": "us-east-1"
  // 	  }
  // 	},
  // 	"status": {
  // 	  "ready": true,
  // 	  "state": "Ready"
  // 	}
  // }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var indexModel = await pinecone.DescribeIndexAsync("docs-example");

  Console.WriteLine(indexModel);

  // Response:
  // {
  //   "name": "docs-example",
  //   "dimension": 1536,
  //   "metric": "cosine",
  //   "host": "docs-example-govk0nt.svc.aped-4627-b74a.pinecone.io",
  //   "deletion_protection": "disabled",
  //   "spec": {
  //     "pod": null,
  //     "serverless": {
  //       "cloud": "aws",
  //       "region": "us-east-1"
  //     }
  //   },
  //   "status": {
  //     "ready": true,
  //     "state": "Ready"
  //   }
  // }
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"

  curl -i -X GET "https://api.pinecone.io/indexes/docs-example-curl" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -H "X-Pinecone-Api-Version: 2025-10"

  # Response:
  # {
  #    "name": "docs-example",
  #    "metric": "cosine",
  #    "dimension": 1536,
  #    "status": {
  #       "ready": true,
  #       "state": "Ready"
  #    },
  #    "host": "docs-example-4zo0ijk.svc.us-east1-aws.pinecone.io",
  #    "spec": {
  #       "serverless": {
  #          "region": "us-east-1",
  #          "cloud": "aws"
  #       }
  #    }
  # }
  ```
</CodeGroup>

## Target by index name

This method is convenient for testing but is not recommended for production:

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  index = pc.Index("docs-example")
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone';

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  // For the Node.js SDK, you must specify both the index host and name.
  const index = pc.index('docs-example');
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.clients.Pinecone;

  public class GenerateEmbeddings {
      public static void main(String[] args) throws ApiException {
          Pinecone pc = new Pinecone.Builder("YOUR_API_KEY").build();
          Index index = pc.getIndexConnection("docs-example");
      }
  }
  ```

  ```go Go theme={null}
  // It is not possible to target an index by name in the Go SDK. 
  // You must target an index by host.
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  var index = pinecone.Index("docs-example");
  ```
</CodeGroup>

---

# Source: https://docs.pinecone.io/reference/cli/target-context.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# CLI target context

<Note>
  This feature is in [public preview](/release-notes/feature-availability).
</Note>

Most CLI operations happen within the context of a specific organization and project. The **target context** is the organization and project scope that the CLI uses when running commands.

Before setting target context, you must authenticate with the CLI. See [CLI authentication](/reference/cli/authentication) for details on authentication methods.

## Why target context matters

By specifying a target organization and/or project, you're telling the CLI which resources to operate on.

* **Control plane operations** (creating indexes, listing indexes): Always happen within a specific project.
* **Admin API operations**:
  * Organization operations don't require target context. They operate on organizations you have access to.
  * Project management operations are scoped to your target organization. For example, running `pc project list` shows projects in the target org.
  * API key operations are scoped to the target project, unless you specify `--id`.

## How target context works

How the CLI sets and uses target context depends on your authentication method and how you configure it.

### User login

When you authenticate with `pc auth login`, the CLI automatically [targets](/reference/cli/target-context):

* The default organization returned by the server for your user.
* The first project in the list of that organization's projects.

However, you can change organization and project as needed:

```bash  theme={null}
pc auth login
pc target -o "my-org" -p "my-project"
```

### Service account

**Configured via CLI command**

When you configure a service account with `pc auth configure --client-id --client-secret`, the CLI automatically targets the organization associated with the service account (service accounts belong to a single organization). Then, to target a project:

* If the selected organization has only one project, the CLI automatically targets it.
* If that organization has multiple projects, the CLI prompts you to select one (or you can use the `--project-id` flag).
* If it doesn't have any projects, create one and then target it manually.

To change to a specific target project:

```bash  theme={null}
pc target -p "different-project"
```

Or, to select a project interactively:

```bash  theme={null}
pc target
```

**Configured via environment variables**

If you set service account credentials via environment variables (`PINECONE_CLIENT_ID` and `PINECONE_CLIENT_SECRET`) without running `pc auth configure`, the CLI does **not** automatically set any target context. You must explicitly set it with `pc target`.

Service accounts are scoped to a single organization. You can only target the organization associated with your service account credentials. To do this, and also target a specific project within that organization:

```bash  theme={null}
pc target -o "my-org" -p "my-project"
```

Or, to select a project interactively (the CLI discovers the organization automatically):

```bash  theme={null}
pc target
```

### API key

When you set a default API key (with `pc auth configure --api-key` or the `PINECONE_API_KEY` environment variable), the CLI does **not** change or clear its stored target context. However, in this scenario, control plane operations **do** use the API key's organization and project — not the CLI's saved target context (regardless of any calls you've made to `pc target`, or the output of `pc target --show`). This happens because Pinecone API keys are always scoped to a specific organization and project, and they cannot access resources outside of that scope.

<Note>
  Because Pinecone API keys cannot be used to authenticate calls to the Admin API, Admin API operations still authentiate with your user login token or service account credentials (if available).
</Note>

```bash  theme={null}
pc auth login
pc target -o "my-org" -p "my-project"
# Setting an API key doesn't change stored target context
pc auth configure --api-key "YOUR_API_KEY"
pc target --show  # Still shows my-org and my-project, but they're not used for control plane operations
```

## Viewing and managing target context

To view your current target organization and project:

```bash  theme={null}
# Show current target context
pc target --show

# Clear target context
pc target --clear
```

---

# Source: https://docs.pinecone.io/integrations/terraform.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Terraform

Terraform is an infrastructure as code tool that lets you create, update, and version infrastructure by defining resources in configuration files. This allows for a repeated workflow for provisioning and managing your infrastructure.

This page describes how to use the [Terraform Provider for Pinecone](https://registry.terraform.io/providers/pinecone-io/pinecone/latest/docs) to manage Pinecone indexes, collections, API keys, and projects.

## Requirements

Ensure you have the following:

* [Terraform](https://developer.hashicorp.com/terraform) >= v1.4.6
* [Go](https://go.dev/doc/install) >= v1.23.7
* A [Pinecone API key](https://app.pinecone.io/organizations/-/keys) for managing indexes and collections
* A [Pinecone service account](https://app.pinecone.io/organizations/-/settings/access/service-accounts) for managing API keys and projects

## Install the provider

1. Configuring the Pinecone provider in your Terraform configuration file:

   ```hcl  theme={null}
   terraform { 
     required_providers { 
       pinecone = { 
         source = "pinecone-io/pinecone" 
         version = "~> 2.0.0"
       } 
     } 
   } 
   ```

2. Run `terraform init` to install the provider from the [Terraform registry](https://registry.terraform.io/providers/pinecone-io/pinecone/latest). Alternatively, you can download the latest binary for your target platfrom the [GitHub repository](https://github.com/pinecone-io/terraform-provider-pinecone/releases).

## Authenticate

For managing indexes and collections, you authenticate with a [Pinecone API key](https://app.pinecone.io/organizations/-/keys). For managing API keys and projects, you authenticate with [Pinecone service account](https://app.pinecone.io/organizations/-/settings/access/service-accounts) credentials (client ID and client secret).

1. Set environment variables for authentication:

   ```bash  theme={null}
   # For indexes and collections  
   export PINECONE_API_KEY="YOUR_API_KEY"

   # For API keys and projects
   export PINECONE_CLIENT_ID="YOUR_CLIENT_ID"
   export PINECONE_CLIENT_SECRET="YOUR_CLIENT_SECRET"
   ```

2. Append the following to your Terraform configuration file:

   ```hcl  theme={null}
   provider "pinecone" {}
   ```

<Note>
  You can also set the API key and/or service account credential as [input variables](https://developer.hashicorp.com/terraform/language/values/variables).
</Note>

## Manage resources

The Terraform Provider for Pinecone allows Terraform to manage indexes, collections, API keys, and projects.

### Indexes

The `pinecone_index` resource lets you create, update, and delete [indexes](/guides/index-data/indexing-overview).

<Note>
  You can [update](/guides/manage-data/manage-indexes) only the index deletion protection, tags, and integrated inference embedding settings of an index.
</Note>

```terraform  theme={null}
# Dense index
resource "pinecone_index" "example-index" {
  name = "example-index"
  dimension = 1536
  metric = "cosine"
  vector_type = "dense"
  spec = {
    serverless = {
      cloud = "aws"
      region = "us-west-2"
    }
  }
  deletion_protection = "disabled"
  tags = {
    environment = "development"
  }
}

# Dense index with integrated embedding
resource "pinecone_index" "example-index-integrated" {
  name = "example-index-integrated"
  spec = {
    serverless = {
      cloud = "aws"
      region = "us-west-2"
    }
  }
  embed = {
    model = "llama-text-embed-v2"
    field_map = {
      text = "chunk_text"
    }
  }
}
```

### Collections

The `pinecone_collection` resource lets you create and delete [collections](/guides/indexes/pods/understanding-collections) for pod-based indexes.

```terraform  theme={null}
resource "pinecone_index" "example-index" {
  name = "example-index"
  dimension = 10
  spec = {
    pod = {
      environment = "us-west4-gcp"
      pod_type = "s1.x1"
    }
  }
}

resource "pinecone_collection" "example-collection" {
  name = "example-collection"
  source = pinecone_index.example-index.name
```

### API keys

The `pinecone_api_key` resource lets you create, update, and delete [API keys](/guides/projects/manage-api-keys).

<Note>
  You can update only the name and roles of an API key.
</Note>

```terraform  theme={null}
# API key with default roles (ProjectEditor)
resource "pinecone_api_key" "example-key" {
  name = "example-key"
  project_id = "YOUR_PROJECT_ID"
}

# API key with custom roles
resource "pinecone_api_key" "example-key-custom_roles" {
  name = "example-key-custom-roles"
  project_id = "YOUR_PROJECT_ID"
  roles = ["ProjectViewer", "DataPlaneViewer"]
}

output "api_key_roles" {
  description = "The roles assigned to the API key"
  value       = pinecone_api_key.example.roles
}
```

### Projects

The `pinecone_project` resource lets you create, update, and delete [projects](/guides/projects/understanding-projects).

<Warning>
  Customers who signed up for a Standard or Enterprise plan on or after August 18, 2025 cannot create [pod-based indexes](/guides/indexes/pods/understanding-pod-based-indexes) and cannot set the max pods for a project.
</Warning>

```terraform  theme={null}
# Basic project
resource "pinecone_project" "example-project" {
  name = "example-project"
}

# Project with CMEK encryption enabled
resource "pinecone_project" "example-project-encrypted" {
  name = "example-project-encrypted"
  force_encryption_with_cmek = true
}

# Project with custom max pods
resource "pinecone_project" "example-project-custom-pods" {
  name = "example-project-custom-pods"
  max_pods = 10
}
```

## Limitations

The Terraform Provider for Pinecone does not support the following resources:

* [Backups for serverless indexes](/guides/manage-data/backups-overview)
* [Service accounts](/guides/projects/manage-service-accounts)
* [Private endpoints](/guides/production/connect-to-aws-privatelink)
* [Assistants](/guides/assistant/overview)

## See also

* Documentation can be found on the [Terraform
  Registry](https://registry.terraform.io/providers/pinecone-io/pinecone/latest/docs).
* See the [GitHub respository](https://github.com/pinecone-io/terraform-provider-pinecone/tree/main/examples)
  for additional usage examples.
* For support requests, create an issue in the [GitHub
  repository](https://github.com/pinecone-io/terraform-provider-pinecone).

---

# Source: https://docs.pinecone.io/guides/get-started/test-at-scale.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Test Pinecone at scale

> Test Pinecone with a real-world dataset and semantic search workload.

This guide walks you through testing Pinecone at production scale. You'll import 10 million vectors, run a benchmark, and analyze the results to verify Pinecone meets production requirements for semantic search applications.

<Note>
  This test requires a Pinecone account on the Standard or Enterprise plan. New users can sign up for the [Standard trial](/guides/organizations/manage-billing/standard-trial) for 21 days and \$300 in credits, more than enough to cover the costs of this test. Existing users on the Starter plan can [upgrade](/guides/organizations/manage-billing/upgrade-billing-plan).
</Note>

## About this test

Semantic search enables finding relevant content based on meaning rather than exact keyword matches, making it ideal for applications like product search, content recommendation, and question-answering systems. This test simulates a production-scale semantic search workload, measuring import time, query throughput, query latency, and associated costs.

The test uses the following configuration:

* **Records**: 10 million records from the [Amazon Reviews 2023](https://amazon-reviews-2023.github.io/) dataset
* **Embedding model**: `llama-text-embed-v2` (1024 dimensions)
* **Similarity metric**: cosine
* **Total size**: 48.8 GB
* **Query load**: 10 queries per second total (across all users)
* **Concurrent users**: 10 users querying simultaneously
* **Test queries**: 100,000 queries
* **Import time target**: \< 30 minutes
* **Query latency target**: p90 latency \< 100ms

**Estimated cost**: \~\$127 (import: \$48.80, queries: \$78.08, storage: \$0.09) — see [detailed cost breakdown](#6-check-costs)

## 1. Get an API key

To follow the steps in this guide, you'll need an API key. Create a new API key in the [Pinecone console](https://app.pinecone.io/organizations/-/keys), or use this widget:

<div style={{minWidth: '450px', minHeight:'152px'}}>
  <div id="pinecone-connect-widget">
    <div class="connect-widget-skeleton">
      <div class="skeleton-content" />
    </div>
  </div>
</div>

Your generated API key:

```shell  theme={null}
"{{YOUR_API_KEY}}"
```

## 2. Create an index

<Note>
  This test requires you to use AWS-based indexes and infrastructure. The sample dataset is only available from Amazon S3, and you can only import from Amazon S3 to Pinecone indexes hosted on AWS. To run the benchmark, you'll need to provision an AWS EC2 instance in the same region as your index.
</Note>

Create an on-demand index that matches the dimensions and similarity metric of the dataset you'll import in later steps.

<Tabs>
  <Tab title="Console">
    1. In the Pinecone console, go to the [Indexes](https://app.pinecone.io/organizations/-/projects/-/indexes) page.
    2. Click **Create index**.
    3. Check **Custom settings**.
    4. Configure the index with the following settings:
       * **Name**: `search-10m`
       * **Vector type**: Dense
       * **Dimensions**: `1024`
       * **Metric**: cosine
       * **Capacity mode**: Serverless (on-demand)
       * **Cloud**: AWS (required for this test)
       * **Region**: Use an AWS region appropriate for your use case (for example, `us-east-1`)
    5. Click **Create index**.
  </Tab>

  <Tab title="Code">
    If using code to create an index, first install the [Python SDK](/reference/python-sdk):

    ```shell Terminal theme={null}
    pip install pinecone
    ```

    Then, create the index:

    <CodeGroup>
      ```python Python theme={null}
      from pinecone import Pinecone, ServerlessSpec

      pc = Pinecone(api_key="{{YOUR_API_KEY}}")

      index_name = "search-10m"

      if not pc.has_index(index_name):
          pc.create_index(
              name=index_name,
              vector_type="dense",
              dimension=1024,
              metric="cosine",
              spec=ServerlessSpec(
                  # AWS is required for this test
                  cloud="aws",   
                  # Use an AWS region appropriate for your use case
                  region="us-east-1" 
              )
          )
      ```
    </CodeGroup>
  </Tab>
</Tabs>

## 3. Import the dataset

Pinecone's [import feature](/guides/index-data/import-data) enables you to load millions of vectors from object storage in parallel. In this step, you'll import 10 million records into a single namespace (`ns_2`) in your index.

### Choose an import source

To import the dataset, you'll need to use the following Amazon S3 import URL:

```
s3://fe-customer-pocs/search/search_10M/dense/
```

### Start and monitor the import

For this dataset, the import should take less than 30 minutes.

<Tabs>
  <Tab title="Console">
    1. In the Pinecone console, go to the [Indexes](https://app.pinecone.io/organizations/-/projects/-/indexes) page.
    2. Find your `search-10m` index and click **... > Import data**.
    3. For **Storage integration**, select **No integration (public bucket)**.
    4. Enter the import URL: `s3://fe-customer-pocs/search/search_10M/dense/`.
    5. For **Error handling**, select **Abort on error (default)**.
    6. Click **Start import**.

    To monitor progress, open your index in the Pinecone console and navigate to the **Imports** tab. After the import completes, compare the **Started time** and **End time** timestamps to see the total time required.

    <Note>
      For this dataset, the import should take around 30 minutes. While the import is running, you can continue with the next step to provision a VM and install VSB. However, wait for the import to complete before running the benchmark.
    </Note>
  </Tab>

  <Tab title="Code">
    Start the import:

    <CodeGroup>
      ```python Python theme={null}
      from pinecone import Pinecone, ImportErrorMode

      pc = Pinecone(api_key="{{YOUR_API_KEY}}")
      index = pc.Index("search-10m")

      import_response = index.start_import(
          uri="s3://fe-customer-pocs/search/search_10M/dense/",
          error_mode=ImportErrorMode.ABORT 
      )
      print(f"Import started: {import_response['id']}")
      ```
    </CodeGroup>

    Monitor import progress:

    <CodeGroup>
      ```python Python theme={null}
      from pinecone import Pinecone
      import time 

      pc = Pinecone(api_key="{{YOUR_API_KEY}}")
      index = pc.Index("search-10m")

      while True:
          status = index.describe_import(id="IMPORT_ID")
          print(f"Status: {status['status']}, Progress: {status['percent_complete']:.1f}%")
          if status['status'] == "Completed":
              print("Import completed successfully!")
              break
          elif status['status'] == "Failed":
              print("Import failed. Check error details.")
              break
          time.sleep(15) # Check every 15 seconds
      ```
    </CodeGroup>

    <Note>
      There are three ways to find the import ID:

      * It's returned when the import is started
      * In the Pinecone console, on the **Imports** tab for your index
      * By calling [List imports](/reference/api/latest/data-plane/list_imports)
    </Note>
  </Tab>
</Tabs>

## 4. Run the benchmark

To simulate realistic query patterns and measure latency and throughput for your Pinecone index, use [Vector Search Bench (VSB)](https://github.com/pinecone-io/VSB). The benchmark runs 100,000 queries at 10 queries per second, which should take just under three hours to complete.

<Steps>
  <Step title="Provision a VM">
    VSB reports latency as the time from when the tool issues a query to when the query is returned by Pinecone.

    To minimize the client-side latency between the tool and Pinecone, run the benchmark on a dedicated AWS EC2 instance that's hosted in the same AWS region as your Pinecone index. This reduces the client-side latency to sub-millisecond range.

    <Note>
      As noted in [section 2](#2-create-an-index), this test requires an AWS EC2 instance in the same region as your index.
    </Note>

    For instructions on how to provision an EC2 instance, see the [AWS documentation](https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/LaunchingAndUsingInstances.html).

    <Note>
      Create a VM that comes with Python 3.11 or higher.
    </Note>
  </Step>

  <Step title="Connect to the VM">
    Connect to the VM using SSH or the cloud provider's console.
  </Step>

  <Step title="Install Vector Search Bench (VSB)">
    [VSB (Vector Search Bench)](https://github.com/pinecone-io/VSB) is a benchmarking suite for testing vector database search performance across different workloads and databases. To install it, you'll first need to install various dependencies.

    1. **Verify Python version**

       VSB requires Python 3.11 or higher to run. Verify your Python version:

       ```bash Terminal theme={null}
       python3 --version
       ```

       If your version is below 3.11, install Python 3.11+ using your distribution's package manager.

    2. **Install git**

       Git is required to clone the VSB repository. Check if git is installed:

       ```bash Terminal theme={null}
       git --version
       ```

       If git is not installed, install it using your system's package manager:

       ```bash Terminal theme={null}
       # Adapt for your VM's package manager (apt/yum/dnf)
       sudo apt-get update && sudo apt-get install git
       ```

    3. **Install pipx**

       pipx is required to install Poetry. First, check if pip3 is installed:

       ```bash Terminal theme={null}
       pip3 --version
       ```

       If pip is not installed, install it using your system's package manager:

       ```bash Terminal theme={null}
       # Adapt for your VM's package manager (apt/yum/dnf)
       sudo apt-get update && sudo apt-get install python3-pip
       ```

       Then check if pipx is installed:

       ```bash Terminal theme={null}
       pipx --version
       ```

       If pipx is not installed, install it via your system's package manager:

       ```bash Terminal theme={null}
       # Adapt for your VM's package manager (apt/yum/dnf)
       sudo apt-get update && sudo apt-get install pipx
       pipx ensurepath
       ```

       After installation, run this command to update the PATH in your current terminal session:

       ```bash Terminal theme={null}
       source ~/.bashrc
       ```

    4. **Install Poetry**

       [Poetry](https://python-poetry.org/) is required to manage [VSB's](https://github.com/pinecone-io/VSB) Python dependencies and virtual environment. If Poetry is not [installed](https://python-poetry.org/docs/), use pipx to install it:

       ```bash Terminal theme={null}
       pipx install poetry
       ```

       Alternatively, use the [official Poetry installer](https://python-poetry.org/docs/#installing-with-the-official-installer).

    5. **Clone the VSB repository**

       To run the benchmark, you'll first need to clone the VSB repository and navigate to it:

       ```bash Terminal theme={null}
       git clone https://github.com/pinecone-io/VSB.git
       cd VSB
       ```

    6. **Configure Poetry**

       Since your VM has Python 3.11 or higher installed (as specified in the VM provisioning step), tell Poetry to use it:

       ```bash Terminal theme={null}
       poetry env use python3
       ```

    7. **Install dependencies**

       VSB requires several Python packages to run. Install all dependencies:

       ```bash Terminal theme={null}
       poetry install
       ```
  </Step>

  <Step title="Benchmark your Pinecone index">
    To test the performance of your Pinecone index, run the following command from within the `VSB` directory. For more information about VSB, see its [GitHub repository](https://github.com/pinecone-io/VSB).

    The following command simulates 10 concurrent users issuing a total of 100,000 queries at 10 queries per second (QPS). Each query performs a vector search for the top 10 most similar 1024-dimensional vectors, using cosine similarity, with query vectors selected uniformly at random. The `--skip_populate` flag skips the data population phase, since you've already imported data into your index.

    ```bash Terminal theme={null}
    poetry run vsb \
        --database="pinecone" \
        --workload=synthetic-proportional \
        --pinecone_api_key="{{YOUR_API_KEY}}" \
        --pinecone_index_name="search-10m" \
        --pinecone_namespace_name="ns_2" \
        --synthetic_dimensions=1024 \
        --synthetic_metric=cosine \
        --synthetic_top_k=10 \
        --synthetic_requests=100000 \
        --users=10 \
        --requests_per_sec=10 \
        --synthetic_query_distribution=uniform \
        --synthetic_query_ratio=1 \
        --synthetic_insert_ratio=0 \
        --synthetic_delete_ratio=0 \
        --synthetic_update_ratio=0 \
        --skip_populate
    ```
  </Step>
</Steps>

## 5. Analyze performance

At the end of the run, VSB prints an operation summary including the requests per second achieved and latencies at different percentiles. Here's an example output:

```shell Terminal theme={null}
2025-12-23T00:34:37 INFO     Completed Run phase, took 9940.14s

                      Operation Summary

  Operation  Requests  Failures  Requests/sec  Failures/sec
 ───────────────────────────────────────────────────────────
  Search        99000     0(0%)            10           0.0

                                                    Metrics Summary

  Operation  Metric         Min  0.1%    1%    5%   10%   25%   50%   75%   90%   95%   99%  99.9%  99.99%   Max  Mean
 ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
  Search     Latency (ms)    23    25    25    26    27    27    29    34    44    81   350    430    1300  4602    43
```

Confirm that the requests per second achieved is around 10 QPS and the p90 latency is less than 100ms.

To see more detailed statistics, you can analyze the `stats.json` file identified in the output.

## 6. Check costs

You can check the costs for the import, queries, and storage in the Pinecone console at [Settings > Usage](https://app.pinecone.io/organizations/-/settings/usage). Cost data is delayed up to three days, but once it's available, compare the actual costs to the estimated costs below.

<Note>For the latest pricing details, see [Pricing](https://www.pinecone.io/pricing).</Note>

| Cost type | Amount          | Pricing                | Estimated cost |
| :-------- | :-------------- | :--------------------- | :------------- |
| Import    | 48.8 GB         | \$1/GB                 | \$48.80        |
| Queries   | 100,000 queries | \$16 per 1M read units | \$78.08        |
| Storage   | 4 hours         | \$0.33/GB/month        | \$0.09         |
| **Total** |                 |                        | **\$126.97**   |

<Steps>
  <Step title="Import costs">
    The current price for import is \$1/GB. The dataset size for this test is 48.8 GB, so the import cost should be \$48.80.
  </Step>

  <Step title="Query costs">
    A query uses 1 [read unit (RU)](/guides/manage-cost/understanding-cost#read-units) for every 1 GB of namespace size. The current price for queries in the `us-east-1` region of AWS is \$16 per 1 million read units (pricing varies by region).

    This test ran 100,000 queries against a namespace size of 48.8 GB. Each query uses 48.8 RUs (1 RU per GB), so the total is 4,880,000 RUs. At \$16 per 1 million RUs, the cost is (4,880,000 / 1,000,000) × \$16 = \$78.08.
  </Step>

  <Step title="Storage costs">
    The current price for storage is \$0.33 per GB per month. The dataset size for this test is 48.8 GB. Assuming a total storage time of 4 hours (including import, benchmark runtime, and cleanup), the storage cost is: \$0.33/GB/month \* 48.8 GB / 730 hours \* 4 hours = \$0.09.
  </Step>

  <Step title="Total costs">
    The total cost for the test is the sum of the import cost, query cost, and storage cost: \$48.80 + \$78.08 + \$0.09 = \$126.97.
  </Step>
</Steps>

## 7. Clean up

When you no longer need your test index, [delete it](/guides/manage-data/manage-indexes#delete-an-index) to avoid incurring unnecessary costs.

---

# Source: https://docs.pinecone.io/integrations/traceloop.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Traceloop

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

[Traceloop](https://www.traceloop.com/) provides observability for your LLM app using OpenTelemetry. Traceloop automatically monitors the quality of your LLM outputs. It helps you to debug and test changes to your models and prompts.

The Pinecone integration with Traceloop produces traces and metrics that can be viewed in any OpenTelemetry-based platform like Datadog, Grafana, Traceloop, and others.

<PrimarySecondaryCTA
  primaryHref={
  "https://www.traceloop.com/docs/openllmetry/getting-started-python"
}
  primaryLabel={"Get started"}
/>

---

# Source: https://docs.pinecone.io/integrations/trulens.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# TruLens

> Using TruLens and Pinecone to evaluate grounded LLM applications

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

TruLens is a powerful open source library for evaluating and tracking large language model-based applications. TruLens provides a set of tools for developing and monitoring neural nets, including large language models (LLMs). This includes both tools for evaluation of LLMs and LLM-based applications with TruLens-Eval and deep learning explainability with TruLens-Explain.

To build an effective RAG-style LLM application, it is important to experiment with various configuration choices while setting up your Pinecone vector database, and study their impact on performance metrics. Tracking and evaluation with TruLens enables fast iteration of your application.

<PrimarySecondaryCTA secondaryHref={"#setup-guide"} secondaryLabel={"View setup guide"} />

## Setup guide

[TruLens](https://github.com/truera/trulens) is a powerful open source library for evaluating and tracking large language model-based applications. In this guide, we will show you how to use TruLens to evaluate applications built on top of a high performance Pinecone vector database.

### Why TruLens?

Systematic evaluation is needed to support reliable, non-hallucinatory LLM-based applications. TruLens contains instrumentation and evaluation tools for large language model (LLM)-based applications. For evaluation, TruLens provides a set of feedback functions, analogous to labeling functions, to programmatically score the input, output and intermediate text of an LLM app. Each LLM application request can be scored on its question-answer relevance, context relevance and groundedness. These feedback functions provide evidence that your LLM-application is non-hallucinatory.

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/1.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=5e9bed118e6b241799bf579d0f528247" alt="diagram-1" data-og-width="1721" width="1721" data-og-height="598" height="598" data-path="images/1.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/1.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=08db4c06a2f1f3eab651e2ee2dce9fd8 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/1.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=616758e4f7f4dfb657d98142a287965e 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/1.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=9e89e1b1c185dddadc25637f87df30a2 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/1.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=3e1471b914e6cffb8d3c629084f01e1f 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/1.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=24f7f161ef9257050fdc062d7c638a5e 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/1.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=32f0d2e3ff4fabe3c7d407a0a3ac4a01 2500w" />

In addition to the above, feedback functions also support the evaluation of ground truth agreement, sentiment, model agreement, language match, toxicity, and a full suite of moderation evaluations, including hate, violence and more. TruLens implements feedback functions as an extensible framework that can evaluate your custom needs as well.

During the development cycle, TruLens supports the iterative development of a wide range of LLM applications by wrapping your application to log cost, latency, key metadata and evaluations of each application run. This allows you to track and identify failure modes, pinpoint their root cause, and measure improvement across experiments.

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/2.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=ab375c103797d6ed5cc3e6a4f86b637f" alt="application-screenshot" data-og-width="1494" width="1494" data-og-height="700" height="700" data-path="images/2.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/2.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=bb1da067913ac5f726dffd8a102357b4 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/2.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=4df448fe4e85641d732fc003fdb0caf6 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/2.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=d795ca7efa2aead62a861f1af8da237d 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/2.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6536021c3b233d53162fee1b82ddb5fa 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/2.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=8cb8cd60ab2f524180715c3e75521ed0 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/2.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=c7784c2edd556209898448807d4196c8 2500w" />

### Why Pinecone?

Large language models alone have a hallucination problem. Several decades of machine learning research have optimized models, including modern LLMs, for generalization, while actively penalizing memorization. However, many of today's applications require factual, grounded answers. LLMs are also expensive to train, and provided by third party APIs. This means the knowledge of an LLM is fixed. Retrieval-augmented generation (RAG) is a way to reliably ensure models are grounded, with Pinecone as the curated source of real world information, long term memory, application domain knowledge, or whitelisted data.

In the RAG paradigm, rather than just passing a user question directly to a language model, the system retrieves any documents that could be relevant in answering the question from the knowledge base, and then passes those documents (along with the original question) to the language model to generate the final response. The most popular method for RAG involves chaining together LLMs with vector databases, such as the widely used Pinecone vector DB.

In this process, a numerical vector (an embedding) is calculated for all documents, and those vectors are then stored in a database optimized for storing and querying vectors. Incoming queries are vectorized as well, typically using an encoder LLM to convert the query into an embedding. The query embedding is then matched via embedding similarity against the document embeddings in the vector database to retrieve the documents that are relevant to the query.

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/3.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=3c3b7194183c239a69db1989b77fa15d" alt="diagram-2" data-og-width="1364" width="1364" data-og-height="623" height="623" data-path="images/3.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/3.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=e5434da70f2237fe8735fe18ac1e2291 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/3.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=00cdb8549012faee0dd649322b9d5da1 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/3.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=a2e803adf53599748a030e9f89259891 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/3.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=4c8d0a28e052574131736cde74cf6479 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/3.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=3835b02af50749b0200cbd39a2de58d3 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/3.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6ab02341c8579fbf5c6b2a121ef52811 2500w" />

Pinecone makes it easy to build high-performance vector search applications, including retrieval-augmented question answering. Pinecone can easily handle very large scales of hundreds of millions and even billions of vector embeddings. Pinecone's large scale allows it to handle long term memory or a large corpus of rich external and domain-appropriate data so that the LLM component of RAG application can focus on tasks like summarization, inference and planning. This setup is optimal for developing a non-hallucinatory application.\
In addition, Pinecone is fully managed, so it is easy to change configurations and components. Combined with the tracking and evaluation with TruLens, this is a powerful combination that enables fast iteration of your application.

### Using Pinecone and TruLens to improve LLM performance and reduce hallucination

To build an effective RAG-style LLM application, it is important to experiment with various configuration choices while setting up the vector database, and study their impact on performance metrics.

In this example, we explore the downstream impact of some of these configuration choices on response quality, cost and latency with a sample LLM application built with Pinecone as the vector DB. The evaluation and experiment tracking is done with the [TruLens](https://www.trulens.org/) open source library. TruLens offers an extensible set of [feedback functions](https://truera.com/ai-quality-education/generative-ai-and-llms/whats-missing-to-evaluate-foundation-models-at-scale/) to evaluate LLM apps and enables developers to easily track their LLM app experiments.

In each component of this application, different configuration choices can be made that can impact downstream performance. Some of these choices include the following:

**Constructing the Vector DB**

* Data preprocessing and selection
* Chunk Size and Chunk Overlap
* Index distance metric
* Selection of embeddings

**Retrieval**

* Amount of context retrieved (top k)
* Query planning

**LLM**

* Prompting
* Model choice
* Model parameters (size, temperature, frequency penalty, model retries, etc.)

These configuration choices are useful to keep in mind when constructing your app. In general, there is no optimal choice for all use cases. Rather, we recommend that you experiment with and evaluate a variety of configurations to find the optimal selection as you are building your application.

#### Creating the index in Pinecone

Here we'll download a pre-embedded dataset from the `pinecone-datasets` library allowing us to skip the embedding and preprocessing steps.

```Python Python theme={null}
import pinecone_datasets

dataset = pinecone_datasets.load_dataset('wikipedia-simple-text-embedding-ada-002-100K')
dataset.head()
```

After downloading the data, we can initialize our pinecone environment and create our first index. Here, we have our first potentially important choice, by selecting the **distance metric** used for our index.

```Python Python theme={null}
pinecone.create_index(
        name=index_name_v1,
        metric='cosine', # We'll try each distance metric here.
        dimension=1536  # 1536 dim of text-embedding-ada-002.
)
```

Then, we can upsert our documents into the index in batches.

```Python Python theme={null}
for batch in dataset.iter_documents(batch_size=100):
    index.upsert(batch)
```

#### Build the vector store

Now that we've built our index, we can start using LangChain to initialize our vector store.

```Python Python theme={null}
embed = OpenAIEmbeddings(
    model='text-embedding-ada-002',
    openai_api_key=OPENAI_API_KEY
)

from langchain.vectorstores import Pinecone

text_field = "text"

# Switch back to a normal index for LangChain.
index = pinecone.Index(index_name_v1)

vectorstore = Pinecone(
    index, embed.embed_query, text_field
)
```

In RAG, we take the query as a question that is to be answered by an LLM, but the LLM must answer the question based on the information it receives from the `vectorstore`.

#### Initialize our RAG application

To do this, we initialize a `RetrievalQA` as our app:

```Python Python theme={null}
from langchain.chat_models import ChatOpenAI
from langchain.chains import RetrievalQA

# completion llm
llm = ChatOpenAI(
    model_name='gpt-3.5-turbo',
    temperature=0.0
)

qa = RetrievalQA.from_chain_type(
    llm=llm,
    chain_type="stuff",
    retriever=vectorstore.as_retriever()
)
```

#### TruLens for evaluation and tracking of LLM experiments

Once we've set up our app, we should put together our [feedback functions](https://truera.com/ai-quality-education/generative-ai-and-llms/whats-missing-to-evaluate-foundation-models-at-scale/). As a reminder, feedback functions are an extensible method for evaluating LLMs. Here we'll set up two feedback functions: `qs_relevance` and `qa_relevance`. They're defined as follows:

*QS Relevance: query-statement relevance is the average of relevance (0 to 1) for each context chunk returned by the semantic search.*
*QA Relevance: question-answer relevance is the relevance (again, 0 to 1) of the final answer to the original question.*

```Python Python theme={null}
# Imports main tools for eval
from trulens_eval import TruChain, Feedback, Tru, feedback, Select
import numpy as np
tru = Tru()

# OpenAI as feedback provider
openai = feedback.OpenAI()

# Question/answer relevance between overall question and answer.
qa_relevance = Feedback(openai.relevance).on_input_output()

# Question/statement relevance between question and each context chunk.
qs_relevance = 
Feedback(openai.qs_relevance).
on_input()
# See explanation below 
.on(Select.Record.app.combine_documents_chain._call.args.inputs.input_documents[:].page_content)
.aggregate(np.mean)

```

Our use of selectors here also requires an explanation.

QA Relevance is the simpler of the two. Here, we are using `.on_input_output()` to specify that the feedback function should be applied on both the input and output of the application.

For QS Relevance, we use TruLens selectors to locate the context chunks retrieved by our application. Let's break it down into simple parts:

1. Argument Specification – The `on_input` which appears first is a convenient shorthand and states that the first argument to `qs_relevance` (the question) is to be the main input of the app.

2. Argument Specification – The `on(Select...)` line specifies where the statement argument to the implementation comes from. We want to evaluate the context chunks, which are an intermediate step of the LLM app. This form references the langchain app object call chain, which can be viewed from `tru.run_dashboard()`. This flexibility allows you to apply a feedback function to any intermediate step of your LLM app. Below is an example where TruLens displays how to select each piece of the context.

   <img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/4.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=5822c0cf4ac339a43386120b9c8bdea0" alt="subcomponents" data-og-width="1850" width="1850" data-og-height="336" height="336" data-path="images/4.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/4.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=1491ad9e2af87a72d4d94eb7dbd1ab59 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/4.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=a3cfbb5990e305c451d5bf3a713d4d2e 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/4.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=e3a8b5e552edc5ece5c28d76d88aa264 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/4.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=3739a31a93ddd97eafed5fe7d3880551 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/4.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=e0b55c3382063504157437bbdb66f8d0 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/4.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=8d0811e7ecd13123debafd0ab0eeffac 2500w" />

3. Aggregation specification -- The last line aggregate (`np.mean`) specifies how feedback outputs are to be aggregated. This only applies to cases where the argument specification names more than one value for an input or output.

The result of these lines is that `f_qs_relevance` can be now be run on apps/records and will automatically select the specified components of those apps/records

To finish up, we just wrap our Retrieval QA app with TruLens along with a list of the feedback functions we will use for eval.

```Python Python theme={null}
# wrap with TruLens
truchain = TruChain(qa,
    app_id='Chain1_WikipediaQA',
    feedbacks=[qa_relevance, qs_relevance])

truchain(“Which state is Washington D.C. in?”)
```

After submitting a number of queries to our application, we can track our experiment and evaluations with the TruLens dashboard.

```Python Python theme={null}
tru.run_dashboard()
```

Here is a view of our first experiment:

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/5.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=25f3f2684029f23e72189334db906f49" alt="trulens-dashboard-1" data-og-width="1017" width="1017" data-og-height="143" height="143" data-path="images/5.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/5.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=835238582da8592aa840dff87c69d283 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/5.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=20e7c6e981a9657b8e4304ee221725dc 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/5.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=8c4423f3be2668ae666485c98c1b5a47 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/5.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=9f4eb89c6c00735276f5064000e4ce3c 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/5.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=dc0b96deb57e9ac8c711afe0698669d2 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/5.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=e25fde50e8e1c25a4d0d8edb102b6b6a 2500w" />

#### Experiment with distance metrics

Now that we've walked through the process of building our tracked RAG application using cosine as the distance metric, all we have to do for the next two experiments is to rebuild the index with `euclidean` or `dotproduct` as the metric and follow the rest of the steps above as is.

Because we are using OpenAI embeddings, which are normalized to length 1, dot product and cosine distance are equivalent - and Euclidean will also yield the same ranking. See the OpenAI docs for more information. With the same document ranking, we should not expect a difference in response quality, but computation latency may vary across the metrics. Indeed, OpenAI advises that dot product computation may be a bit faster than cosine. We will be able to confirm this expected latency difference with TruLens.

```Python Python theme={null}
index_name_v2 = 'langchain-rag-euclidean'
pinecone.create_index(
        name=index_name_v2,
        metric='euclidean', # metric='dotproduct',
        dimension=1536,  # 1536 dim of text-embedding-ada-002
    )
```

After doing so, we can view our evaluations for all three LLM apps sitting on top of the different indexes. All three apps are struggling with query-statement relevance. In other words, the context retrieved is only somewhat relevant to the original query.

**We can also see that both the Euclidean and dot-product metrics performed at a lower latency than cosine at roughly the same evaluation quality.**

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=cf0234cb8924dc23cf364387f9126881" alt="trulens-dashboard-2" data-og-width="1013" width="1013" data-og-height="523" height="523" data-path="images/6.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=0d1fd6b71d55bf6984310e89b2ecb5d0 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=0320201157b8d94a87114b3cd83590c3 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=1f3790b99f600f60a7bffc26043d85be 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6888d3ded645a16c00f8b050cac5b76f 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=7b309425a15f78d512e6fe6f31136bb0 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/6.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=27ce449f1ece420647c93eb824a09809 2500w" />

### Problem: hallucination

Digging deeper into the Query Statement Relevance, we notice one problem in particular with a question about famous dental floss brands. The app responds correctly, but is not backed up by the context retrieved, which does not mention any specific brands.

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/7.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=9de7098b912d0249ba900299660f921b" alt="trulens-dashboard-feedback-1" data-og-width="586" width="586" data-og-height="415" height="415" data-path="images/7.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/7.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=bf334c79229d877737f0e3c185ca4fc1 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/7.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=04cf66c99f0a9eb630dccd679d8e449b 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/7.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=748327f7716489fcb324eab2cef25530 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/7.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=206f5881eccf29d1cd20f237f7c2fb70 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/7.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=da5fe01f0e281b7b8d9a2d24cf6ab2b9 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/7.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=993e9de4e3545808d384fd2939dd5eb7 2500w" />

#### Quickly evaluate app components with LangChain and TruLens

Using a less powerful model is a common way to reduce hallucination for some applications. We'll evaluate ada-001 in our next experiment for this purpose.

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/8.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=7dea53e93c7189bb81de5a2d331961a0" alt="trulens-dashboard-3" data-og-width="992" width="992" data-og-height="128" height="128" data-path="images/8.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/8.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=e49e676788190ab84141fcc7ccf6793d 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/8.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=ad8b8a8f72e58a446be8b40857d82ac6 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/8.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=795c0bbbef65f993640e8cfe866d5580 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/8.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=367b1f5de867b4dcebf896ffda9be192 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/8.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=165d075015baf3a519f38e5cc7d28082 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/8.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=f949f8604d0bfa2008304a6a3b893eb2 2500w" />

Changing different components of apps built with frameworks like LangChain is really easy. In this case we just need to call `text-ada-001` from the LangChain LLM store. Adding in easy evaluation with TruLens allows us to quickly iterate through different components to find our optimal app configuration.

```Python Python theme={null}
# completion llm
from langchain.llms import OpenAI

llm = OpenAI(
    model_name='text-ada-001',
    temperature=0
)

from langchain.chains import RetrievalQAWithSourcesChain
qa_with_sources = RetrievalQA.from_chain_type(
    llm=llm,
    chain_type="stuff",
    retriever=vectorstore.as_retriever()
)

# wrap with TruLens
truchain = TruChain(qa_with_sources,
    app_id='Chain4_WikipediaQA',
    feedbacks=[qa_relevance, qs_relevance])
```

**However, this configuration with a less powerful model struggles to return a relevant answer given the context provided.**

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/9.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=73f93e1a26ded0ae6a52c5a9a750884f" alt="trulens-dashboard-4" data-og-width="1022" width="1022" data-og-height="147" height="147" data-path="images/9.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/9.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=3e4b4149212a2da4004067d610f0afec 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/9.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=cd650d3ae9c67b99aacd1738d22ea742 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/9.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=f0a5a73fc8c792fff9d462110328382c 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/9.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=9bba67d6abfaa9fbfe752995ace04cfc 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/9.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=607a0defab012a6e6095d0667f358bb1 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/9.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=7783a78d4ffd11c269d2a297f9bc0ecb 2500w" />

For example, when asked “Which year was Hawaii's state song written?”, the app retrieves context that contains the correct answer but fails to respond with that answer, instead simply responding with the name of the song.

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/10.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=20bd62ec726af77b872cfd34b3f6370e" alt="trulens-dashboard-feedback-2" data-og-width="648" width="648" data-og-height="430" height="430" data-path="images/10.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/10.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=e0ed4a0b3a43bbbf1032717a69543790 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/10.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=b21ca5cfacc36d408d3cb9fd9b341484 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/10.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=0908b2ad2fe0fff3f2733917293afdb6 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/10.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=5770a6f19f8856971415daaa5fd12a54 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/10.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=736c2727fa3f5f8c62f6236d8327b166 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/10.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=a17ccd36ca1a0d7792674c18b4c90b33 2500w" />

While our relevance function is not doing a great job here in differentiating which context chunks are relevant, we can manually see that only the one (the 4th chunk) mentions the year the song was written. Narrowing our `top_k`, or the number of context chunks retrieved by the semantic search, may help.

We can do so as follows:

```Python Python theme={null}
qa = RetrievalQA.from_chain_type(
    llm=llm,
    chain_type="stuff",
    retriever=vectorstore.as_retriever(top_k = 1)
)
```

The way the `top_k` is implemented in LangChain's RetrievalQA is that the documents are still retrieved by semantic search and only the `top_k` are passed to the LLM. Therefore, TruLens also captures all of the context chunks that are being retrieved. In order to calculate an accurate QS Relevance metric that matches what's being passed to the LLM, we only calculate the relevance of the top context chunk retrieved by slicing the `input_documents` passed into the TruLens Select function:

```Python Python theme={null}
qs_relevance = Feedback(openai.qs_relevance).on_input().on(
    Select.Record.app.combine_documents_chain._call.args.inputs.input_documents[:1].page_content
).aggregate(np.mean)
```

Once we've done so, our final application has much improved `qs_relevance`, `qa_relevance` and latency!

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/11.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=8d020d3573789f3a84a33ab2454c6017" alt="trulens-dashboard-5" data-og-width="832" width="832" data-og-height="114" height="114" data-path="images/11.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/11.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=fa84e46a4c4ca2de65e52a3a4ff7510e 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/11.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=bcb077e2814852755c8052d98a1d2c00 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/11.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=b8518e0ca388d9b1379507f2f5cc8288 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/11.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=4b0f852fb9db98a5b73bba39805b95d2 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/11.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=9540196c0ca72a2db5bc3d9f4fd9bd4f 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/11.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=fb17431f19c302f6be5ebe9d5cc67372 2500w" />

With that change, our application is successfully retrieving the one piece of context it needs, and successfully forming an answer from that context.

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/12.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=d304b502bf601fb1f4fe24fb05daa325" alt="trulens-dashboard-feedback-3" data-og-width="653" width="653" data-og-height="413" height="413" data-path="images/12.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/12.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=ad0f9bd7fcfacd0d8964f94b5dcf9944 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/12.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=fc2e4954cd701b53def3ed93f25a600c 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/12.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=7511483fb00d8a089205ac2654e9719e 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/12.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=c03adca2cbba373c46e3b426621e1696 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/12.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=1440bace9aedb35a9d82d962f18e9e2d 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/12.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=8f8934d528f22980ca4a4337d3db5d53 2500w" />

Even better, the application now knows what it doesn't know:

<img src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/13.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=9752560cdd280531c09feecf1d4492af" alt="trulens-dashboard-feedback-4" data-og-width="494" width="494" data-og-height="217" height="217" data-path="images/13.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/13.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=0fc55c8534f61a5e58e505c4dc6ecbf9 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/13.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=14629e78fcadde9f8d782c4e5a0bda3a 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/13.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=f6d11230665696a6fe965c6da2333d6e 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/13.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=6a9c57780f6a5f0678ce4de25fdee4e3 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/13.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=499aad55bf8c556818dc9cf184f77055 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/13.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=84973891f083f8a33d45090fa0b7c20e 2500w" />

### Summary

In conclusion, we note that exploring the downstream impact of some Pinecone configuration choices on response quality, cost and latency is an important part of the LLM app development process, ensuring that we make the choices that lead to the app performing the best. Overall, TruLens and Pinecone are the perfect combination for building reliable RAG-style applications. Pinecone provides a way to efficiently store and retrieve context used by LLM apps, and TruLens provides a way to track and evaluate each iteration of your application.

---

# Source: https://docs.pinecone.io/integrations/twelve-labs.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Twelve Labs

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

[Twelve Labs](https://twelvelabs.io) is an AI company that provides state-of-the-art video understanding capabilities through its easy-to-use APIs. Our newly released product is the Embed API, which enables developers to create high-quality multimodal embeddings that capture the rich context and interactions between different modalities in videos, such as visual expressions, body language, spoken words, and overall context.

By integrating Twelve Labs' Embed API with Pinecone's vector database, developers can efficiently store, index, and retrieve these multimodal embeddings at scale. This integration empowers developers to build cutting-edge AI applications that leverage video data, such as video search, recommendation systems, content moderation, and more. Developers can seamlessly generate embeddings using Twelve Labs' API and store them in Pinecone for fast and accurate similarity search and retrieval.

The integration of Twelve Labs and Pinecone offers developers a powerful toolkit to process and understand video content in a more human-like manner. By combining Twelve Labs' video-native approach with Pinecone's purpose-built vector search capabilities, developers can unlock new possibilities and build innovative applications across various industries, including media and entertainment, e-commerce, education, and beyond.

<PrimarySecondaryCTA secondaryHref={"#setup-guide"} secondaryLabel={"View setup guide"} />

## Setup guide

To integrate Twelve Labs' Embed API with Pinecone:

1. Sign up for a [Twelve Labs](https://twelvelabs.io) and obtain your API key.
2. Install the [Twelve Labs Python client library](https://github.com/twelvelabs-io/twelvelabs-python).
3. Sign up for a [Pinecone account](https://app.pinecone.io/) and [create an index](/guides/index-data/create-an-index).
4. Install the [Pinecone client library](/reference/pinecone-sdks).
5. Use the [Twelve Labs Embed API](https://docs.twelvelabs.io/docs/create-embeddings) to generate multimodal embeddings for your videos.
6. Connect to your Pinecone index and [upsert the embeddings](/guides/index-data/upsert-data).
7. [Query the Pinecone index](/guides/search/search-overview) to retrieve similar videos based on embeddings.

For more detailed information and code examples, please see the [Twelve Labs documentation](https://docs.twelvelabs.io).

---

# Source: https://docs.pinecone.io/troubleshooting/unable-to-pip-install.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Unable to pip install

Python `3.x` uses `pip3`. Use the following commands in your terminal to install the latest version of the [Pinecone Python SDK](/reference/sdks/python/overview):

```Shell Shell theme={null}
# If you are connecting to Pinecone via gRPC:
pip3 install -U pinecone[grpc]
```

```Shell Shell theme={null}
# If you are connecting to Pinecone via HTTP:
pip3 install -U pinecone
```

---

# Source: https://docs.pinecone.io/guides/indexes/pods/understanding-collections.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Understanding collections

> Create static backups of pod-based indexes with collections

<Warning>
  Customers who sign up for a Standard or Enterprise plan on or after August 18, 2025 cannot create pod-based indexes. Instead, create [serverless indexes](/guides/index-data/create-an-index), and consider using [dedicated read nodes](/guides/index-data/dedicated-read-nodes) for large workloads (millions of records or more, and moderate or high query rates).
</Warning>

A collection is a static copy of a pod-based index that only consumes storage. It is a non-queryable representation of a set of records. You can [create a collection](/guides/indexes/pods/back-up-a-pod-based-index) of a pod-based index, and you can [create a new pod-based index from a collection](/guides/manage-data/restore-an-index). This allows you to restore the index with the same or different configurations.

<Note>
  Once a collection is created, it cannot be moved to a different project.
</Note>

<PuPr />

## Use cases

Creating a collection is useful when performing tasks like the following:

* Protecting an index from manual or system failures.
* Temporarily shutting down an index.
* Copying the data from one index into a different index.
* Making a backup of your index.
* Experimenting with different index configurations.

## Performance

Collections operations perform differently, depending on the pod type of the index:

* Creating a `p1` or `s1` index from a collection takes approximately 10 minutes.
* Creating a `p2` index from a collection can take several hours when the number of vectors is on the order of 1,000,000.

## Limitations

Collection limitations are as follows:

* You can only perform operations on collections in the current Pinecone project.

## Pricing

See [Pricing](https://www.pinecone.io/pricing/) for up-to-date pricing information.

---

# Source: https://docs.pinecone.io/guides/manage-cost/understanding-cost.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Understanding cost

> Understand how costs are incurred in Pinecone.

For the latest pricing details, see [Pricing](https://www.pinecone.io/pricing/).

## Minimum usage

The Standard and Enterprise [pricing plans](https://www.pinecone.io/pricing/) include a monthly minimum usage committment:

| Plan       | Minimum usage |
| ---------- | ------------- |
| Starter    | \$0/month     |
| Standard   | \$50/month    |
| Enterprise | \$500/month   |

Beyond the monthly minimum, customers are charged for what they use each month.

**Examples**

<AccordionGroup>
  <Accordion title="Usage below monthly minimum">
    * You are on the Standard plan.
    * Your usage for the month of August amounts to \$20.
    * Your usage is below the \$50 monthly minimum, so your total for the month is \$50.

    In this case, the August invoice would include line items for each service you used (totaling \$20), plus a single line item covering the rest of the minimum usage commitment (\$30).
  </Accordion>

  <Accordion title="Usage exceeds monthly minimum">
    * You are on the Standard plan.
    * Your usage for the month of August amounts to \$100.
    * Your usage exceeds the \$50 monthly minimum, so your total for the month is \$100.

    In this case, the August invoice would only show line items for each service you used (totaling \$100). Since your usage exceeds the minimum usage commitment, you are only charged for your actual usage and no additional minimum usage line item appears on your invoice.
  </Accordion>
</AccordionGroup>

## Discounts

Pinecone offers a discount for customers who enroll in an annual plan with an upfront payment. Eligibility requires a minimum annual commitment of \$8,000.

Throughout the contract term, the discount applies to List Price for all Pinecone services, excluding Storage and Support. Any usage that exceeds the prepaid credits will be billed at full List Price and will not qualify for the contractual discount.

Customers on the Standard and Enterprise pay-as-you-go plans may upgrade directly by navigating in the Pinecone console to **[Settings > Billing > Plans](https://app.pinecone.io/organizations/-/settings/billing/plans)**.

<Note>
  The self-serve annual plan is not available through cloud marketplace billing. To purchase an annual plan through a cloud marketplace, contact [ar@pinecone.io](mailto:ar@pinecone.io).
</Note>

## Serverless indexes

With serverless indexes, you pay for the amount of data stored and operations performed, based on three usage metrics: [read units](#read-units), [write units](#write-units), and [storage](#storage).

For the latest serverless pricing rates, see [Pricing](https://www.pinecone.io/pricing/).

### Read units

Read units (RUs) measure the compute, I/O, and network resources consumed by the following read requests:

* [Query](#query)
* [Fetch](#fetch)
* [List](#list)

<Tip>
  Read requests return the number of RUs used. You can use this information to [monitor read costs](/guides/manage-cost/monitor-usage-and-costs#read-units).
</Tip>

<Note>
  Indexes built on [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) are not subject to read unit limits for query, fetch, and list operations. For sizing and capacity planning guidance, see the [Dedicated Read Nodes](/guides/index-data/dedicated-read-nodes) guide.
</Note>

#### Query

The cost of a query scales linearly with the size of the targeted namespace. Specifically, a query uses 1 RU for every 1 GB of namespace size, with a minimum of 0.25 RUs per query.

| Namespace size | Read units per query |
| :------------- | :------------------- |
| \< 0.25 GB     | 0.25 RUs (minimum)   |
| 1 GB           | 1 RU                 |
| 10 GB          | 10 RUs               |
| 50 GB          | 50 RUs               |
| 100 GB         | 100 RUs              |

To learn how to calculate your namespace size, see [Storage](#storage).

<Note>
  Parameters that affect the size of the query response, such as `top_k`, `include_metadata`, and `include_values`, are not relevant for query cost; only the size of the namespace determines the number of RUs used.
</Note>

#### Fetch

A fetch request uses 1 RU for every 10 records fetched, for example:

| Fetched records | RUs |
| --------------- | --- |
| 10              | 1   |
| 50              | 5   |
| 107             | 11  |

Specifying a non-existent ID or adding the same ID more than once does not increase the number of RUs used. However, a fetch request will always use at least 1 RU.

<Note>
  [Fetching records by metadata](/guides/manage-data/fetch-data#fetch-records-by-metadata) uses the same cost model as fetching by ID: 1 RU for every 10 records fetched.
</Note>

#### List

List has a fixed cost of 1 RU per call, with up to 100 records per call.

### Write units

Write units (WUs) measure the storage and compute resources used by the following write requests:

* [Upsert](#upsert)
* [Update](#update)
* [Delete](#delete)

#### Upsert

An upsert request uses 1 WU for each 1 KB of the request, with a minimum of 5 WUs per request. When an upsert modifies an existing record, the request uses 1 WU for each 1 KB of the existing record as well.

For example, the following table shows the WUs used by upsert requests at different batch sizes and record sizes, assuming all records are new:

| Records per batch | Dimension | Avg. metadata size | Avg. record size | WUs  |
| :---------------- | :-------- | :----------------- | :--------------- | :--- |
| 1                 | 768       | 100 bytes          | 3.2 KB           | 5    |
| 2                 | 768       | 100 bytes          | 3.2 KB           | 7    |
| 10                | 1024      | 15,000 bytes       | 19.10 KB         | 191  |
| 100               | 768       | 500 bytes          | 3.57 KB          | 357  |
| 1000              | 1536      | 1000 bytes         | 7.14 KB          | 7140 |

#### Update

An update request uses 1 WU for each 1 KB of the new and existing record, with a minimum of 5 WUs per request.

For example, the following table shows the WUs used by an update at different record sizes:

| New record size | Previous record size | WUs |
| :-------------- | :------------------- | :-- |
| 6.24 KB         | 6.50 KB              | 13  |
| 19.10 KB        | 15 KB                | 25  |
| 3.57 KB         | 5 KB                 | 9   |
| 7.14 KB         | 10 KB                | 18  |
| 3.17 KB         | 3.17 KB              | 7   |

<Note>
  [Updating records by metadata](/guides/manage-data/update-data#update-by-metadata) uses the same cost model as updating by ID: 1 WU for each 1 KB of the new and existing record.
</Note>

#### Delete

A delete request uses 1 WU for each 1 KB of records deleted, with a minimum of 5 WUs per request.

For example, the following table shows the WUs used by delete requests at different batch sizes and record sizes:

| Records per batch | Dimension | Avg. metadata size | Avg. record size | WUs  |
| :---------------- | :-------- | :----------------- | :--------------- | :--- |
| 1                 | 768       | 100 bytes          | 3.2 KB           | 5    |
| 2                 | 768       | 100 bytes          | 3.2 KB           | 7    |
| 10                | 1024      | 15,000 bytes       | 19.10 KB         | 191  |
| 100               | 768       | 500 bytes          | 3.57 KB          | 357  |
| 1000              | 1536      | 1000 bytes         | 7.14 KB          | 7140 |

Specifying a non-existent ID or adding the same ID more than once does not increase WU use.

[Deleting a namespace](/guides/manage-data/manage-namespaces#delete-a-namespace) or [deleting all records in a namespace using `deleteAll`](/guides/manage-data/delete-data#delete-all-records-in-a-namespace) uses 5 WUs.

<Note>
  [Deleting records by metadata](/guides/manage-data/delete-data#delete-records-by-metadata) uses the same cost model as deleting by ID: 1 WU for each 1 KB of records deleted.
</Note>

### Storage

Storage costs are based on the size of an index on a per-gigabyte (GB) monthly rate. The size of an index is defined as the total size of its records across all namespaces. For the latest storage pricing rates, see [Pricing](https://www.pinecone.io/pricing/).

A record can include a dense vector, a sparse vector, or both. Use the formula for your index type to calculate total size:

<div className="border border-gray-200 dark:border-gray-700 rounded-lg p-4 my-4">
  <Tabs>
    <Tab title="Dense index">
      A [dense index](/guides/index-data/indexing-overview#dense-indexes) contains records with one dense vector each.

      <Note>
        Dense index records can also contain sparse vectors (when the index metric is set to `dotproduct`), which can be useful for [hybrid search](/guides/search/hybrid-search#use-a-single-hybrid-index). To learn how to calculate the size of a hybrid index, see [Hybrid index](#hybrid-index).
      </Note>

      **Calculate dense index size (assuming no sparse vectors)**

      ```
      Index size = Number of records × (
                     ID size + 
                     Metadata size +
                     Dense vector dimensions × 4 bytes
                   )
      ```

      Where:

      * `ID size` and `Metadata size` are measured in bytes, averaged across all records.
      * Each `Dense vector dimension` uses 4 bytes.

      **Example dense index calculations**

      These examples assume 8-byte IDs:

      | Records    | Dense vector dimensions | Avg metadata size | Index size |
      | :--------- | :---------------------- | :---------------- | :--------- |
      | 500,000    | 768                     | 500 bytes         | 1.79 GB    |
      | 1,000,000  | 1536                    | 1,000 bytes       | 7.15 GB    |
      | 5,000,000  | 1024                    | 15,000 bytes      | 95.5 GB    |
      | 10,000,000 | 1536                    | 1,000 bytes       | 71.5 GB    |

      <Note>
        Example: 500,000 records × (8-byte ID + (768 dense vector dimensions × 4 bytes) + 500 bytes of metadata) = 1.79 GB
      </Note>
    </Tab>

    <Tab title="Sparse index">
      A [sparse index](/guides/index-data/indexing-overview#sparse-indexes) contains records with one sparse vector each.

      **Calculate sparse index size**

      ```
      Index size = Number of records × (
                     ID size + 
                     Metadata size +
                     Number of non-zero sparse values × 8 bytes
                   )
      ```

      Where:

      * `ID size` and `Metadata size` are measured in bytes, averaged across all records.
      * `Number of non-zero sparse values`: Average number across all records. To find the count for a single record, check the length of the sparse vector's `indices` or `values` array. Each non-zero value uses 8 bytes.

      **Example sparse index calculations**

      These examples assume 8-byte IDs:

      | Records    | Avg number of non-zero sparse values | Avg metadata size | Index size |
      | :--------- | :----------------------------------- | :---------------- | :--------- |
      | 500,000    | 10                                   | 500 bytes         | 0.29 GB    |
      | 1,000,000  | 50                                   | 1,000 bytes       | 1.41 GB    |
      | 5,000,000  | 100                                  | 15,000 bytes      | 79.0 GB    |
      | 10,000,000 | 50                                   | 1,000 bytes       | 14.1 GB    |

      <Note>
        Example: 500,000 records × (8-byte ID + (10 non-zero sparse values × 8 bytes) + 500 bytes of metadata) = 0.29 GB
      </Note>
    </Tab>

    <Tab title="Hybrid index">
      A [hybrid index](/guides/search/hybrid-search#use-a-single-hybrid-index) contains records that each have one dense vector and an optional sparse vector.

      **Calculate hybrid index size**

      ```
      Index size = Number of records × (
                     ID size + 
                     Metadata size +
                     Dense vector dimensions × 4 bytes + 
                     Number of non-zero sparse values × 8 bytes
                   )
      ```

      Where:

      * `ID size` and `Metadata size` are measured in bytes, averaged across all records.
      * Each `Dense vector dimension` uses 4 bytes.
      * `Number of non-zero sparse values`: Average number across all records, including those without sparse vectors. To find the count for a single record, check the length of the sparse vector's `indices` or `values` array. Each non-zero value uses 8 bytes.

      **Example hybrid index calculations**

      These examples assume 8-byte IDs:

      | Records    | Dense vector dimensions | Avg number of non-zero sparse values | Avg metadata size | Index size |
      | :--------- | :---------------------- | :----------------------------------- | :---------------- | :--------- |
      | 500,000    | 768                     | 10                                   | 500 bytes         | 1.83 GB    |
      | 1,000,000  | 1536                    | 50                                   | 1,000 bytes       | 7.54 GB    |
      | 5,000,000  | 1024                    | 100                                  | 15,000 bytes      | 99.5 GB    |
      | 10,000,000 | 1536                    | 50                                   | 1,000 bytes       | 75.4 GB    |

      <Note>
        Example: 500,000 records × (8-byte ID + (768 dense vector dimensions × 4 bytes) + (10 non-zero sparse values × 8 bytes) + 500 bytes of metadata) = 1.83 GB
      </Note>
    </Tab>
  </Tabs>
</div>

## Imports

[Importing from object storage](/guides/index-data/import-data) is the most efficient and cost-effective method to load large numbers of records into an index. The cost of an import is based on the size of the records read, whether the records were imported successfully or not.

If the import operation fails (e.g., after encountering a vector of the wrong dimension in an import with `on_error="abort"`), you will still be charged for the records read. However, if the import fails because of an internal system error, you will not incur charges. In this case, the import will return the error message `"We were unable to process your request. If the problem persists, please contact us at https://support.pinecone.io"`.

For the latest import pricing rates, see [Pricing](https://www.pinecone.io/pricing/).

## Backups and restores

A [backup](/guides/manage-data/backups-overview) is a static copy of a serverless index. Both the cost of storing a backup and [restoring an index](/guides/manage-data/restore-an-index) from a backup is based on the size of the index. For the latest backup and restore pricing rates, see [Pricing](https://www.pinecone.io/pricing/).

## Embedding

Pinecone hosts several [embedding models](/guides/index-data/create-an-index#embedding-models) so it's easy to manage your vector storage and search process on a single platform. You can use a hosted model to embed your data as an integrated part of upserting and querying, or you can use a hosted model to embed your data as a standalone operation.

Embedding costs are determined by how many [tokens](https://www.pinecone.io/learn/tokenization/) are in a request. In general, the more words contained in your passage or query, the more tokens you generate.

For example, if you generate embeddings for the query, "What is the maximum diameter of a red pine?", Pinecone Inference generates 10 tokens, then converts them into an embedding. If the price per token for your billing plan is \$.08 per million tokens, then this API call costs \$.00001.

To learn more about tokenization, see [Choosing an embedding model](https://www.pinecone.io/learn/series/rag/embedding-models-rundown/). For the latest embed pricing rates, see [Pricing](https://www.pinecone.io/pricing/).

<Tip>
  Embedding requests returns the total tokens generated. You can use this information to [monitor and manage embedding costs](/guides/manage-cost/monitor-usage-and-costs#embedding-tokens).
</Tip>

## Reranking

Pinecone hosts several [reranking models](/guides/search/rerank-results#reranking-models) so it's easy to manage two-stage vector retrieval on a single platform. You can use a hosted model to rerank results as an integrated part of a query, or you can use a hosted model to rerank results as a standalone operation.

Reranking costs are determined by the number of requests to the reranking model. For the latest rerank pricing rates, see [Pricing](https://www.pinecone.io/pricing/).

## Assistant

For details on how costs are incurred in Pinecone Assistant, see [Assistant pricing](/guides/assistant/pricing-and-limits).

## See also

* [Manage cost](/guides/manage-cost/manage-cost)
* [Monitor usage](/guides/manage-cost/monitor-usage-and-costs)
* [Pricing](https://www.pinecone.io/pricing/)

---

# Source: https://docs.pinecone.io/guides/organizations/understanding-organizations.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Understanding organizations

> Understand organization structure, projects, and billing.

A Pinecone organization is a set of [projects](/guides/projects/understanding-projects) that use the same billing. Organizations allow one or more users to control billing and project permissions for all of the projects belonging to the organization. Each project belongs to an organization.

<Note>
  While an email address can be associated with multiple organizations, it cannot be used to create more than one organization. For information about managing organization members, see [Manage organization members](/guides/organizations/manage-organization-members).
</Note>

## Projects in an organization

Each organization contains one or more projects that share the same organization owners and billing settings. Each project belongs to exactly one organization. If you need to move a project from one organization to another, [contact Support](https://app.pinecone.io/organizations/-/settings/support/ticket).

## Billing settings

All of the projects in an organization share the same billing method and settings. The billing settings for the organization are controlled by the organization owners.

Organization owners can update the billing contact information, update the payment method, and view and download invoices using the [Pinecone console](https://app.pinecone.io/organizations/-/settings/billing).

## Organization roles

Organization owners can manage access to their organizations and projects by assigning roles to organization members and service accounts. The role determines the entity's permissions within Pinecone. The organization roles are as follows:

* **Organization owner**: Organization owners have global permissions across the organization. This includes managing billing details, organization members, and all projects. Organization owners are automatically [project owners](/guides/projects/understanding-projects#project-roles) and, therefore, have all project owner permissions as well.

* **Organization user**: Organization users have restricted organization-level permissions. When inviting organization users, you also choose the projects they belong to and the project role they should have.

* **Billing admin**: Billing admins have permissions to view and update billing details, but they cannot manage organization members. Billing admins cannot manage projects unless they are also [project owners](/guides/projects/understanding-projects#project-roles).

The following table summarizes the permissions for each organization role:

| Permission                           | Org Owner | Org User | Billing Admin |
| ------------------------------------ | --------- | -------- | ------------- |
| View account details                 | ✓         | ✓        | ✓             |
| Update organization name             | ✓         |          |               |
| Delete the organization              | ✓         |          |               |
| View billing details                 | ✓         |          | ✓             |
| Update billing details               | ✓         |          | ✓             |
| View usage details                   | ✓         |          | ✓             |
| View support plans                   | ✓         |          | ✓             |
| Invite members to the organization   | ✓         |          |               |
| Delete pending member invites        | ✓         |          |               |
| Remove members from the organization | ✓         |          |               |
| Update organization member roles     | ✓         |          |               |
| Create projects                      | ✓         | ✓        |               |

## Organization single sign-on (SSO)

SSO allows organizations to manage their teams' access to Pinecone through their identity management solution. Once your integration is configured, you can specify a default role for teammates when they sign up.

For more information, see [Configure single sign-on](/guides/production/configure-single-sign-on/okta).

<Note>SSO is available on Standard and Enterprise plans.</Note>

## Service accounts

<Note>
  This feature is in [public preview](/release-notes/feature-availability) and available only on [Enterprise plans](https://www.pinecone.io/pricing/).
</Note>

[Service accounts](/guides/organizations/manage-service-accounts) enable programmatic access to Pinecone's Admin API, which can be used to create and manage projects and API keys.

Use service accounts to automate infrastructure management and integrate Pinecone into your deployment workflows, rather than through manual actions in the Pinecone console. Service accounts use the [organization roles](/guides/organizations/understanding-organizations#organization-roles) and [project role](/guides/projects/understanding-projects#project-roles) for permissioning, and provide a secure and auditable way to handle programmatic access.

## See also

* [Manage organization members](/guides/organizations/manage-organization-members)
* [Manage project members](/guides/projects/manage-project-members)

---

# Source: https://docs.pinecone.io/guides/indexes/pods/understanding-pod-based-indexes.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Understanding pod-based indexes

> Learn about pod-based index architecture and types

<Warning>
  Customers who sign up for a Standard or Enterprise plan on or after August 18, 2025 cannot create pod-based indexes. Instead, create [serverless indexes](/guides/index-data/create-an-index), and consider using [dedicated read nodes](/guides/index-data/dedicated-read-nodes) for large workloads (millions of records or more, and moderate or high query rates).
</Warning>

With pod-based indexes, you choose one or more pre-configured units of hardware (pods). Depending on the pod type, pod size, and number of pods used, you get different amounts of storage and higher or lower latency and throughput. Be sure to [choose an appropriate pod type and size](/guides/indexes/pods/choose-a-pod-type-and-size) for your dataset and workload.

## Pod types

Different pod types are priced differently. See [Understanding cost](/guides/manage-cost/understanding-cost) for more details.

<Note>
  Once a pod-based index is created, you cannot change its pod type. However, you can create a collection from an index and then [create a new index with a different pod type](/guides/indexes/pods/create-a-pod-based-index#create-a-pod-index-from-a-collection) from the collection.
</Note>

### s1 pods

These storage-optimized pods provide large storage capacity and lower overall costs with slightly higher query latencies than p1 pods. They are ideal for very large indexes with moderate or relaxed latency requirements.

Each s1 pod has enough capacity for around 5M vectors of 768 dimensions.

### p1 pods

These performance-optimized pods provide very low query latencies, but hold fewer vectors per pod than s1 pods. They are ideal for applications with low latency requirements (\<100ms).

Each p1 pod has enough capacity for around 1M vectors of 768 dimensions.

### p2 pods

The p2 pod type provides greater query throughput with lower latency. For vectors with fewer than 128 dimension and queries where `topK` is less than 50, p2 pods support up to 200 QPS per replica and return queries in less than 10ms. This means that query throughput and latency are better than s1 and p1.

Each p2 pod has enough capacity for around 1M vectors of 768 dimensions. However, capacity may vary with dimensionality.

The data ingestion rate for p2 pods is significantly slower than for p1 pods; this rate decreases as the number of dimensions increases. For example, a p2 pod containing vectors with 128 dimensions can upsert up to 300 updates per second; a p2 pod containing vectors with 768 dimensions or more supports upsert of 50 updates per second. Because query latency and throughput for p2 pods vary from p1 pods, test p2 pod performance with your dataset.

The p2 pod type does not support sparse vector values.

## Pod size and performance

Each pod type supports four pod sizes: `x1`, `x2`, `x4`, and `x8`. Your index storage and compute capacity doubles for each size step. The default pod size is `x1`. You can increase the size of a pod after index creation.

To learn about changing the pod size of an index, see [Configure an index](/guides/indexes/pods/scale-pod-based-indexes#increase-pod-size).

## Pod environments

When creating a pod-based index, you must choose the cloud environment where you want the index to be hosted. The project environment can affect your [pricing](https://pinecone.io/pricing). The following table lists the available cloud regions and the corresponding values of the `environment` parameter for the [`create_index`](/guides/index-data/create-an-index#create-a-pod-based-index) endpoint:

| Cloud | Region                       | Environment                   |
| ----- | ---------------------------- | ----------------------------- |
| GCP   | us-west-1 (N. California)    | `us-west1-gcp`                |
| GCP   | us-central-1 (Iowa)          | `us-central1-gcp`             |
| GCP   | us-west-4 (Las Vegas)        | `us-west4-gcp`                |
| GCP   | us-east-4 (Virginia)         | `us-east4-gcp`                |
| GCP   | northamerica-northeast-1     | `northamerica-northeast1-gcp` |
| GCP   | asia-northeast-1 (Japan)     | `asia-northeast1-gcp`         |
| GCP   | asia-southeast-1 (Singapore) | `asia-southeast1-gcp`         |
| GCP   | us-east-1 (South Carolina)   | `us-east1-gcp`                |
| GCP   | eu-west-1 (Belgium)          | `eu-west1-gcp`                |
| GCP   | eu-west-4 (Netherlands)      | `eu-west4-gcp`                |
| AWS   | us-east-1 (Virginia)         | `us-east-1-aws`               |
| Azure | eastus (Virginia)            | `eastus-azure`                |

[Contact us](http://www.pinecone.io/contact/) if you need a dedicated deployment in other regions.

The environment cannot be changed after the index is created.

## Pod costs

For each pod-based index, billing is determined by the per-minute price per pod and the number of pods the index uses, regardless of index activity. The per-minute price varies by pod type, pod size, account plan, and cloud region. For the latest pod-based index pricing rates, see [Pricing](https://www.pinecone.io/pricing/pods).

Total cost depends on a combination of factors:

* **Pod type.** Each pod type has different per-minute pricing.
* **Number of pods.** This includes replicas, which duplicate pods.
* **Pod size.**  Larger pod sizes have proportionally higher costs per minute.
* **Total pod-minutes.** This includes the total time each pod is running, starting at pod creation and rounded up to 15-minute increments.
* **Cloud provider.** The cost per pod-type and pod-minute varies depending on the cloud provider you choose for your project.
* **Collection storage.** Collections incur costs per GB of data per minute in storage, rounded up to 15-minute increments.
* **Plan.** The free plan incurs no costs; the Standard or Enterprise plans incur different costs per pod-type, pod-minute, cloud provider, and collection storage.

The following equation calculates the total costs accrued over time:

```
(Number of pods) * (pod size) * (number of replicas) * (minutes pod exists) * (pod price per minute) 
+ (collection storage in GB) * (collection storage time in minutes) * (collection storage price per GB per minute)
```

To see a calculation of your current usage and costs, go to [**Settings > Usage**](https://app.pinecone.io/organizations/-/settings/usage) in the Pinecone console.

<Accordion title="Example">
  While our pricing page lists rates on an hourly basis for ease of comparison, this example lists prices per minute, as this is how Pinecone calculates billing.

  An example application has the following requirements:

  * 1,000,000 vectors with 1536 dimensions
  * 150 queries per second with `top_k` = 10
  * Deployment in an EU region
  * Ability to store 1GB of inactive vectors

  [Based on these requirements](/guides/indexes/pods/choose-a-pod-type-and-size), the organization chooses to configure the project to use the Standard billing plan to host one `p1.x2` pod with three replicas and a collection containing 1 GB of data. This project runs continuously for the month of January on the Standard plan. The components of the total cost for this example are given in Table 1 below:

  **Table 1: Example billing components**

  | Billing component             | Value        |
  | ----------------------------- | ------------ |
  | Number of pods                | 1            |
  | Number of replicas            | 3            |
  | Pod size                      | x2           |
  | Total pod count               | 6            |
  | Minutes in January            | 44,640       |
  | Pod-minutes (pods \* minutes) | 267,840      |
  | Pod price per minute          | \$0.0012     |
  | Collection storage            | 1 GB         |
  | Collection storage minutes    | 44,640       |
  | Price per storage minute      | \$0.00000056 |

  The invoice for this example is given in Table 2 below:

  **Table 2: Example invoice**

  | Product       | Quantity | Price per unit | Charge   |
  | ------------- | -------- | -------------- | -------- |
  | Collections   | 44,640   | \$0.00000056   | \$0.025  |
  | P2 Pods (AWS) | 0        |                | \$0.00   |
  | P2 Pods (GCP) | 0        |                | \$0.00   |
  | S1 Pods       | 0        |                | \$0.00   |
  | P1 Pods       | 267,840  | \$0.0012       | \$514.29 |

  Amount due \$514.54
</Accordion>

## Known limitations

* [Pod storage capacity](#pod-types)
  * Each **p1** pod has enough capacity for 1M vectors with 768 dimensions.
  * Each **s1** pod has enough capacity for 5M vectors with 768 dimensions.
* [Metadata](/guides/index-data/indexing-overview#metadata)
  * Metadata with high cardinality, such as a unique value for every vector in a large index, uses more memory than expected and can cause the pods to become full.
* [Collections](/guides/manage-data/back-up-an-index#pod-based-index-backups-using-collections)
  * You cannot query or write to a collection after its creation. For this reason, a collection only incurs storage costs.
  * You can only perform operations on collections in the current Pinecone project.
* [Sparse-dense vectors](/guides/search/hybrid-search#use-a-single-hybrid-index)
  * Only `s1` and `p1` [pod-based indexes](/guides/indexes/pods/understanding-pod-based-indexes#pod-types) using the dotproduct distance metric support sparse-dense vectors.

---

# Source: https://docs.pinecone.io/guides/projects/understanding-projects.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Understanding projects

> Learn about projects, environments, and member roles.

A Pinecone project belongs to an [organization](/guides/organizations/understanding-organizations) and contains a number of [indexes](/guides/index-data/indexing-overview) and users. Only a user who belongs to the project can access the indexes in that project. Each project also has at least one project owner.

## Project environments

You choose a cloud environment for each index in a project. This makes it easy to manage related resources across environments and use the same API key to access them.

## Project roles

If you are an [organization owner](/guides/organizations/understanding-organizations#organization-roles) or project owner, you can manage members in your project. Project members are assigned a role, which determines their permissions within the project. The project roles are as follows:

* **Project owner**: Project owners have global permissions across projects they own.

* **Project user**: Project users have restricted permissions for the specific projects they are invited to.

The following table summarizes the permissions for each project role:

| Permission                  | Owner | User |
| :-------------------------- | ----- | ---- |
| Update project names        | ✓     |      |
| Delete projects             | ✓     |      |
| View project members        | ✓     | ✓    |
| Update project member roles | ✓     |      |
| Delete project members      | ✓     |      |
| View API keys               | ✓     | ✓    |
| Create API keys             | ✓     |      |
| Delete API keys             | ✓     |      |
| View indexes                | ✓     | ✓    |
| Create indexes              | ✓     | ✓    |
| Delete indexes              | ✓     | ✓    |
| Upsert vectors              | ✓     | ✓    |
| Query vectors               | ✓     | ✓    |
| Fetch vectors               | ✓     | ✓    |
| Update a vector             | ✓     | ✓    |
| Delete a vector             | ✓     | ✓    |
| List vector IDs             | ✓     | ✓    |
| Get index stats             | ✓     | ✓    |

Specific to pod-based indexes:

<Warning>
  Customers who sign up for a Standard or Enterprise plan on or after August 18, 2025 cannot create pod-based indexes. Instead, create [serverless indexes](/guides/index-data/create-an-index), and consider using [dedicated read nodes](/guides/index-data/dedicated-read-nodes) for large workloads (millions of records or more, and moderate or high query rates).
</Warning>

| Permission                | Owner | User |
| :------------------------ | ----- | ---- |
| Update project pod limits | ✓     |      |
| View project pod limits   | ✓     | ✓    |
| Update index size         | ✓     | ✓    |

## API keys

Each Pinecone [project](/guides/projects/understanding-projects) has one or more API keys. In order to [make calls to the Pinecone API](/guides/get-started/quickstart), you must provide a valid API key for the relevant Pinecone project.

For more information, see [Manage API keys](/guides/projects/manage-api-keys).

## Project IDs

Each Pinecone project has a unique product ID.

To find the ID of a project, go to the project list in the [Pinecone console](https://app.pinecone.io/organizations/-/projects).

## See also

* [Understanding organizations](guides/organizations/understanding-organizations)
* [Manage organization members](guides/organizations/manage-organization-members)

---

# Source: https://docs.pinecone.io/integrations/unstructured.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Unstructured

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Unstructured builds ETL tools for LLMs, including an open source Python library, a SaaS API, and an ETL platform. Unstructured extracts content and metadata from 25+ document types, including PDFs, Word documents and PowerPoints. After extracting content and metadata, Unstructured performs additional preprocessing steps for LLMs such as chunking. Unstructured maintains upstream connections to data sources such as SharePoint and Google drive, and downstream connections to databases such as Pinecone.

Integrating Pinecone with Unstructured enables developers to load data from an source or document type into Pinecone with a single click, accelerating the building of LLM apps that connect to organizational data.

<PrimarySecondaryCTA primaryLabel={"Get started"} primaryHref={"https://docs.unstructured.io/ui/destinations/pinecone-destination-quickstart"} />

---

# Source: https://docs.pinecone.io/guides/manage-data/update-data.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Update records

> Update vectors and metadata for existing records

You can [update](/reference/api/latest/data-plane/update) a single record using the record ID or multiple records using a metadata filter.

* **Update by ID**: Update a single record's metadata (add or change fields) or vector values.
* **Update by metadata**: Update metadata (add or change fields) across multiple records using a metadata filter. Vector values cannot be updated.

To update entire records, use the [upsert](/guides/index-data/upsert-data) operation instead.

## Update by ID

To update the vector and/or metadata of a single record, use the [`update`](/reference/api/latest/data-plane/update) operation with the following parameters:

* `namespace`: The [namespace](/guides/index-data/indexing-overview#namespaces) containing the record to update. To use the default namespace, set the namespace to `"__default__"`.
* `id`: The ID of the record to update.
* One or both of the following:
  * Updated values for the vector. Specify one of the following:
    * `values`: For dense vectors. Must have the same length as the existing vector.
    * `sparse_values`: For sparse vectors.
  * `setMetadata`: The metadata to add or change. When updating metadata, only the specified metadata fields are modified, and if a specified metadata field does not exist, it is added.

<Warning>
  If a non-existent record ID is specified, no records are affected and a `200 OK` status is returned.
</Warning>

In this example, assume you are updating the dense vector values and one metadata value of the following record in the `example-namespace` namespace:

```
(
    namespace="example-namespace",
    id="id-3", 
    values=[4.0, 2.0], 
    setMetadata={"type": "doc", "genre": "drama"}
)
```

<CodeGroup>
  ```Python Python theme={null}
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  index.update(
  	namespace="example-namespace",
  	id="id-3", 
  	values=[5.0, 3.0], 
  	set_metadata={"genre": "comedy"}
  )
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  await index.namespace('example-namespace').update({
    id: 'id-3',
    values: [5.0, 3.0],
    metadata: {
      genre: "comedy",
    },
  });
  ```

  ```java Java theme={null}
  import com.google.protobuf.Struct;
  import com.google.protobuf.Value;
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import io.pinecone.proto.UpdateResponse;

  import java.util.Arrays;
  import java.util.List;

  public class UpdateExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");
          List<Float> values = Arrays.asList(5.0f, 3.0f);
  		Struct metaData = Struct.newBuilder()
  			.putFields("genre",
  					Value.newBuilder().setStringValue("comedy").build())
  			.build();
          UpdateResponse updateResponse = index.update("id-3", values, metaData, "example-namespace", null, null);
          System.out.println(updateResponse);
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
      "google.golang.org/protobuf/types/known/structpb"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
    	}

      id := "id-3"

      metadataMap := map[string]interface{}{
          "genre": "comedy",
      }

      metadataFilter, err := structpb.NewStruct(metadataMap)
      if err != nil {
          log.Fatalf("Failed to create metadata map: %v", err)
      }

      err = idxConnection.UpdateVector(ctx, &pinecone.UpdateVectorRequest{
          Id:       id,
          Values: []float32{5.0, 3.0},
          Metadata: metadataFilter,
      })
      if err != nil {
          log.Fatalf("Failed to update vector with ID %v: %v", id, err)
      }
  }
  ```

  ```csharp C# theme={null}
  using Pinecone;

  var pinecone = new PineconeClient("YOUR_API_KEY");

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  var index = pinecone.Index(host: "INDEX_HOST");

  var updateResponse = await index.UpdateAsync(new UpdateRequest {
      Id = "id-3",
      Namespace = "example-namespace",
      Values = new[] { 5.0f, 3.0f },
      SetMetadata = new Metadata {
          ["genre"] = new("comedy")
      }
  });
  ```

  ```bash curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  # Update both values and metadata
  curl "https://$INDEX_HOST/vectors/update" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "id": "id-3",
          "values": [5.0, 3.0],
          "setMetadata": {"genre": "comedy"},
          "namespace": "example-namespace"
        }'
  ```
</CodeGroup>

After the update, the dense vector values and the `genre` metadata value are changed, but the `type` metadata value is unchanged:

```
(
    id="id-3", 
    values=[5.0, 3.0], 
    metadata={"type": "doc", "genre": "comedy"}
)
```

## Update by metadata

To add or change metadata across multiple records in a namespace, use the `update` operation with the following parameters:

* `namespace`: The [namespace](/guides/index-data/indexing-overview#namespaces) containing the records to update. To use the default namespace, set this to `"__default__"`.
* `filter`: A [metadata filter expression](/guides/index-data/indexing-overview#metadata-filter-expressions) to match the records to update.
* `setMetadata`: The metadata to add or change. When updating metadata, only the specified metadata fields are modified. If a specified metadata field does not exist, it is added.
* `dry_run`: Optional. If `true`, the number of records that match the filter expression is returned, but the records are not updated.

  <Note>
    Each request updates a maximum of 100,000 records. Use `"dry_run": true` to check if you need to run the request multiple times. See the example below for details.
  </Note>

For example, let's say you have records that represent chunks of a single document with metadata that keeps track of chunk and document details, and you want to store the author's name with each chunk of the document:

```json  theme={null}
{
    "id": "document1#chunk1", 
    "values": [0.0236663818359375, -0.032989501953125, ..., -0.01041412353515625, 0.0086669921875], 
    "metadata": {
        "document_id": "document1",
        "document_title": "Introduction to Vector Databases",
        "chunk_number": 1,
        "chunk_text": "First chunk of the document content...",
        "document_url": "https://example.com/docs/document1"
    }
},
{
    "id": "document1#chunk2", 
    "values": [-0.0412445068359375, 0.028839111328125, ..., 0.01953125, -0.0174560546875],
    "metadata": {
        "document_id": "document1",
        "document_title": "Introduction to Vector Databases", 
        "chunk_number": 2,
        "chunk_text": "Second chunk of the document content...",
        "document_url": "https://example.com/docs/document1"
    }
},
...
```

1. To check how many records match the filter expression, send a request with `"dry_run": true`:

   ```bash curl {11} theme={null}
   # To get the unique host for an index,
   # see https://docs.pinecone.io/guides/manage-data/target-an-index
   PINECONE_API_KEY="YOUR_API_KEY"
   INDEX_HOST="INDEX_HOST"

   curl "https://$INDEX_HOST/vectors/update" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H 'Content-Type: application/json' \
       -H "X-Pinecone-Api-Version: 2025-10" \
       -d '{
               "dry_run": true,
               "namespace": "example-namespace",
               "filter": {
                   "document_title": {"$eq": "Introduction to Vector Databases"}
               },
               "setMetadata": {
                   "author": "Del Klein"
               } 
           }'
   ```

   The response contains the number of records that match the filter expression:

   ```json  theme={null}
   {
       "matchedVectors": 150000
   }
   ```

   Since this number exceeds the 100,000 record limit, you'll need to run the update request multiple times.

2. Initiate the first update by sending the request without the `dry_run` parameter:

   ```bash curl theme={null}
   curl "https://$INDEX_HOST/vectors/update" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H 'Content-Type: application/json' \
       -H "X-Pinecone-Api-Version: 2025-10" \
       -d '{
               "namespace": "example-namespace",
               "filter": {
                   "document_title": {"$eq": "Introduction to Vector Databases"}
               },
               "setMetadata": {
                   "author": "Del Klein"
               } 
           }'
   ```

   Again, the response contains the total number of records that match the filter expression, but only 100,000 will be updated:

   ```json  theme={null}
   {
       "matchedVectors": 150000
   }
   ```

3. Pinecone is eventually consistent, so there can be a slight delay before your update request is processed. Repeat the `dry_run` request until the number of matching records shows that the first 100,000 records have been updated:

   ```bash curl {11} theme={null}
   # To get the unique host for an index,
   # see https://docs.pinecone.io/guides/manage-data/target-an-index
   PINECONE_API_KEY="YOUR_API_KEY"
   INDEX_HOST="INDEX_HOST"

   curl "https://$INDEX_HOST/vectors/update" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H 'Content-Type: application/json' \
       -H "X-Pinecone-Api-Version: 2025-10" \
       -d '{
               "dry_run": true,
               "namespace": "example-namespace",
               "filter": {
                   "document_title": {"$eq": "Introduction to Vector Databases"}
               },
               "setMetadata": {
                   "author": "Del Klein"
               } 
           }'
   ```

   ```json  theme={null}
   {
       "matchedVectors": 50000
   }
   ```

4. Once the first 100,000 records have been updated, update the remaining records:

   ```bash curl theme={null}
   curl "https://$INDEX_HOST/vectors/update" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H 'Content-Type: application/json' \
       -H "X-Pinecone-Api-Version: 2025-10" \
       -d '{
               "namespace": "example-namespace",
               "filter": {
                   "document_title": {"$eq": "Introduction to Vector Databases"}
               },
               "setMetadata": {
                   "author": "Del Klein"
               } 
           }'
   ```

5. Repeat the `dry_run` request until the number of matching records shows that the remaining records have been updated:

   ```bash curl {11} theme={null}
   # To get the unique host for an index,
   # see https://docs.pinecone.io/guides/manage-data/target-an-index
   PINECONE_API_KEY="YOUR_API_KEY"
   INDEX_HOST="INDEX_HOST"

   curl "https://$INDEX_HOST/vectors/update" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -H 'Content-Type: application/json' \
       -H "X-Pinecone-Api-Version: 2025-10" \
       -d '{
               "dry_run": true,
               "namespace": "example-namespace",
               "filter": {
                   "document_title": {"$eq": "Introduction to Vector Databases"}
               },
               "setMetadata": {
                   "author": "Del Klein"
               } 
           }'
   ```

   ```json  theme={null}
   {
       "matchedVectors": 0
   }
   ```

   Once the request has completed, all matching records include the author name as metadata:

   ```json {10,22} theme={null}
   {
       "id": "document1#chunk1", 
       "values": [0.0236663818359375, -0.032989501953125, ..., -0.01041412353515625, 0.0086669921875], 
       "metadata": {
           "document_id": "document1",
           "document_title": "Introduction to Vector Databases",
           "chunk_number": 1,
           "chunk_text": "First chunk of the document content...",
           "document_url": "https://example.com/docs/document1",
           "author": "Del Klein"
       }
   },
   {
       "id": "document1#chunk2", 
       "values": [-0.0412445068359375, 0.028839111328125, ..., 0.01953125, -0.0174560546875],
       "metadata": {
           "document_id": "document1",
           "document_title": "Introduction to Vector Databases", 
           "chunk_number": 2,
           "chunk_text": "Second chunk of the document content...",
           "document_url": "https://example.com/docs/document1",
           "author": "Del Klein"
       }
   },
   ...
   ```

### Limitations

* Each request updates a maximum of 100,000 records. Use `"dry_run": true` to check if you need to run the request multiple times. See the example above for details.
* You can add or change metadata across multiple records, but you cannot remove metadata fields.

## Data freshness

Pinecone is eventually consistent, so there can be a slight delay before updates are visible to queries. You can [use log sequence numbers](/guides/index-data/check-data-freshness#check-the-log-sequence-number) to check whether an update request has completed.

## See also

* [Update an entire document](/guides/index-data/data-modeling#update-an-entire-document)

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/update.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Update a vector

> Update a vector in a namespace. If a value is included, it will overwrite the previous value. If a `set_metadata` is included, the values of the fields specified in it will be added or overwrite the previous value.

For guidance and examples, see [Update data](https://docs.pinecone.io/guides/manage-data/update-data).

<RequestExample>
  ```shell curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/vectors/update" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
          "id": "id-3",
          "values": [4.0, 2.0],
          "setMetadata": {"type": "comedy"},
          "namespace": "example-namespace"
        }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {}
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml post /vectors/update
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /vectors/update:
    post:
      tags:
        - Vector Operations
      summary: Update a vector
      description: >-
        Update a vector in a namespace. If a value is included, it will
        overwrite the previous value. If a `set_metadata` is included, the
        values of the fields specified in it will be added or overwrite the
        previous value.


        For guidance and examples, see [Update
        data](https://docs.pinecone.io/guides/manage-data/update-data).
      operationId: updateVector
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/UpdateRequest'
        required: true
      responses:
        '200':
          description: A successful response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/UpdateResponse'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    UpdateRequest:
      description: The request for the `update` operation.
      type: object
      properties:
        id:
          example: example-vector-1
          description: Vector's unique id.
          type: string
          minLength: 1
          maxLength: 512
        values:
          example:
            - 0.1
            - 0.2
            - 0.3
            - 0.4
            - 0.5
            - 0.6
            - 0.7
            - 0.8
          description: Vector data.
          type: array
          items:
            type: number
            format: float
          minLength: 1
          maxLength: 20000
        sparseValues:
          $ref: '#/components/schemas/SparseValues'
        setMetadata:
          example:
            genre: documentary
            year: 2019
          description: Metadata to set for the vector.
          type: object
        namespace:
          example: example-namespace
          description: The namespace containing the vector to update.
          type: string
        filter:
          example:
            genre:
              $in:
                - comedy
                - documentary
                - drama
            year:
              $eq: 2019
          description: >-
            A metadata filter expression. When updating metadata across records
            in a namespace,  the update is applied to all records that match the
            filter.  See [Understanding
            metadata](https://docs.pinecone.io/guides/index-data/indexing-overview#metadata).
          type: object
        dryRun:
          example: false
          description: >-
            If `true`, return the number of records that match the `filter`, but
            do not execute the update.  Default is `false`.
          default: false
          type: boolean
    UpdateResponse:
      description: The response for the `update` operation.
      type: object
      properties:
        matchedRecords:
          example: 42
          description: >-
            The number of records that matched the filter (if a filter was
            provided).
          type: integer
          format: int32
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    SparseValues:
      description: >-
        Vector sparse data. Represented as a list of indices and a list of 
        corresponded values, which must be with the same length.
      type: object
      properties:
        indices:
          example:
            - 1
            - 312
            - 822
            - 14
            - 980
          description: The indices of the sparse data.
          type: array
          required:
            - indices
          items:
            type: integer
            format: int64
          minLength: 1
          maxLength: 1000
        values:
          example:
            - 0.1
            - 0.2
            - 0.3
            - 0.4
            - 0.5
          description: >-
            The corresponding values of the sparse data, which must be with the
            same length as the indices.
          type: array
          required:
            - values
          items:
            type: number
            format: float
          minLength: 1
          maxLength: 1000
      required:
        - indices
        - values
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/admin/update_api_key.md

# Source: https://docs.pinecone.io/reference/api/2025-10/admin-assistant/update_api_key.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Update an API key

> Update the name and roles of an API key.


<RequestExample>
  ```bash curl theme={null}
  PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"
  PINECONE_API_KEY_ID="YOUR_API_KEY_ID"

  curl -X PATCH "https://api.pinecone.io/admin/api-keys/$PINECONE_API_KEY_ID" \
  	-H "X-Pinecone-Api-Version: 2025-10" \
  	-H "Authorization: Bearer $PINECONE_ACCESS_TOKEN" \
  	-d '{
  		"name": "New API key name",
  		"roles": ["ProjectEditor"]
  	}'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "key": {
      "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
      "name": "New API key name",
      "project_id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
      "roles": [
        "ProjectEditor"
      ]
    },
    "value": "string"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/admin_2025-10.oas.yaml patch /admin/api-keys/{api_key_id}
openapi: 3.0.3
info:
  title: Pinecone Admin API
  description: |
    Provides an API for managing a Pinecone organization and its resources.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - BearerAuth: []
tags:
  - name: API Keys
    description: Actions that manage API Keys.
  - name: Organizations
    description: Actions that manage organizations.
  - name: Projects
    description: Actions that manage projects.
paths:
  /admin/api-keys/{api_key_id}:
    patch:
      tags:
        - API Keys
      summary: Update an API key
      description: |
        Update the name and roles of an API key.
      operationId: update_api_key
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: api_key_id
          description: API key ID
          required: true
          schema:
            type: string
            format: uuid
          style: simple
      requestBody:
        description: Updated name and roles for the API key.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/UpdateAPIKeyRequest'
        required: true
      responses:
        '200':
          description: API key updated successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/APIKey'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-metric-validation-error:
                  summary: Validation error
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Bad request. The request body included invalid request
                        parameters.
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '403':
          description: Forbidden
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
        4XX:
          description: Unexpected error on request.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
components:
  schemas:
    UpdateAPIKeyRequest:
      type: object
      properties:
        name:
          example: devkey
          description: >
            A new name for the API key. The name must be 1-80 characters long.
            If omitted, the name will not be updated.
          type: string
          minLength: 1
          maxLength: 80
        roles:
          description: >
            A new set of roles for the API key. Existing roles will be removed
            if not included.

            If this field is omitted, the roles will not be updated.
          type: array
          items:
            example: ProjectEditor
            description: >-
              A role that can be assigned to an API key.

              Possible values: `ProjectEditor`, `ProjectViewer`,
              `ControlPlaneEditor`, `ControlPlaneViewer`, `DataPlaneEditor`, or
              `DataPlaneViewer`.
            x-enum:
              - ProjectEditor
              - ProjectViewer
              - ControlPlaneEditor
              - ControlPlaneViewer
              - DataPlaneEditor
              - DataPlaneViewer
            type: string
    APIKey:
      description: The details of an API key, without the secret.
      type: object
      properties:
        id:
          description: The unique ID of the API key.
          type: string
          format: uuid
        name:
          description: The name of the API key.
          type: string
        project_id:
          description: The ID of the project containing the API key.
          type: string
          format: uuid
        roles:
          description: The roles assigned to the API key.
          type: array
          items:
            example: ProjectEditor
            description: >-
              A role that can be assigned to an API key.

              Possible values: `ProjectEditor`, `ProjectViewer`,
              `ControlPlaneEditor`, `ControlPlaneViewer`, `DataPlaneEditor`, or
              `DataPlaneViewer`.
            x-enum:
              - ProjectEditor
              - ProjectViewer
              - ControlPlaneEditor
              - ControlPlaneViewer
              - DataPlaneEditor
              - DataPlaneViewer
            type: string
      required:
        - id
        - name
        - project_id
        - roles
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, or `UNPROCESSABLE_ENTITY`.        
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    BearerAuth:
      type: http
      scheme: bearer
      description: >
        An [access
        token](https://docs.pinecone.io/guides/organizations/manage-service-accounts#retrieve-an-access-token)
        must be provided in the `Authorization` header using the `Bearer`
        scheme.

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/assistant/update_assistant.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Update an assistant

> Update an existing assistant. You can modify the assistant's instructions.

For guidance and examples, see [Manage assistants](https://docs.pinecone.io/guides/assistant/manage-assistants#add-instructions-to-an-assistant).

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"

  curl -X PATCH "https://api.pinecone.io/assistant/assistants/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
    "instructions": "Use American English for spelling and grammar.",
    "metadata": {"team": "customer-support", "version": "1.1"}
  }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "name": "example-assistant",
    "instructions": "Use American English for spelling and grammar.",
    "metadata": {"team": "customer-support", "version": "1.1"},
    "status": "Ready",
    "host": "https://prod-1-data.ke.pinecone.io",
    "created_at": "2025-10-01T12:30:00Z",
    "updated_at": "2025-10-01T12:45:00Z"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/assistant_control_2025-10.oas.yaml patch /assistants/{assistant_name}
openapi: 3.0.3
info:
  title: Pinecone Assistant Control Plane API
  description: >-
    Pinecone Assistant Engine is a context engine to store and retrieve relevant
    knowledge  from millions of documents at scale. This API supports creating
    and managing assistants. 
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io/assistant
    description: Production API endpoints
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Assistants
    description: Actions that manage Assistants
paths:
  /assistants/{assistant_name}:
    patch:
      tags:
        - Manage Assistants
      summary: Update an assistant
      description: >-
        Update an existing assistant. You can modify the assistant's
        instructions.


        For guidance and examples, see [Manage
        assistants](https://docs.pinecone.io/guides/assistant/manage-assistants#add-instructions-to-an-assistant).
      operationId: update_assistant
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: assistant_name
          description: The name of the assistant to update.
          required: true
          schema:
            type: string
          style: simple
      requestBody:
        description: The desired configuration updates for the assistant.
        content:
          application/json:
            schema:
              description: The configuration updates for the assistant.
              type: object
              properties:
                instructions:
                  nullable: true
                  description: >-
                    Description or directive for the assistant to apply to all
                    responses. Maximum 16 KB.
                  type: string
                metadata:
                  description: >-
                    Optional metadata associated with the assistant. Metadata is
                    a JSON object that can store custom organizational data,
                    tags, and attributes. Maximum size is 16KB.
                  nullable: true
                  type: object
        required: true
      responses:
        '200':
          description: Update request successful.
          content:
            application/json:
              schema:
                type: object
                properties:
                  assistant_name:
                    type: string
                  instructions:
                    description: >-
                      Description or directive for the assistant to apply to all
                      responses.
                    type: string
                  metadata:
                    description: >-
                      Optional metadata associated with the assistant. Metadata
                      is a JSON object that can store custom organizational
                      data, tags, and attributes.
                    type: object
        '400':
          description: Bad Request
        '401':
          description: Unauthorized
        '404':
          description: Assistant not found
        '500':
          description: Internal Server Error
components:
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: Pinecone API Key

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/admin/update_project.md

# Source: https://docs.pinecone.io/reference/api/2025-10/admin-assistant/update_project.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Update a project

> Update a project's configuration details.
You can update the project's name, maximum number of Pods, or enable encryption with a customer-managed encryption key (CMEK).


<RequestExample>
  ```bash curl theme={null}
  PINECONE_ACCESS_TOKEN="YOUR_ACCESS_TOKEN"
  PROJECT_ID="YOUR_PROJECT_ID"

  curl -X PATCH "https://api.pinecone.io/admin/projects/$PROJECT_ID" \
    -H "accept: application/json" \
    -H "Content-Type: application/json" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "name": "updated-example-project"
      }'
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
    "name": "updated-example-project",
    "max_pods": 0,
    "force_encryption_with_cmek": false,
    "organization_id": "string",
    "created_at": "2025-03-17T00:42:31.912Z"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/admin_2025-10.oas.yaml patch /admin/projects/{project_id}
openapi: 3.0.3
info:
  title: Pinecone Admin API
  description: |
    Provides an API for managing a Pinecone organization and its resources.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://api.pinecone.io
    description: Production API endpoints
security:
  - BearerAuth: []
tags:
  - name: API Keys
    description: Actions that manage API Keys.
  - name: Organizations
    description: Actions that manage organizations.
  - name: Projects
    description: Actions that manage projects.
paths:
  /admin/projects/{project_id}:
    patch:
      tags:
        - Projects
      summary: Update a project
      description: >
        Update a project's configuration details.

        You can update the project's name, maximum number of Pods, or enable
        encryption with a customer-managed encryption key (CMEK).
      operationId: update_project
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: project_id
          description: Project ID
          required: true
          schema:
            type: string
            format: uuid
          style: simple
      requestBody:
        description: >
          Project details to be updated. Fields that are omitted will not be
          updated.
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/UpdateProjectRequest'
        required: true
      responses:
        '200':
          description: The project was successfully updated.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Project'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                index-metric-validation-error:
                  summary: Validation error
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Bad request. The request body included invalid request
                        parameters.
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '403':
          description: Forbidden
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '404':
          description: Not found
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
        4XX:
          description: Unexpected error on request.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
components:
  schemas:
    UpdateProjectRequest:
      type: object
      properties:
        name:
          example: chatbot-prod
          description: The name of the new project.
          type: string
          minLength: 1
          maxLength: 512
        max_pods:
          description: |
            The maximum number of Pods that can be created in the project.
          type: integer
        force_encryption_with_cmek:
          description: >
            Whether to force encryption with a customer-managed encryption key
            (CMEK). Once enabled, CMEK encryption cannot be disabled.
          type: boolean
    Project:
      description: The details of a project.
      type: object
      properties:
        id:
          description: The unique ID of the project.
          type: string
          format: uuid
        name:
          description: The name of the project.
          type: string
          minLength: 1
          maxLength: 512
        max_pods:
          description: The maximum number of Pods that can be created in the project.
          type: integer
        force_encryption_with_cmek:
          description: >-
            Whether to force encryption with a customer-managed encryption key
            (CMEK).
          type: boolean
        organization_id:
          description: The unique ID of the organization that the project belongs to.
          type: string
        created_at:
          description: The date and time when the project was created.
          type: string
          format: date-time
      required:
        - id
        - name
        - max_pods
        - force_encryption_with_cmek
        - organization_id
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: >-
              Index name must contain only lowercase alphanumeric characters or
              hyphens, and must not begin or end with a hyphen.
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              description: >-
                The error code.

                Possible values: `OK`, `UNKNOWN`, `INVALID_ARGUMENT`,
                `DEADLINE_EXCEEDED`, `QUOTA_EXCEEDED`, `NOT_FOUND`,
                `ALREADY_EXISTS`, `PERMISSION_DENIED`, `UNAUTHENTICATED`,
                `RESOURCE_EXHAUSTED`, `FAILED_PRECONDITION`, `ABORTED`,
                `OUT_OF_RANGE`, `UNIMPLEMENTED`, `INTERNAL`, `UNAVAILABLE`,
                `DATA_LOSS`, `FORBIDDEN`, or `UNPROCESSABLE_ENTITY`.        
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
                - UNPROCESSABLE_ENTITY
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    BearerAuth:
      type: http
      scheme: bearer
      description: >
        An [access
        token](https://docs.pinecone.io/guides/organizations/manage-service-accounts#retrieve-an-access-token)
        must be provided in the `Authorization` header using the `Bearer`
        scheme.

````

---

# Source: https://docs.pinecone.io/guides/organizations/manage-billing/upgrade-billing-plan.md

# Source: https://docs.pinecone.io/guides/assistant/admin/upgrade-billing-plan.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Upgrade your plan

> Upgrade to a paid plan to access advanced features and limits.

This page describes how to upgrade from the free Starter plan to the [Standard or Enterprise plan](https://www.pinecone.io/pricing/), paying either with a credit/debit card or through a supported cloud marketplace.

<Note>
  To change your plan, you must be an [organization owner or billing admin](/guides/organizations/understanding-organizations#organization-roles).
</Note>

<Tip>
  To commit to annual spending, [contact Pinecone](https://www.pinecone.io/contact).
</Tip>

## Pay with a credit/debit card

To upgrade your plan to Standard or Enterprise and pay with a credit/debit card, do the following:

1. In the Pinecone console, go to [Settings > Billing > Plans](https://app.pinecone.io/organizations/-/settings/billing/plans).
2. Click **Upgrade** in the **Standard** or **Enterprise** plan section.
3. Click **Credit / Debit card**.
4. Enter your credit card information.
5. Click **Upgrade**.

After upgrading, you will immediately start paying for usage of your Pinecone indexes, including the serverless indexes that were free on the Starter plan. For more details about how costs are calculated, see [Understanding cost](/guides/manage-cost/understanding-cost).

## Pay through the Google Cloud Marketplace

To upgrade your plan to Standard or Enterprise and pay through the Google Cloud Marketplace, do the following:

1. In the Pinecone console, go to [Settings > Billing > Plans](https://app.pinecone.io/organizations/-/settings/billing/plans).
2. Click **Upgrade** in the **Standard** or **Enterprise** plan section.
3. Click **Billing through GCP**. This takes you to the [Pinecone listing](https://console.cloud.google.com/marketplace/product/pinecone-public/pinecone) in the Google Cloud Marketplace.
4. Click **Subscribe**.
5. On the **Order Summary** page, select a billing account, accept the terms and conditions, and click **Subscribe**.

   <Note>The billing unit listed does not reflect the actual cost or metering of costs for Pinecone. See the [Pinecone Pricing page](https://www.pinecone.io/pricing/) for accurate details.</Note>
6. On the **Your order request has been sent to Pinecone** modal, click **Sign up with Pinecone**. This takes you to a Google-specific Pinecone login page.
7. Log in to your Pinecone account. Use the same authentication method as your existing Pinecone organization.
8. Select an organization from the list. You can only connect to organizations that are on the [Starter plan](https://www.pinecone.io/pricing/). Alternatively, you can opt to create a new organization.
9. Click **Connect to Pinecone** and follow the prompts.

Once your organization is connected and upgraded, you will receive a confirmation message. You will then immediately start paying for usage of your Pinecone indexes, including the serverless indexes that were free on the Starter plan. For more details about how costs are calculated, see [Understanding cost](/guides/manage-cost/understanding-cost).

## Pay through the AWS Marketplace

To upgrade your plan to Standard or Enterprise and pay through the AWS Marketplace, do the following:

1. In the Pinecone console, go to [Settings > Billing > Plans](https://app.pinecone.io/organizations/-/settings/billing/plans).
2. Click **Upgrade** in the **Standard** or **Enterprise** plan section.
3. Click **Billing through AWS**. This takes you to the [Pinecone listing](https://aws.amazon.com/marketplace/pp/prodview-xhgyscinlz4jk) in the AWS Marketplace.
4. Click **View purchase options**.
5. On the **Subscribe to Pinecone Vector Database** page, review the offer and then click **Subscribe**.

   <Note>The billing unit listed does not reflect the actual cost or metering of costs for Pinecone. See the [Pinecone Pricing page](https://www.pinecone.io/pricing/) for accurate details.</Note>
6. You'll see a message stating that your subscription is in process. Click **Set up your account**. This takes you to an AWS-specific Pinecone login page.

   <Warning>
     If the [Pinecone subscription page](https://aws.amazon.com/marketplace/saas/ordering?productId=738798c3-eeca-494a-a2a9-161bee9450b2) shows a message stating, “You are currently subscribed to this offer,” contact your team members to request an invitation to the existing AWS-linked organization. The **Set up your account** button is clickable, but Pinecone does not create a new AWS-linked organization.
   </Warning>
7. Log in to your Pinecone account. Use the same authentication method as your existing Pinecone organization.
8. Select an organization from the list. You can only connect to organizations that are on the [Starter plan](https://www.pinecone.io/pricing/). Alternatively, you can opt to create a new organization.
9. Click **Connect to Pinecone** and follow the prompts.

Once your organization is connected and upgraded, you will receive a confirmation message. You will then immediately start paying for usage of your Pinecone indexes, including the serverless indexes that were free on the Starter plan. For more details about how costs are calculated, see [Understanding cost](/guides/manage-cost/understanding-cost).

## Pay through the Microsoft Marketplace

To upgrade your plan to Standard or Enterprise and pay through the Microsoft Marketplace, do the following:

1. In the Pinecone console, go to [Settings > Billing > Plans](https://app.pinecone.io/organizations/-/settings/billing/plans).
2. Click **Upgrade** in the **Standard** or **Enterprise** plan section.
3. Click **Billing through Azure**. This takes you to the [Pinecone listing](https://marketplace.microsoft.com/product/saas/pineconesystemsinc1688761585469.pineconesaas) in the Microsoft Marketplace.
4. Click **Get it now**.
5. Select the **Pinecone - Pay As You Go** plan.
6. Click **Subscribe**.
7. On the **Subscribe to Pinecone** page, select the required details and click **Review + subscribe**.

   <Note>The billing unit listed does not reflect the actual cost or metering of costs for Pinecone. See the [Pinecone Pricing page](https://www.pinecone.io/pricing/) for accurate details.</Note>
8. Click **Subscribe**.
9. After the subscription is approved, click **Configure account now**. This redirects you to an Microsoft-specific Pinecone login page.
10. Log in to your Pinecone account. Use the same authentication method as your existing Pinecone organization.
11. Select an organization from the list. You can only connect to organizations that are on the [Starter plan](https://www.pinecone.io/pricing/). Alternatively, you can opt to create a new organization.
12. Click **Connect to Pinecone** and follow the prompts.

Once your organization is connected and upgraded, you will receive a confirmation message. You will then immediately start paying for usage of your Pinecone indexes, including the serverless indexes that were free on the Starter plan. For more details about how costs are calculated, see [Understanding cost](/guides/manage-cost/understanding-cost).

---

# Source: https://docs.pinecone.io/guides/assistant/upload-files.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Upload files

> Upload local files to an assistant.

<Note>
  File upload limitations depend on the plan you are using. For more information, see [Pricing and limitations](/guides/assistant/pricing-and-limits#limits).
</Note>

## Upload a local file

You can [upload a file to your assistant](/reference/api/latest/assistant/upload_file) from your local device, as in the following example:

<CodeGroup>
  ```python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  pc = Pinecone(api_key="YOUR_API_KEY")

  # Get an assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant", 
  )

  # Upload a file.
  response = assistant.upload_file(
      file_path="/Users/jdoe/Downloads/example_file.txt",
      timeout=None
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  await assistant.uploadFile({
    path: '/Users/jdoe/Downloads/example_file.txt'
  });
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"
  LOCAL_FILE_PATH="/Users/jdoe/Downloads/example_file.txt"

  curl -X POST "https://prod-1-data.ke.pinecone.io/assistant/files/$ASSISTANT_NAME" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -F "file=@$LOCAL_FILE_PATH"
  ```
</CodeGroup>

It may take several minutes for your assistant to process your file. You can [check the status of your file](/guides/assistant/manage-files#get-the-status-of-a-file) to determine if it is ready to use.

<Tip>
  You can upload a file to an assistant using the [Pinecone console](https://app.pinecone.io/organizations/-/projects/-/assistant). Select the assistant you want to upload to and add the file in the Assistant playground.
</Tip>

## Upload a file with metadata

You can upload a file with metadata. Metadata is a dictionary of key-value pairs that you can use to store additional information about the file. For example, you can use metadata to store the file's name, document type, publish date, or any other relevant information.

<CodeGroup>
  ```Python Python theme={null}
  # To use the Python SDK, install the plugin:
  # pip install --upgrade pinecone pinecone-plugin-assistant

  from pinecone import Pinecone
  pc = Pinecone(api_key="YOUR_API_KEY")

  # Get the assistant.
  assistant = pc.assistant.Assistant(
      assistant_name="example-assistant", 
  )

  # Upload a file.
  response = assistant.upload_file(
      file_path="/Users/jdoe/Downloads/example_file.txt",
      metadata={"published": "2024-01-01", "document_type": "manuscript"},
      timeout=None
  )
  ```

  ```javascript JavaScript theme={null}
  import { Pinecone } from '@pinecone-database/pinecone'

  const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

  const assistantName = 'example-assistant';
  const assistant = pc.Assistant(assistantName);
  await assistant.uploadFile({
    path: '/Users/jdoe/Downloads/example_file.txt',
    metadata: { 'published': '2024-01-01', 'document_type': 'manuscript' },
  });
  ```

  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"
  ENCODED_METADATA="%7B%22published%22%3A%222024-01-01%22%2C%22document_type%22%3A%22script%22%7D" # URL encoded metadata - See w3schools.com/tags/ref_urlencode.ASP
  LOCAL_FILE_PATH="/Users/jdoe/Downloads/example_file.txt"

  curl -X POST "https://prod-1-data.ke.pinecone.io/assistant/files/$ASSISTANT_NAME?metadata=$ENCODED_METADATA" \
      -H "Api-Key: $PINECONE_API_KEY" \
      -F "file=@$LOCAL_FILE_PATH"
  ```
</CodeGroup>

When a file is uploaded with metadata, you can use the metadata to [filter a list of files](/guides/assistant/manage-files#view-a-filtered-list-of-files) and [filter chat responses](/guides/assistant/chat-with-assistant#filter-chat-with-metadata).

## Upload a PDF with multimodal context

Assistants can gather context from images contained in PDF files. To learn more about this feature, see [Multimodal context for assistants](/guides/assistant/multimodal).

## Upload from a binary stream

You can upload a file directly from an in-memory binary stream using the Python SDK and the [BytesIO class](https://docs.python.org/3/library/io.html#io.BytesIO).

<Note>
  When uploading text-based files (like .txt, .md, .json, etc.) through BytesIO streams, make sure the content is encoded in UTF-8 format.
</Note>

```python Python theme={null}
from pinecone import Pinecone
from io import BytesIO

pc = Pinecone(api_key="YOUR_API_KEY")

# Get an assistant
assistant = pc.assistant.Assistant(
    assistant_name="example-assistant", 
)

# Create a BytesIO stream with some content
md_text = "# Title\n\ntext"
# Note: Assistant currently supports only utf-8 for text-based files
stream = BytesIO(md_text.encode("utf-8"))

# Upload the stream
response = assistant.upload_bytes_stream(
    stream=stream,
    file_name="example_file.md",
    timeout=None
)
```

---

# Source: https://docs.pinecone.io/reference/api/2025-10/assistant/upload_file.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Upload file to assistant

> Upload a file to the specified assistant.

For guidance and examples, see [Manage files](https://docs.pinecone.io/guides/assistant/manage-files#upload-a-local-file).

<RequestExample>
  ```bash curl theme={null}
  PINECONE_API_KEY="YOUR_API_KEY"
  ASSISTANT_NAME="example-assistant"
  ENCODED_METADATA="%7B%22published%22%3A%222024-01-01%22%2C%22document_type%22%3A%22script%22%7D" # URL encoded metadata - See w3schools.com/tags/ref_urlencode.ASP
  LOCAL_FILE_PATH="/Users/jdoe/Downloads/example_file.txt"

  curl -X POST "https://prod-1-data.ke.pinecone.io/assistant/files/$ASSISTANT_NAME?metadata=$ENCODED_METADATA" \
       -H "Api-Key: $PINECONE_API_KEY" \
       -F "file=@$LOCAL_FILE_PATH"
  ```
</RequestExample>

<ResponseExample>
  ```json curl theme={null}
  {
    "name": "example-file.txt",
    "id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
    "metadata": "{ 'published': '2024-01-01', 'document_type': 'manuscript' }",
    "created_on": "2023-11-07T05:31:56Z",
    "updated_on": "2023-11-07T05:31:56Z",
    "status": "Processing",
    "percent_done": 50,
    "signed_url": "https://storage.googleapis.com/bucket/file.pdf",
    "error_message": "<string>"
  }
  ```
</ResponseExample>


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/assistant_data_2025-10.oas.yaml post /files/{assistant_name}
openapi: 3.0.3
info:
  title: Pinecone Assistant Data Plane API
  description: >-
    Pinecone Assistant Engine is a context engine to store and retrieve relevant
    knowledge from millions of documents at scale. This API supports
    interactions with assistants.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{assistant_host}
    variables:
      assistant_host:
        default: unknown
        description: host of the created assistant
security:
  - ApiKeyAuth: []
tags:
  - name: Manage Assistants
    description: Actions that manage Assistants
paths:
  /files/{assistant_name}:
    post:
      tags:
        - Manage Assistants
      summary: Upload file to assistant
      description: >-
        Upload a file to the specified assistant.


        For guidance and examples, see [Manage
        files](https://docs.pinecone.io/guides/assistant/manage-files#upload-a-local-file).
      operationId: upload_file
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: assistant_name
          description: The name of the assistant to upload files to.
          required: true
          schema:
            type: string
          example: test-model
          style: simple
        - in: query
          name: metadata
          description: Optional JSON-encoded metadata for files.
          schema:
            type: string
            format: json
          example: '{"genre":{"$eq":"comedy"}}'
          style: form
        - in: query
          name: multimodal
          description: >-
            Optional flag to opt in to multimodal file processing (PDFs only).
            Can be either `true` or `false`. Default is `false`.
          schema:
            type: string
          style: form
      requestBody:
        description: The desired file to be uploaded and processed into the assistant.
        content:
          multipart/form-data:
            schema:
              type: object
              properties:
                file:
                  description: The file to upload.
                  type: string
                  format: binary
              required:
                - file
        required: true
      responses:
        '200':
          description: File upload has been accepted.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/AssistantFileModel'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                files-validation-error:
                  summary: Validation error on ingest.
                  value:
                    error:
                      code: INVALID_ARGUMENT
                      message: >-
                        Uploaded file can only currently be either a pdf or txt
                        file
                    status: 400
        '401':
          description: 'Unauthorized. Possible causes: Invalid API key.'
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                unauthorized:
                  summary: Unauthorized
                  value:
                    error:
                      code: UNAUTHENTICATED
                      message: Invalid API key.
                    status: 401
        '404':
          description: Assistant not found.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                assistant-not-found:
                  summary: Assistant not found.
                  value:
                    error:
                      code: NOT_FOUND
                      message: Assistant "example-assistant" not found.
                    status: 404
        '500':
          description: Internal server error.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              examples:
                internal-server-error:
                  summary: Internal server error
                  value:
                    error:
                      code: UNKNOWN
                      message: Internal server error
                    status: 500
components:
  schemas:
    AssistantFileModel:
      description: The response format for a successful file upload request.
      type: object
      properties:
        name:
          type: string
        id:
          type: string
          format: uuid
        metadata:
          nullable: true
          type: object
        created_on:
          type: string
          format: date-time
        updated_on:
          type: string
          format: date-time
        status:
          description: >-
            The current state of the uploaded file. Possible values:

            - `Processing`: File is being processed (parsed, chunked, embedded)

            - `Available`: Processing completed successfully; file is ready for
            use

            - `Deleting`: Deletion has been initiated but not yet completed

            - `ProcessingFailed`: Processing failed with an error


            Note: Once a file is deleted, the API returns 404 Not Found instead
            of a file object.
          x-enum:
            - Processing
            - Available
            - Deleting
            - ProcessingFailed
          type: string
        percent_done:
          nullable: true
          description: The percentage of the file that has been processed
          type: number
          format: double
        signed_url:
          nullable: true
          example: https://storage.googleapis.com/bucket/file.pdf?...
          description: >-
            A [signed
            URL](https://cloud.google.com/storage/docs/access-control/signed-urls)
            that provides temporary, read-only access to the underlying file.
            Anyone with the link can access the file, so treat it as sensitive
            data. Expires after a short time.
          type: string
        error_message:
          nullable: true
          description: >-
            A message describing any error during file processing. Provided only
            if an error occurs.
          type: string
        multimodal:
          description: Indicates whether the file was processed as multimodal.
          type: boolean
      required:
        - id
        - name
    ErrorResponse:
      example:
        error:
          code: QUOTA_EXCEEDED
          message: >-
            The index exceeds the project quota of 5 pods by 2 pods. Upgrade
            your account or change the project settings to increase the quota.
        status: 429
      description: The response shape used for all error responses.
      type: object
      properties:
        status:
          example: 500
          description: The HTTP status code of the error.
          type: integer
        error:
          example:
            code: INVALID_ARGUMENT
            message: Uploaded file can only currently be either a pdf or txt file
          description: Detailed information about the error that occurred.
          type: object
          properties:
            code:
              x-enum:
                - OK
                - UNKNOWN
                - INVALID_ARGUMENT
                - DEADLINE_EXCEEDED
                - QUOTA_EXCEEDED
                - NOT_FOUND
                - ALREADY_EXISTS
                - PERMISSION_DENIED
                - UNAUTHENTICATED
                - RESOURCE_EXHAUSTED
                - FAILED_PRECONDITION
                - ABORTED
                - OUT_OF_RANGE
                - UNIMPLEMENTED
                - INTERNAL
                - UNAVAILABLE
                - DATA_LOSS
                - FORBIDDEN
              type: string
            message:
              example: >-
                Index name must contain only lowercase alphanumeric characters
                or hyphens, and must not begin or end with a hyphen.
              type: string
            details:
              description: >-
                Additional information about the error. This field is not
                guaranteed to be present.
              type: object
          required:
            - code
            - message
      required:
        - status
        - error
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: Pinecone API Key

````

---

# Source: https://docs.pinecone.io/guides/index-data/upsert-data.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Upsert records

> Add or update records in Pinecone indexes and manage data with namespaces.

This page shows you how to upsert records into a namespace in an index. [Namespaces](/guides/index-data/indexing-overview#namespaces) let you partition records within an index and are essential for [implementing multitenancy](/guides/index-data/implement-multitenancy) when you need to isolate the data of each customer/user.

If a record ID already exists, upserting overwrites the entire record. To change only part of a record, [update ](/guides/manage-data/update-data) the record.

<Tip>
  To control costs when ingesting large datasets (10,000,000+ records), use [import](/guides/index-data/import-data) instead of upsert.
</Tip>

## Upsert dense vectors

<Tabs>
  <Tab title="Upsert text">
    <Note>
      Upserting text is supported only for [indexes with integrated embedding](/guides/index-data/indexing-overview#integrated-embedding).
    </Note>

    To upsert source text into a [dense index with integrated embedding](/guides/index-data/create-an-index#create-a-dense-index), use the [`upsert_records`](/reference/api/latest/data-plane/upsert_records) operation. Pinecone converts the text to dense vectors automatically using the hosted dense embedding model associated with the index.

    * Specify the [`namespace`](/guides/index-data/indexing-overview#namespaces) to upsert into. If the namespace doesn't exist, it is created. To use the default namespace, set the namespace to `"__default__"`.
    * Format your input data as records, each with the following:
      * An `_id` field with a unique record identifier for the index namespace. `id` can be used as an alias for `_id`.
      * A field with the source text to convert to a vector. This field must match the `field_map` specified in the index.
      * Additional fields are stored as record [metadata](/guides/index-data/indexing-overview#metadata) and can be returned in search results or used to [filter search results](/guides/search/filter-by-metadata).

    For example, the following code converts the sentences in the `chunk_text` fields to dense vectors and then upserts them into `example-namespace` in an example index. The additional `category` field is stored as metadata.

    <CodeGroup>
      ```python Python theme={null}
      from pinecone import Pinecone

      pc = Pinecone(api_key="YOUR_API_KEY")

      # To get the unique host for an index, 
      # see https://docs.pinecone.io/guides/manage-data/target-an-index
      index = pc.Index(host="INDEX_HOST")

      # Upsert records into a namespace
      # `chunk_text` fields are converted to dense vectors
      # `category` fields are stored as metadata
      index.upsert_records(
          "example-namespace",
          [
              {
                  "_id": "rec1",
                  "chunk_text": "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut.",
                  "category": "digestive system", 
              },
              {
                  "_id": "rec2",
                  "chunk_text": "Apples originated in Central Asia and have been cultivated for thousands of years, with over 7,500 varieties available today.",
                  "category": "cultivation",
              },
              {
                  "_id": "rec3",
                  "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases.",
                  "category": "immune system",
              },
              {
                  "_id": "rec4",
                  "chunk_text": "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes.",
                  "category": "endocrine system",
              },
          ]
      ) 
      ```

      ```javascript JavaScript theme={null}
      import { Pinecone } from '@pinecone-database/pinecone'

      const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      const namespace = pc.index("INDEX_NAME", "INDEX_HOST").namespace("example-namespace");

      // Upsert records into a namespace
      // `chunk_text` fields are converted to dense vectors
      // `category` is stored as metadata
      await namespace.upsertRecords([
              {
                  "_id": "rec1",
                  "chunk_text": "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut.",
                  "category": "digestive system", 
              },
              {
                  "_id": "rec2",
                  "chunk_text": "Apples originated in Central Asia and have been cultivated for thousands of years, with over 7,500 varieties available today.",
                  "category": "cultivation",
              },
              {
                  "_id": "rec3",
                  "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases.",
                  "category": "immune system",
              },
              {
                  "_id": "rec4",
                  "chunk_text": "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes.",
                  "category": "endocrine system",
              }
      ]);
      ```

      ```java Java theme={null}
      import io.pinecone.clients.Index;
      import io.pinecone.configs.PineconeConfig;
      import io.pinecone.configs.PineconeConnection;
      import org.openapitools.db_data.client.ApiException;

      import java.util.*;

      public class UpsertText {
          public static void main(String[] args) throws ApiException {
              PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
              config.setHost("INDEX_HOST");
              PineconeConnection connection = new PineconeConnection(config);

              Index index = new Index(config, connection, "integrated-dense-java");
              ArrayList<Map<String, String>> upsertRecords = new ArrayList<>();

              HashMap<String, String> record1 = new HashMap<>();
              record1.put("_id", "rec1");
              record1.put("category", "digestive system");
              record1.put("chunk_text", "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut.");

              HashMap<String, String> record2 = new HashMap<>();
              record2.put("_id", "rec2");
              record2.put("category", "cultivation");
              record2.put("chunk_text", "Apples originated in Central Asia and have been cultivated for thousands of years, with over 7,500 varieties available today.");

              HashMap<String, String> record3 = new HashMap<>();
              record3.put("_id", "rec3");
              record3.put("category", "immune system");
              record3.put("chunk_text", "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases.");

              HashMap<String, String> record4 = new HashMap<>();
              record4.put("_id", "rec4");
              record4.put("category", "endocrine system");
              record4.put("chunk_text", "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes.");

              upsertRecords.add(record1);
              upsertRecords.add(record2);
              upsertRecords.add(record3);
              upsertRecords.add(record4);

              index.upsertRecords("example-namespace", upsertRecords);
          }
      }
      ```

      ```go Go theme={null}
      package main

      import (
          "context"
          "fmt"
          "log"

          "github.com/pinecone-io/go-pinecone/v4/pinecone"
      )

      func main() {
          ctx := context.Background()

          pc, err := pinecone.NewClient(pinecone.NewClientParams{
              ApiKey: "YOUR_API_KEY",
          })
          if err != nil {
              log.Fatalf("Failed to create Client: %v", err)
          }

          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
          if err != nil {
              log.Fatalf("Failed to create IndexConnection for Host: %v", err)
      	  }

          // Upsert records into a namespace
          // `chunk_text` fields are converted to dense vectors
          // `category` is stored as metadata
      	records := []*pinecone.IntegratedRecord{
              {
                  "_id": "rec1",
                  "chunk_text": "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut.",
                  "category": "digestive system", 
              },
              {
                  "_id": "rec2",
                  "chunk_text": "Apples originated in Central Asia and have been cultivated for thousands of years, with over 7,500 varieties available today.",
                  "category": "cultivation",
              },
              {
                  "_id": "rec3",
                  "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases.",
                  "category": "immune system",
              },
              {
                  "_id": "rec4",
                  "chunk_text": "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes.",
                  "category": "endocrine system",
              },
      	}

      	err = idxConnection.UpsertRecords(ctx, records)
      	if err != nil {
      		log.Fatalf("Failed to upsert vectors: %v", err)
      	}
      }
      ```

      ```csharp C# theme={null}
      using Pinecone;

      var pinecone = new PineconeClient("YOUR_API_KEY");

      var index = pinecone.Index(host: "INDEX_HOST");

      await index.UpsertRecordsAsync(
          "example-namespace",
          [
              new UpsertRecord
              {
                  Id = "rec1",
                  AdditionalProperties =
                  {
                      ["chunk_text"] = "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
                      ["category"] = "technology",
                      ["quarter"] = "Q3",
                  },
              },
              new UpsertRecord
              {
                  Id = "rec2",
                  AdditionalProperties =
                  {
                      ["chunk_text"] = "AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space.",
                      ["category"] = "technology",
                      ["quarter"] = "Q4",
                  },
              },  
              new UpsertRecord
              {
                  Id = "rec3",
                  AdditionalProperties =
                  {
                      ["chunk_text"] = "AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space.",
                      ["category"] = "technology",
                      ["quarter"] = "Q4",
                  },
              },
              new UpsertRecord
              {
                  Id = "rec4",
                  AdditionalProperties =
                  {
                      ["chunk_text"] = "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production",
                      ["category"] = "technology",
                      ["quarter"] = "Q3",
                  },
              },
          ]
      );
      ```

      ```shell curl theme={null}
      # To get the unique host for an index,
      # see https://docs.pinecone.io/guides/manage-data/target-an-index
      INDEX_HOST="INDEX_HOST"
      NAMESPACE="YOUR_NAMESPACE"
      PINECONE_API_KEY="YOUR_API_KEY"

      # Upsert records into a namespace
      # `chunk_text` fields are converted to dense vectors
      # `category` is stored as metadata
      curl "https://$INDEX_HOST/records/namespaces/$NAMESPACE/upsert" \
        -H "Content-Type: application/x-ndjson" \
        -H "Api-Key: $PINECONE_API_KEY" \
        -H "X-Pinecone-Api-Version: 2025-10" \
        -d '{"_id": "rec1", "chunk_text": "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut.", "category": "digestive system"}
            {"_id": "rec2", "chunk_text": "Apples originated in Central Asia and have been cultivated for thousands of years, with over 7,500 varieties available today.", "category": "cultivation"}
            {"_id": "rec3", "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases.", "category": "immune system"}
            {"_id": "rec4", "chunk_text": "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes.", "category": "endocrine system"}'
      ```
    </CodeGroup>
  </Tab>

  <Tab title="Upsert vectors">
    To upsert dense vectors into a [dense index](/guides/index-data/create-an-index#create-a-dense-index), use the [`upsert`](/reference/api/latest/data-plane/upsert) operation as follows:

    * Specify the [`namespace`](/guides/index-data/indexing-overview#namespaces) to upsert into. If the namespace doesn't exist, it is created. To use the default namespace, set the namespace to `"__default__"`.
    * Format your input data as records, each with the following:
      * An `id` field with a unique record identifier for the index namespace.
      * A `values` field with the dense vector values.
      * Optionally, a `metadata` field with [key-value pairs](/guides/index-data/indexing-overview#metadata) to store additional information or context. When you query the index, you can use metadata to [filter search results](/guides/search/filter-by-metadata).

    <CodeGroup>
      ```Python Python theme={null}
      from pinecone.grpc import PineconeGRPC as Pinecone

      pc = Pinecone(api_key="YOUR_API_KEY")

      # To get the unique host for an index, 
      # see https://docs.pinecone.io/guides/manage-data/target-an-index
      index = pc.Index(host="INDEX_HOST")

      index.upsert(
        vectors=[
          {
            "id": "A", 
            "values": [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1], 
            "metadata": {"genre": "comedy", "year": 2020}
          },
          {
            "id": "B", 
            "values": [0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2],
            "metadata": {"genre": "documentary", "year": 2019}
          },
          {
            "id": "C", 
            "values": [0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3],
            "metadata": {"genre": "comedy", "year": 2019}
          },
          {
            "id": "D", 
            "values": [0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4],
            "metadata": {"genre": "drama"}
          }
        ],
        namespace="example-namespace"
      )
      ```

      ```javascript JavaScript theme={null}
      import { Pinecone } from '@pinecone-database/pinecone'

      const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      const index = pc.index("INDEX_NAME", "INDEX_HOST")

      const records = [
          {
            id: 'A',
            values: [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1],
            metadata: { genre: "comedy", year: 2020 },
          },
          {
            id: 'B',
            values: [0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2],
            metadata: { genre: "documentary", year: 2019 },
          },
          {
            id: 'C',
            values: [0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3],
            metadata: { genre: "comedy", year: 2019 },
          },
          {
            id: 'D',
            values: [0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4],
            metadata: { genre: "drama" },
          }
      ]

      await index.('example-namespace').upsert(records);
      ```

      ```java Java theme={null}
      import com.google.protobuf.Struct;
      import com.google.protobuf.Value;
      import io.pinecone.clients.Index;
      import io.pinecone.configs.PineconeConfig;
      import io.pinecone.configs.PineconeConnection;

      import java.util.Arrays;
      import java.util.List;

      public class UpsertExample {
          public static void main(String[] args) {
              PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
              // To get the unique host for an index, 
              // see https://docs.pinecone.io/guides/manage-data/target-an-index
              config.setHost("INDEX_HOST");
              PineconeConnection connection = new PineconeConnection(config);
              Index index = new Index(connection, "INDEX_NAME");
              List<Float> values1 = Arrays.asList(0.1f, 0.1f, 0.1f, 0.1f, 0.1f, 0.1f, 0.1f, 0.1f);
              List<Float> values2 = Arrays.asList(0.2f, 0.2f, 0.2f, 0.2f, 0.2f, 0.2f, 0.2f, 0.2f);
              List<Float> values3 = Arrays.asList(0.3f, 0.3f, 0.3f, 0.3f, 0.3f, 0.3f, 0.3f, 0.3f);
              List<Float> values4 = Arrays.asList(0.4f, 0.4f, 0.4f, 0.4f, 0.4f, 0.4f, 0.4f, 0.4f);
              Struct metaData1 = Struct.newBuilder()
                      .putFields("genre", Value.newBuilder().setStringValue("comedy").build())
                      .putFields("year", Value.newBuilder().setNumberValue(2020).build())
                      .build();
              Struct metaData2 = Struct.newBuilder()
                      .putFields("genre", Value.newBuilder().setStringValue("documentary").build())
                      .putFields("year", Value.newBuilder().setNumberValue(2019).build())
                      .build();
              Struct metaData3 = Struct.newBuilder()
                      .putFields("genre", Value.newBuilder().setStringValue("comedy").build())
                      .putFields("year", Value.newBuilder().setNumberValue(2019).build())
                      .build();
              Struct metaData4 = Struct.newBuilder()
                      .putFields("genre", Value.newBuilder().setStringValue("drama").build())
                      .build();

              index.upsert("A", values1, null, null, metaData1, 'example-namespace');
              index.upsert("B", values2, null, null, metaData2, 'example-namespace');
              index.upsert("C", values3, null, null, metaData3, 'example-namespace');
              index.upsert("D", values4, null, null, metaData4, 'example-namespace');
          }
      }
      ```

      ```go Go theme={null}
      package main

      import (
          "context"
          "fmt"
          "log"

          "github.com/pinecone-io/go-pinecone/v4/pinecone"
          "google.golang.org/protobuf/types/known/structpb"
      )

      func main() {
          ctx := context.Background()

          pc, err := pinecone.NewClient(pinecone.NewClientParams{
              ApiKey: "YOUR_API_KEY",
          })
          if err != nil {
              log.Fatalf("Failed to create Client: %v", err)
          }

          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
          if err != nil {
              log.Fatalf("Failed to create IndexConnection for Host: %v", err)
      	  }

          metadataMap1 := map[string]interface{}{
              "genre": "comedy",
              "year": 2020,
          }

          metadata1, err := structpb.NewStruct(metadataMap1)
          if err != nil {
              log.Fatalf("Failed to create metadata map: %v", err)
          }

          metadataMap2 := map[string]interface{}{
              "genre": "documentary",
              "year": 2019,
          }

          metadata2, err := structpb.NewStruct(metadataMap2)
          if err != nil {
              log.Fatalf("Failed to create metadata map: %v", err)
          }

          metadataMap3 := map[string]interface{}{
              "genre": "comedy",
              "year": 2019,
          }

          metadata3, err := structpb.NewStruct(metadataMap3)
          if err != nil {
              log.Fatalf("Failed to create metadata map: %v", err)
          }

          metadataMap4 := map[string]interface{}{
              "genre": "drama",
          }

          metadata4, err := structpb.NewStruct(metadataMap4)
          if err != nil {
              log.Fatalf("Failed to create metadata map: %v", err)
          }

          vectors := []*pinecone.Vector{
              {
                  Id:     "A",
                  Values: []float32{0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1},
                  Metadata: metadata1,
              },
              {
                  Id:     "B",
                  Values: []float32{0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2},
                  Metadata: metadata2,
              },
              {
                  Id:     "C",
                  Values: []float32{0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3},
                  Metadata: metadata3,
              },   
              {
                  Id:     "D",
                  Values: []float32{0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4},
                  Metadata: metadata4,
              },   
          }

          count, err := idxConnection.UpsertVectors(ctx, vectors)
          if err != nil {
              log.Fatalf("Failed to upsert vectors: %v", err)
          } else {
              fmt.Printf("Successfully upserted %d vector(s)!\n", count)
          }
      }
      ```

      ```csharp C# theme={null}
      using Pinecone;

      var pinecone = new PineconeClient("YOUR_API_KEY");

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      var index = pinecone.Index(host: "INDEX_HOST");

      var upsertResponse = await index.UpsertAsync(new UpsertRequest {
          Vectors = new[]
          {
              new Vector
              {
                  Id = "A",
                  Values = new[] { 0.1f, 0.1f, 0.1f, 0.1f, 0.1f, 0.1f, 0.1f, 0.1f },
                  Metadata = new Metadata {
                      ["genre"] = new("comedy"),
                      ["year"] = new(2020),
                  },
              },
              new Vector
              {
                  Id = "B",
                  Values = new[] { 0.2f, 0.2f, 0.2f, 0.2f, 0.2f, 0.2f, 0.2f, 0.2f },
                  Metadata = new Metadata {
                      ["genre"] = new("documentary"),
                      ["year"] = new(2019),
                  },
              },
              new Vector
              {
                  Id = "C",
                  Values = new[] { 0.3f, 0.3f, 0.3f, 0.3f, 0.3f, 0.3f, 0.3f, 0.3f },
                  Metadata = new Metadata {
                      ["genre"] = new("comedy"),
                      ["year"] = new(2019),
                  },
              },
              new Vector
              {
                  Id = "D",
                  Values = new[] { 0.4f, 0.4f, 0.4f, 0.4f, 0.4f, 0.4f, 0.4f, 0.4f },
                  Metadata = new Metadata {
                      ["genre"] = new("drama"),
                  },
              }
          },
          Namespace = "example-namespace",
      });
      ```

      ```bash curl theme={null}
      # To get the unique host for an index,
      # see https://docs.pinecone.io/guides/manage-data/target-an-index
      PINECONE_API_KEY="YOUR_API_KEY"
      INDEX_HOST="INDEX_HOST"

      curl "https://$INDEX_HOST/vectors/upsert" \
        -H "Api-Key: $PINECONE_API_KEY" \
        -H 'Content-Type: application/json' \
        -H "X-Pinecone-Api-Version: 2025-10" \
        -d '{
          "vectors": [
            {
              "id": "A",
              "values": [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1],
              "metadata": {"genre": "comedy", "year": 2020}
            },
            {
              "id": "B",
              "values": [0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2],
              "metadata": {"genre": "documentary", "year": 2019}
            },
            {
              "id": "C",
              "values": [0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3],
              "metadata": {"genre": "comedy", "year": 2019}
            },
            {
              "id": "D",
              "values": [0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4],
              "metadata": {"genre": "drama"}
            }
          ],
          "namespace": "example-namespace"
        }'
      ```
    </CodeGroup>
  </Tab>
</Tabs>

## Upsert sparse vectors

<Tabs>
  <Tab title="Upsert text">
    <Note>
      Upserting text is supported only for [indexes with integrated embedding](/guides/index-data/indexing-overview#integrated-embedding).
    </Note>

    To upsert source text into a [sparse index with integrated embedding](/guides/index-data/create-an-index#create-a-sparse-index), use the [`upsert_records`](/reference/api/latest/data-plane/upsert_records) operation. Pinecone converts the text to sparse vectors automatically using the hosted sparse embedding model associated with the index.

    * Specify the [`namespace`](/guides/index-data/indexing-overview#namespaces) to upsert into. If the namespace doesn't exist, it is created. To use the default namespace, set the namespace to `"__default__"`.
    * Format your input data as records, each with the following:
      * An `_id` field with a unique record identifier for the index namespace. `id` can be used as an alias for `_id`.
      * A field with the source text to convert to a vector. This field must match the `field_map` specified in the index.
      * Additional fields are stored as record [metadata](/guides/index-data/indexing-overview#metadata) and can be returned in search results or used to [filter search results](/guides/search/filter-by-metadata).

    For example, the following code converts the sentences in the `chunk_text` fields to sparse vectors and then upserts them into `example-namespace` in an example index. The additional `category` and `quarter` fields are stored as metadata.

    <CodeGroup>
      ```python Python theme={null}
      from pinecone import Pinecone

      pc = Pinecone(api_key="YOUR_API_KEY")

      # To get the unique host for an index, 
      # see https://docs.pinecone.io/guides/manage-data/target-an-index
      index = pc.Index(host="INDEX_HOST")

      # Upsert records into a namespace
      # `chunk_text` fields are converted to sparse vectors
      # `category` and `quarter` fields are stored as metadata
      index.upsert_records(
          "example-namespace",
          [
              { 
                  "_id": "vec1", 
                  "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.", 
                  "category": "technology",
                  "quarter": "Q3"
              },
              { 
                  "_id": "vec2", 
                  "chunk_text": "Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.", 
                  "category": "technology",
                  "quarter": "Q4"
              },
              { 
                  "_id": "vec3", 
                  "chunk_text": "AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production.",
                  "category": "technology",
                  "quarter": "Q3"
              },
              { 
                  "_id": "vec4", 
                  "chunk_text": "AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space.", 
                  "category": "technology",
                  "quarter": "Q4"
              }
          ]
      )

      time.sleep(10) # Wait for the upserted vectors to be indexed
      ```

      ```javascript JavaScript theme={null}
      import { Pinecone } from '@pinecone-database/pinecone'

      const pc = new Pinecone({ apiKey: "YOUR_API_KEY" })

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      const namespace = pc.index("INDEX_NAME", "INDEX_HOST").namespace("example-namespace");

      // Upsert records into a namespace
      // `chunk_text` fields are converted to sparse vectors
      // `category` and `quarter` fields are stored as metadata
      await namespace.upsertRecords([
          { 
              "_id": "vec1", 
              "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.", 
              "category": "technology",
              "quarter": "Q3"
          },
          { 
              "_id": "vec2", 
              "chunk_text": "Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.", 
              "category": "technology",
              "quarter": "Q4"
          },
          { 
              "_id": "vec3", 
              "chunk_text": "AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production.",
              "category": "technology",
              "quarter": "Q3"
          },
          { 
              "_id": "vec4", 
              "chunk_text": "AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space.", 
              "category": "technology",
              "quarter": "Q4"
          }
      ]);
      ```

      ```java Java theme={null}
      import io.pinecone.clients.Index;
      import io.pinecone.configs.PineconeConfig;
      import io.pinecone.configs.PineconeConnection;
      import org.openapitools.db_data.client.ApiException;

      import java.util.*;

      public class UpsertText {
          public static void main(String[] args) throws ApiException {
              PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
              config.setHost("INDEX_HOST");
              PineconeConnection connection = new PineconeConnection(config);

              Index index = new Index(config, connection, "integrated-sparse-java");
              ArrayList<Map<String, String>> upsertRecords = new ArrayList<>();

              HashMap<String, String> record1 = new HashMap<>();
              record1.put("_id", "rec1");
              record1.put("category", "digestive system");
              record1.put("chunk_text", "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut.");

              HashMap<String, String> record2 = new HashMap<>();
              record2.put("_id", "rec2");
              record2.put("category", "cultivation");
              record2.put("chunk_text", "Apples originated in Central Asia and have been cultivated for thousands of years, with over 7,500 varieties available today.");

              HashMap<String, String> record3 = new HashMap<>();
              record3.put("_id", "rec3");
              record3.put("category", "immune system");
              record3.put("chunk_text", "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases.");

              HashMap<String, String> record4 = new HashMap<>();
              record4.put("_id", "rec4");
              record4.put("category", "endocrine system");
              record4.put("chunk_text", "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes.");

              upsertRecords.add(record1);
              upsertRecords.add(record2);
              upsertRecords.add(record3);
              upsertRecords.add(record4);

              index.upsertRecords("example-namespace", upsertRecords);
          }
      }
      ```

      ```go Go theme={null}
      package main

      import (
          "context"
          "fmt"
          "log"

          "github.com/pinecone-io/go-pinecone/v4/pinecone"
      )

      func main() {
          ctx := context.Background()

          pc, err := pinecone.NewClient(pinecone.NewClientParams{
              ApiKey: "YOUR_API_KEY",
          })
          if err != nil {
              log.Fatalf("Failed to create Client: %v", err)
          }

          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
          if err != nil {
              log.Fatalf("Failed to create IndexConnection for Host: %v", err)
      	  }

          // Upsert records into a namespace
          // `chunk_text` fields are converted to sparse vectors
          // `category` and `quarter` fields are stored as metadata
      	records := []*pinecone.IntegratedRecord{
      		{
      			"_id":        "vec1",
      			"chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
      			"category":   "technology",
      			"quarter":    "Q3",
      		},
      		{
      			"_id":        "vec2",
      			"chunk_text": "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
      			"category":   "technology",
      			"quarter":    "Q4",
      		},
      		{
      			"_id":        "vec3",
      			"chunk_text": "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production.",
      			"category":   "technology",
      			"quarter":    "Q3",
      		},
      		{
      			"_id":        "vec4",
      			"chunk_text": "AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space.",
      			"category":   "technology",
      			"quarter":    "Q4",
      		},
      	}

      	err = idxConnection.UpsertRecords(ctx, records)
      	if err != nil {
      		log.Fatalf("Failed to upsert vectors: %v", err)
      	}
      }
      ```

      ```csharp C# theme={null}
      using Pinecone;

      var pinecone = new PineconeClient("YOUR_API_KEY");

      var index = pinecone.Index(host: "INDEX_HOST");

      await index.UpsertRecordsAsync(
          "example-namespace",
          [
              new UpsertRecord
              {
                  Id = "rec1",
                  AdditionalProperties =
                  {
                      ["chunk_text"] = "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
                      ["category"] = "technology",
                      ["quarter"] = "Q3",
                  },
              },
              new UpsertRecord
              {
                  Id = "rec2",
                  AdditionalProperties =
                  {
                      ["chunk_text"] = "AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space.",
                      ["category"] = "technology",
                      ["quarter"] = "Q4",
                  },
              },  
              new UpsertRecord
              {
                  Id = "rec3",
                  AdditionalProperties =
                  {
                      ["chunk_text"] = "AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space.",
                      ["category"] = "technology",
                      ["quarter"] = "Q4",
                  },
              },
              new UpsertRecord
              {
                  Id = "rec4",
                  AdditionalProperties =
                  {
                      ["chunk_text"] = "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production",
                      ["category"] = "technology",
                      ["quarter"] = "Q3",
                  },
              },
          ]
      );
      ```

      ```shell curl theme={null}
      INDEX_HOST="INDEX_HOST"
      NAMESPACE="YOUR_NAMESPACE"
      PINECONE_API_KEY="YOUR_API_KEY"

      curl  "https://$INDEX_HOST/records/namespaces/$NAMESPACE/upsert" \
          -H "Content-Type: application/x-ndjson" \
          -H "Api-Key: $PINECONE_API_KEY" \
          -H "X-Pinecone-Api-Version: 2025-10" \
          -d '{ "_id": "vec1", "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.", "category": "technology", "quarter": "Q3" }
            { "_id": "vec2", "chunk_text": "Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.", "category": "technology", "quarter": "Q4" }
            { "_id": "vec3", "chunk_text": "AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production.", "category": "technology", "quarter": "Q3" }
            { "_id": "vec4", "chunk_text": "AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space.", "category": "technology", "quarter": "Q4" }'
      ```
    </CodeGroup>
  </Tab>

  <Tab title="Upsert vectors">
    To upsert sparse vectors into a [sparse index](/guides/index-data/create-an-index#create-a-sparse-index), use the [`upsert`](/reference/api/latest/data-plane/upsert) operation as follows:

    * Specify the [`namespace`](/guides/index-data/indexing-overview#namespaces) to upsert into. If the namespace doesn't exist, it is created. To use the default namespace, set the namespace to `"__default__"`.
    * Format your input data as records, each with the following:
      * An `id` field with a unique record identifier for the index namespace.
      * A `sparse_values` field with the sparse vector values and indices.
      * Optionally, a `metadata` field with [key-value pairs](/guides/index-data/indexing-overview#metadata) to store additional information or context. When you query the index, you can use metadata to [filter search results](/guides/search/filter-by-metadata).

    For example, the following code upserts sparse vector representations of sentences related to the term "apple", with the source text and additional fields stored as metadata:

    <CodeGroup>
      ```python Python theme={null}
      from pinecone import Pinecone, SparseValues, Vector

      pc = Pinecone(api_key="YOUR_API_KEY")

      # To get the unique host for an index, 
      # see https://docs.pinecone.io/guides/manage-data/target-an-index
      index = pc.Index(host="INDEX_HOST")

      index.upsert(
          namespace="example-namespace",
          vectors=[
              {
                  "id": "vec1",
                  "sparse_values": {
                      "values": [1.7958984, 0.41577148, 2.828125, 2.8027344, 2.8691406, 1.6533203, 5.3671875, 1.3046875, 0.49780273, 0.5722656, 2.71875, 3.0820312, 2.5019531, 4.4414062, 3.3554688],
                      "indices": [822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191]
                  },
                  "metadata": {
                      "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
                      "category": "technology",
                      "quarter": "Q3"
                  }
              },
              {
                  "id": "vec2",
                  "sparse_values": {
                      "values": [0.4362793, 3.3457031, 2.7714844, 3.0273438, 3.3164062, 5.6015625, 2.4863281, 0.38134766, 1.25, 2.9609375, 0.34179688, 1.4306641, 0.34375, 3.3613281, 1.4404297, 2.2558594, 2.2597656, 4.8710938, 0.5605469],
                      "indices": [131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697]
                  },
                  "metadata": {
                      "chunk_text": "Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
                      "category": "technology",
                      "quarter": "Q4"
                  }
              },
              {
                  "id": "vec3",
                  "sparse_values": {
                      "values": [2.6875, 4.2929688, 3.609375, 3.0722656, 2.1152344, 5.78125, 3.7460938, 3.7363281, 1.2695312, 3.4824219, 0.7207031, 0.0826416, 4.671875, 3.7011719, 2.796875, 0.61621094],
                      "indices": [8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697]
                  },
                  "metadata": {
                      "chunk_text": "AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production",
                      "category": "technology",
                      "quarter": "Q3"
                  }
              },
              {
                  "id": "vec4",
                  "sparse_values": {
                      "values": [0.73046875, 0.46972656, 2.84375, 5.2265625, 3.3242188, 1.9863281, 0.9511719, 0.5019531, 4.4257812, 3.4277344, 0.41308594, 4.3242188, 2.4179688, 3.1757812, 1.0224609, 2.0585938, 2.5859375],
                      "indices": [131900689, 152217691, 441495248, 1640781426, 1851149807, 2263326288, 2502307765, 2641553256, 2684780967, 2966813704, 3162218338, 3283104238, 3488055477, 3530642888, 3888762515, 4152503047, 4177290673]
                  },
                  "metadata": {
                      "chunk_text": "AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space.",
                      "category": "technology",
                      "quarter": "Q4"
                  }
              }
          ]
      )
      ```

      ```javascript JavaScript theme={null}
      import { Pinecone } from '@pinecone-database/pinecone'

      const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      const index = pc.index("INDEX_NAME", "INDEX_HOST")

      await index.namespace('example-namespace').upsert([
        {
          id: 'vec1',
          sparseValues: {
            indices: [822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191],
            values: [1.7958984, 0.41577148, 2.828125, 2.8027344, 2.8691406, 1.6533203, 5.3671875, 1.3046875, 0.49780273, 0.5722656, 2.71875, 3.0820312, 2.5019531, 4.4414062, 3.3554688]
          },
          metadata: { 
            chunk_text: 'AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.', 
            category: 'technology',
            quarter: 'Q3' 
          }
        },
        {
          id: 'vec2',
          sparseValues: {
            indices: [131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697],
            values: [0.4362793, 3.3457031, 2.7714844, 3.0273438, 3.3164062, 5.6015625, 2.4863281, 0.38134766, 1.25, 2.9609375, 0.34179688, 1.4306641, 0.34375, 3.3613281, 1.4404297, 2.2558594, 2.2597656, 4.8710938, 0.5605469]
          },
          metadata: { 
            chunk_text: "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market.", 
            category: 'technology',
            quarter: 'Q4' 
          }
        },
        {
          id: 'vec3',
          sparseValues: {
            indices: [8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697],
            values: [2.6875, 4.2929688, 3.609375, 3.0722656, 2.1152344, 5.78125, 3.7460938, 3.7363281, 1.2695312, 3.4824219, 0.7207031, 0.0826416, 4.671875, 3.7011719, 2.796875, 0.61621094]
          },
          metadata: { 
            chunk_text: "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production", 
            category: 'technology',
            quarter: 'Q3' 
          }
        },
        {
          id: 'vec4',
          sparseValues: {
            indices: [131900689, 152217691, 441495248, 1640781426, 1851149807, 2263326288, 2502307765, 2641553256, 2684780967, 2966813704, 3162218338, 3283104238, 3488055477, 3530642888, 3888762515, 4152503047, 4177290673],
            values: [0.73046875, 0.46972656, 2.84375, 5.2265625, 3.3242188, 1.9863281, 0.9511719, 0.5019531, 4.4257812, 3.4277344, 0.41308594, 4.3242188, 2.4179688, 3.1757812, 1.0224609, 2.0585938, 2.5859375]
          },
          metadata: { 
            chunk_text: 'AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space.', 
            category: 'technology',
            quarter: 'Q4' 
          }
        }
      ]);
      ```

      ```java Java theme={null}
      import io.pinecone.clients.Pinecone;
      import io.pinecone.clients.Index;
      import com.google.protobuf.Struct;
      import com.google.protobuf.Value;

      import java.util.*;

      public class UpsertSparseVectors {
          public static void main(String[] args) throws InterruptedException {
              // Instantiate Pinecone class
              Pinecone pinecone = new Pinecone.Builder("YOUR_API)KEY").build();
              
              Index index = pinecone.getIndexConnection("docs-example");

              // Record 1
              ArrayList<Long> indices1 = new ArrayList<>(Arrays.asList(
                      822745112L, 1009084850L, 1221765879L, 1408993854L, 1504846510L,
                      1596856843L, 1640781426L, 1656251611L, 1807131503L, 2543655733L,
                      2902766088L, 2909307736L, 3246437992L, 3517203014L, 3590924191L
              ));

              ArrayList<Float> values1 = new ArrayList<>(Arrays.asList(
                      1.7958984f, 0.41577148f, 2.828125f, 2.8027344f, 2.8691406f,
                      1.6533203f, 5.3671875f, 1.3046875f, 0.49780273f, 0.5722656f,
                      2.71875f, 3.0820312f, 2.5019531f, 4.4414062f, 3.3554688f
              ));

              Struct metaData1 = Struct.newBuilder()
                      .putFields("chunk_text", Value.newBuilder().setStringValue("AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.").build())
                      .putFields("category", Value.newBuilder().setStringValue("technology").build())
                      .putFields("quarter", Value.newBuilder().setStringValue("Q3").build())
                      .build();

              // Record 2
              ArrayList<Long> indices2 = new ArrayList<>(Arrays.asList(
                      131900689L, 592326839L, 710158994L, 838729363L, 1304885087L,
                      1640781426L, 1690623792L, 1807131503L, 2066971792L, 2428553208L,
                      2548600401L, 2577534050L, 3162218338L, 3319279674L, 3343062801L,
                      3476647774L, 3485013322L, 3517203014L, 4283091697L
              ));

              ArrayList<Float> values2 = new ArrayList<>(Arrays.asList(
                      0.4362793f, 3.3457031f, 2.7714844f, 3.0273438f, 3.3164062f,
                      5.6015625f, 2.4863281f, 0.38134766f, 1.25f, 2.9609375f,
                      0.34179688f, 1.4306641f, 0.34375f, 3.3613281f, 1.4404297f,
                      2.2558594f, 2.2597656f, 4.8710938f, 0.5605469f
              ));

              Struct metaData2 = Struct.newBuilder()
                      .putFields("chunk_text", Value.newBuilder().setStringValue("Analysts suggest that AAPL'\\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.").build())
                      .putFields("category", Value.newBuilder().setStringValue("technology").build())
                      .putFields("quarter", Value.newBuilder().setStringValue("Q4").build())
                      .build();

              // Record 3
              ArrayList<Long> indices3 = new ArrayList<>(Arrays.asList(
                      8661920L, 350356213L, 391213188L, 554637446L, 1024951234L,
                      1640781426L, 1780689102L, 1799010313L, 2194093370L, 2632344667L,
                      2641553256L, 2779594451L, 3517203014L, 3543799498L,
                      3837503950L, 4283091697L
              ));

              ArrayList<Float> values3 = new ArrayList<>(Arrays.asList(
                      2.6875f, 4.2929688f, 3.609375f, 3.0722656f, 2.1152344f,
                      5.78125f, 3.7460938f, 3.7363281f, 1.2695312f, 3.4824219f,
                      0.7207031f, 0.0826416f, 4.671875f, 3.7011719f, 2.796875f,
                      0.61621094f
              ));

              Struct metaData3 = Struct.newBuilder()
                      .putFields("chunk_text", Value.newBuilder().setStringValue("AAPL'\\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production").build())
                      .putFields("category", Value.newBuilder().setStringValue("technology").build())
                      .putFields("quarter", Value.newBuilder().setStringValue("Q3").build())
                      .build();

              // Record 4
              ArrayList<Long> indices4 = new ArrayList<>(Arrays.asList(
                      131900689L, 152217691L, 441495248L, 1640781426L, 1851149807L,
                      2263326288L, 2502307765L, 2641553256L, 2684780967L, 2966813704L,
                      3162218338L, 3283104238L, 3488055477L, 3530642888L, 3888762515L,
                      4152503047L, 4177290673L
              ));

              ArrayList<Float> values4 = new ArrayList<>(Arrays.asList(
                      0.73046875f, 0.46972656f, 2.84375f, 5.2265625f, 3.3242188f,
                      1.9863281f, 0.9511719f, 0.5019531f, 4.4257812f, 3.4277344f,
                      0.41308594f, 4.3242188f, 2.4179688f, 3.1757812f, 1.0224609f,
                      2.0585938f, 2.5859375f
              ));

              Struct metaData4 = Struct.newBuilder()
                      .putFields("chunk_text", Value.newBuilder().setStringValue("AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space").build())
                      .putFields("category", Value.newBuilder().setStringValue("technology").build())
                      .putFields("quarter", Value.newBuilder().setStringValue("Q4").build())
                      .build();

              index.upsert("vec1", Collections.emptyList(), indices1, values1, metaData1, "example-namespace");
              index.upsert("vec2", Collections.emptyList(), indices2, values2, metaData2, "example-namespace");
              index.upsert("vec3", Collections.emptyList(), indices3, values3, metaData3, "example-namespace");
              index.upsert("vec4", Collections.emptyList(), indices4, values4, metaData4, "example-namespace");
      ```

      ```go Go theme={null}
      package main

      import (
      	"context"
      	"fmt"
      	"log"

      	"github.com/pinecone-io/go-pinecone/v4/pinecone"
      	"google.golang.org/protobuf/types/known/structpb"
      )

      func main() {
      	ctx := context.Background()

      	pc, err := pinecone.NewClient(pinecone.NewClientParams{
      		ApiKey: "YOUR_API_KEY",
      	})
      	if err != nil {
      		log.Fatalf("Failed to create Client: %v", err)
      	}

      	// To get the unique host for an index,
      	// see https://docs.pinecone.io/guides/manage-data/target-an-index
      	idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST", Namespace: "example-namespace"})
      	if err != nil {
      		log.Fatalf("Failed to create IndexConnection for Host: %v", err)
      	}

      	sparseValues1 := pinecone.SparseValues{
      		Indices: []uint32{822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191},
      		Values:  []float32{1.7958984, 0.41577148, 2.828125, 2.8027344, 2.8691406, 1.6533203, 5.3671875, 1.3046875, 0.49780273, 0.5722656, 2.71875, 3.0820312, 2.5019531, 4.4414062, 3.3554688},
      	}

      	metadataMap1 := map[string]interface{}{
      		"chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones",
      		"category":    "technology",
      		"quarter":     "Q3",
      	}

      	metadata1, err := structpb.NewStruct(metadataMap1)
      	if err != nil {
      		log.Fatalf("Failed to create metadata map: %v", err)
      	}

      	sparseValues2 := pinecone.SparseValues{
      		Indices: []uint32{131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697},
      		Values:  []float32{0.4362793, 3.3457031, 2.7714844, 3.0273438, 3.3164062, 5.6015625, 2.4863281, 0.38134766, 1.25, 2.9609375, 0.34179688, 1.4306641, 0.34375, 3.3613281, 1.4404297, 2.2558594, 2.2597656, 4.8710938, 0.560546},
      	}

      	metadataMap2 := map[string]interface{}{
      		"chunk_text": "Analysts suggest that AAPL's upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
      		"category":    "technology",
      		"quarter":     "Q4",
      	}

      	metadata2, err := structpb.NewStruct(metadataMap2)
      	if err != nil {
      		log.Fatalf("Failed to create metadata map: %v", err)
      	}

      	sparseValues3 := pinecone.SparseValues{
      		Indices: []uint32{8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697},
      		Values:  []float32{2.6875, 4.2929688, 3.609375, 3.0722656, 2.1152344, 5.78125, 3.7460938, 3.7363281, 1.2695312, 3.4824219, 0.7207031, 0.0826416, 4.671875, 3.7011719, 2.796875, 0.61621094},
      	}

      	metadataMap3 := map[string]interface{}{
      		"chunk_text": "AAPL's strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production",
      		"category":    "technology",
      		"quarter":     "Q3",
      	}

      	metadata3, err := structpb.NewStruct(metadataMap3)
      	if err != nil {
      		log.Fatalf("Failed to create metadata map: %v", err)
      	}

      	sparseValues4 := pinecone.SparseValues{
      		Indices: []uint32{131900689, 152217691, 441495248, 1640781426, 1851149807, 2263326288, 2502307765, 2641553256, 2684780967, 2966813704, 3162218338, 3283104238, 3488055477, 3530642888, 3888762515, 4152503047, 4177290673},
      		Values:  []float32{0.73046875, 0.46972656, 2.84375, 5.2265625, 3.3242188, 1.9863281, 0.9511719, 0.5019531, 4.4257812, 3.4277344, 0.41308594, 4.3242188, 2.4179688, 3.1757812, 1.0224609, 2.0585938, 2.5859375},
      	}

      	metadataMap4 := map[string]interface{}{
      		"chunk_text": "AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space.",
      		"category":    "technology",
      		"quarter":     "Q4",
      	}

      	metadata4, err := structpb.NewStruct(metadataMap4)
      	if err != nil {
      		log.Fatalf("Failed to create metadata map: %v", err)
      	}

      	vectors := []*pinecone.Vector{
      		{
      			Id:           "vec1",
      			SparseValues: &sparseValues1,
      			Metadata:     metadata1,
      		},
      		{
      			Id:           "vec2",
      			SparseValues: &sparseValues2,
      			Metadata:     metadata2,
      		},
      		{
      			Id:           "vec3",
      			SparseValues: &sparseValues3,
      			Metadata:     metadata3,
      		},
      		{
      			Id:           "vec4",
      			SparseValues: &sparseValues4,
      			Metadata:     metadata4,
      		},
      	}

      	count, err := idxConnection.UpsertVectors(ctx, vectors)
      	if err != nil {
      		log.Fatalf("Failed to upsert vectors: %v", err)
      	} else {
      		fmt.Printf("Successfully upserted %d vector(s)!\n", count)
      	}
      }
      ```

      ```csharp C# theme={null}
      using Pinecone;

      var pinecone = new PineconeClient("YOUR_API_KEY");

      var index = pinecone.Index("docs-example");

      var vector1 = new Vector
      {
          Id = "vec1",
          SparseValues = new SparseValues
          {
              Indices = new uint[] { 822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191 },
              Values = new ReadOnlyMemory<float>([1.7958984f, 0.41577148f, 2.828125f, 2.8027344f, 2.8691406f, 1.6533203f, 5.3671875f, 1.3046875f, 0.49780273f, 0.5722656f, 2.71875f, 3.0820312f, 2.5019531f, 4.4414062f, 3.3554688f])
          },
          Metadata = new Metadata {
              ["chunk_text"] = new("AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones."),
              ["category"] = new("technology"),
              ["quarter"] = new("Q3"),
          },
      };

      var vector2 = new Vector
      {
          Id = "vec2",
          SparseValues = new SparseValues
          {
              Indices = new uint[] { 131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697 },
              Values = new ReadOnlyMemory<float>([0.4362793f, 3.3457031f, 2.7714844f, 3.0273438f, 3.3164062f, 5.6015625f, 2.4863281f, 0.38134766f, 1.25f, 2.9609375f, 0.34179688f, 1.4306641f, 0.34375f, 3.3613281f, 1.4404297f, 2.2558594f, 2.2597656f, 4.8710938f, 0.5605469f])
          },
          Metadata = new Metadata {
              ["chunk_text"] = new("Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market."),
              ["category"] = new("technology"),
              ["quarter"] = new("Q4"),
          },
      };

      var vector3 = new Vector
      {
          Id = "vec3",
          SparseValues = new SparseValues
          {
              Indices = new uint[] { 8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697 },
              Values = new ReadOnlyMemory<float>([2.6875f, 4.2929688f, 3.609375f, 3.0722656f, 2.1152344f, 5.78125f, 3.7460938f, 3.7363281f, 1.2695312f, 3.4824219f, 0.7207031f, 0.0826416f, 4.671875f, 3.7011719f, 2.796875f, 0.61621094f])
          },
          Metadata = new Metadata {
              ["chunk_text"] = new("AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production"),
              ["category"] = new("technology"),
              ["quarter"] = new("Q3"),
          },    
      };

      var vector4 = new Vector
      {
          Id = "vec4",
          SparseValues = new SparseValues
          {
              Indices = new uint[] { 131900689, 152217691, 441495248, 1640781426, 1851149807, 2263326288, 2502307765, 2641553256, 2684780967, 2966813704, 3162218338, 3283104238, 3488055477, 3530642888, 3888762515, 4152503047, 4177290673 },
              Values = new ReadOnlyMemory<float>([0.73046875f, 0.46972656f, 2.84375f, 5.2265625f, 3.3242188f, 1.9863281f, 0.9511719f, 0.5019531f, 4.4257812f, 3.4277344f, 0.41308594f, 4.3242188f, 2.4179688f, 3.1757812f, 1.0224609f, 2.0585938f, 2.5859375f])
          },
          Metadata = new Metadata {
              ["chunk_text"] = new("AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space."),
              ["category"] = new("technology"),
              ["quarter"] = new("Q4"),
          },
      };

      // Upsert vector
      Console.WriteLine("Upserting vector...");
      var upsertResponse = await index.UpsertAsync(new UpsertRequest
      {
          Vectors = new List<Vector> { vector1, vector2, vector3, vector4 },
          Namespace = "example-namespace"
      });
      Console.WriteLine($"Upserted {upsertResponse.UpsertedCount} vector");
      ```

      ```shell curl theme={null}
      INDEX_HOST="INDEX_HOST"
      PINECONE_API_KEY="YOUR_API_KEY"

      curl "http://$INDEX_HOST/vectors/upsert" \
        -H "Content-Type: application/json" \
        -H "Api-Key: $PINECONE_API_KEY" \
        -H "X-Pinecone-Api-Version: 2025-10" \
        -d '{
              "namespace": "example-namespace",
              "vectors": [
                  {
                      "id": "vec1",
                      "sparseValues": {
                          "values": [1.7958984, 0.41577148, 2.828125, 2.8027344, 2.8691406, 1.6533203, 5.3671875, 1.3046875, 0.49780273, 0.5722656, 2.71875, 3.0820312, 2.5019531, 4.4414062, 3.3554688],
                          "indices": [822745112, 1009084850, 1221765879, 1408993854, 1504846510, 1596856843, 1640781426, 1656251611, 1807131503, 2543655733, 2902766088, 2909307736, 3246437992, 3517203014, 3590924191]
                      },
                      "metadata": {
                          "chunk_text": "AAPL reported a year-over-year revenue increase, expecting stronger Q3 demand for its flagship phones.",
                          "category": "technology",
                          "quarter": "Q3"
                      }
                  },
                  {
                      "id": "vec2",
                      "sparseValues": {
                          "values": [0.4362793, 3.3457031, 2.7714844, 3.0273438, 3.3164062, 5.6015625, 2.4863281, 0.38134766, 1.25, 2.9609375, 0.34179688, 1.4306641, 0.34375, 3.3613281, 1.4404297, 2.2558594, 2.2597656, 4.8710938, 0.5605469],
                          "indices": [131900689, 592326839, 710158994, 838729363, 1304885087, 1640781426, 1690623792, 1807131503, 2066971792, 2428553208, 2548600401, 2577534050, 3162218338, 3319279674, 3343062801, 3476647774, 3485013322, 3517203014, 4283091697]
                      },
                      "metadata": {
                          "chunk_text": "Analysts suggest that AAPL'\''s upcoming Q4 product launch event might solidify its position in the premium smartphone market.",
                          "category": "technology",
                          "quarter": "Q4"
                      }
                  },
                  {
                      "id": "vec3",
                      "sparseValues": {
                          "values": [2.6875, 4.2929688, 3.609375, 3.0722656, 2.1152344, 5.78125, 3.7460938, 3.7363281, 1.2695312, 3.4824219, 0.7207031, 0.0826416, 4.671875, 3.7011719, 2.796875, 0.61621094],
                          "indices": [8661920, 350356213, 391213188, 554637446, 1024951234, 1640781426, 1780689102, 1799010313, 2194093370, 2632344667, 2641553256, 2779594451, 3517203014, 3543799498, 3837503950, 4283091697]
                      },
                      "metadata": {
                          "chunk_text": "AAPL'\''s strategic Q3 partnerships with semiconductor suppliers could mitigate component risks and stabilize iPhone production",
                          "category": "technology",
                          "quarter": "Q3"
                      }
                  },
                  {
                      "id": "vec4",
                      "sparseValues": {
                          "values": [0.73046875, 0.46972656, 2.84375, 5.2265625, 3.3242188, 1.9863281, 0.9511719, 0.5019531, 4.4257812, 3.4277344, 0.41308594, 4.3242188, 2.4179688, 3.1757812, 1.0224609, 2.0585938, 2.5859375],
                          "indices": [131900689, 152217691, 441495248, 1640781426, 1851149807, 2263326288, 2502307765, 2641553256, 2684780967, 2966813704, 3162218338, 3283104238, 3488055477, 3530642888, 3888762515, 4152503047, 4177290673]
                      },
                      "metadata": {
                          "chunk_text": "AAPL may consider healthcare integrations in Q4 to compete with tech rivals entering the consumer wellness space.",
                          "category": "technology",
                          "quarter": "Q4"
                      }
                  },
              ]
          }'
      ```
    </CodeGroup>
  </Tab>
</Tabs>

## Upsert in batches

<Tip>
  To control costs when ingesting large datasets (10,000,000+ records), use [import](/guides/index-data/import-data) instead of upsert.
</Tip>

Send upserts in batches to help increase throughput.

* When upserting records with vectors, a batch should be as large as possible (up to 1000 records) without exceeding the [max request size of 2 MB](#upsert-limits).

  To understand the number of records you can fit into one batch based on the vector dimensions and metadata size, see the following table:

  | Dimension | Metadata (bytes) | Max batch size |
  | :-------- | :--------------- | :------------- |
  | 386       | 0                | 1000           |
  | 768       | 500              | 559            |
  | 1536      | 2000             | 245            |

* When upserting records with text, a batch can contain up to 96 records. This limit comes from the [hosted embedding models](/guides/index-data/create-an-index#embedding-models) used during integrated embedding rather than the batch size limit for upserting raw vectors.

<CodeGroup>
  ```Python Python theme={null}
  import random
  import itertools
  from pinecone.grpc import PineconeGRPC as Pinecone

  pc = Pinecone(api_key="YOUR_API_KEY")

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  def chunks(iterable, batch_size=200):
      """A helper function to break an iterable into chunks of size batch_size."""
      it = iter(iterable)
      chunk = tuple(itertools.islice(it, batch_size))
      while chunk:
          yield chunk
          chunk = tuple(itertools.islice(it, batch_size))

  vector_dim = 128
  vector_count = 10000

  # Example generator that generates many (id, vector) pairs
  example_data_generator = map(lambda i: (f'id-{i}', [random.random() for _ in range(vector_dim)]), range(vector_count))

  # Upsert data with 200 vectors per upsert request
  for ids_vectors_chunk in chunks(example_data_generator, batch_size=200):
      index.upsert(vectors=ids_vectors_chunk) 
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from "@pinecone-database/pinecone";

  const RECORD_COUNT = 10000;
  const RECORD_DIMENSION = 128;

  const client = new Pinecone({ apiKey: "YOUR_API_KEY" });
  const index = client.index("docs-example");

  // A helper function that breaks an array into chunks of size batchSize
  const chunks = (array, batchSize = 200) => {
    const chunks = [];

    for (let i = 0; i < array.length; i += batchSize) {
      chunks.push(array.slice(i, i + batchSize));
    }

    return chunks;
  };

  // Example data generation function, creates many (id, vector) pairs
  const generateExampleData = () =>
    Array.from({ length: RECORD_COUNT }, (_, i) => {
      return {
        id: `id-${i}`,
        values: Array.from({ length: RECORD_DIMENSION }, (_, i) => Math.random()),
      };
    });

  const exampleRecordData = generateExampleData();
  const recordChunks = chunks(exampleRecordData);

  // Upsert data with 200 records per upsert request
  for (const chunk of recordChunks) {
    await index.upsert(chunk)
  }
  ```

  ```java Java theme={null}
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import io.pinecone.unsigned_indices_model.VectorWithUnsignedIndices;

  import java.util.Arrays;
  import java.util.List;

  public class UpsertBatchExample  {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");

          ArrayList<VectorWithUnsignedIndices> vectors = generateVectors();
          ArrayList<ArrayList<VectorWithUnsignedIndices>> chunks = chunks(vectors, BATCH_SIZE);

          for (ArrayList<VectorWithUnsignedIndices> chunk : chunks) {
              index.upsert(chunk, "example-namespace");
          }
      }

      // A helper function that breaks an ArrayList into chunks of batchSize
      private static ArrayList<ArrayList<VectorWithUnsignedIndices>> chunks(ArrayList<VectorWithUnsignedIndices> vectors, int batchSize) {
          ArrayList<ArrayList<VectorWithUnsignedIndices>> chunks = new ArrayList<>();
          ArrayList<VectorWithUnsignedIndices> chunk = new ArrayList<>();

          for (int i = 0; i < vectors.size(); i++) {
              if (i % BATCH_SIZE == 0 && i != 0) {
                  chunks.add(chunk);
                  chunk = new ArrayList<>();
              }

              chunk.add(vectors.get(i));
          }

          return chunks;
      }

      // Example data generation function, creates many (id, vector) pairs
      private static ArrayList<VectorWithUnsignedIndices> generateVectors() {
          Random random = new Random();
          ArrayList<VectorWithUnsignedIndices> vectors = new ArrayList<>();


          for (int i = 0; i <= RECORD_COUNT; i++) {
              String id = "id-" + i;
              ArrayList<Float> values = new ArrayList<>();

              for (int j = 0; j < RECORD_DIMENSION; j++) {
                  values.add(random.nextFloat());
              }

              VectorWithUnsignedIndices vector = new VectorWithUnsignedIndices();
              vector.setId(id);
              vector.setValues(values);
              vectors.add(vector);
          }

          return vectors;
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"

      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
  	  }

      // Generate a large number of vectors to upsert
      vectorCount := 10000
      vectorDim := idx.Dimension

      vectors := make([]*pinecone.Vector, vectorCount)
      for i := 0; i < int(vectorCount); i++ {
          randomFloats := make([]float32, vectorDim)

          for i := int32(0); i < vectorDim; i++ {
              randomFloats[i] = rand.Float32()
          }

          vectors[i] = &pinecone.Vector{
              Id:     fmt.Sprintf("doc1#-vector%d", i),
              Values: randomFloats,
          }
      }

      // Break the vectors into batches of 200
      var batches [][]*pinecone.Vector
      batchSize := 200

      for len(vectors) > 0 {
          batchEnd := batchSize
          if len(vectors) < batchSize {
              batchEnd = len(vectors)
          }
          batches = append(batches, vectors[:batchEnd])
          vectors = vectors[batchEnd:]
      }

      // Upsert batches
      for i, batch := range batches {
          upsertResp, err := idxConn.UpsertVectors(context.Background(), batch)
          if err != nil {
              panic(err)
          }

          fmt.Printf("upserted %d vectors (%v of %v batches)\n", upsertResp, i+1, len(batches))
      }
  }
  ```
</CodeGroup>

## Upsert in parallel

<Tip>
  Python SDK v6.0.0 and later provide `async` methods for use with [asyncio](https://docs.python.org/3/library/asyncio.html). Asyncio support makes it possible to use Pinecone with modern async web frameworks such as FastAPI, Quart, and Sanic. For more details, see [Async requests](/reference/sdks/python/overview#async-requests).
</Tip>

Send multiple upserts in parallel to help increase throughput. Vector operations block until the response has been received. However, they can be made asynchronously as follows:

<CodeGroup>
  ```Python Python theme={null}
  # This example uses `async_req=True` and multiple threads.
  # For a single-threaded approach compatible with modern async web frameworks, 
  # see https://docs.pinecone.io/reference/sdks/python/overview#async-requests
  import random
  import itertools
  from pinecone import Pinecone

  # Initialize the client with pool_threads=30. This limits simultaneous requests to 30.
  pc = Pinecone(api_key="YOUR_API_KEY", pool_threads=30)

  # To get the unique host for an index, 
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  index = pc.Index(host="INDEX_HOST")

  def chunks(iterable, batch_size=200):
      """A helper function to break an iterable into chunks of size batch_size."""
      it = iter(iterable)
      chunk = tuple(itertools.islice(it, batch_size))
      while chunk:
          yield chunk
          chunk = tuple(itertools.islice(it, batch_size))

  vector_dim = 128
  vector_count = 10000

  example_data_generator = map(lambda i: (f'id-{i}', [random.random() for _ in range(vector_dim)]), range(vector_count))

  # Upsert data with 200 vectors per upsert request asynchronously
  # - Pass async_req=True to index.upsert()
  with pc.Index(host="INDEX_HOST", pool_threads=30) as index:
      # Send requests in parallel
      async_results = [
          index.upsert(vectors=ids_vectors_chunk, async_req=True)
          for ids_vectors_chunk in chunks(example_data_generator, batch_size=200)
      ]
      # Wait for and retrieve responses (this raises in case of error)
      [async_result.get() for async_result in async_results]
  ```

  ```JavaScript JavaScript theme={null}
  import { Pinecone } from "@pinecone-database/pinecone";

  const RECORD_COUNT = 10000;
  const RECORD_DIMENSION = 128;

  const client = new Pinecone({ apiKey: "YOUR_API_KEY" });

  // To get the unique host for an index, 
  // see https://docs.pinecone.io/guides/manage-data/target-an-index
  const index = pc.index("INDEX_NAME", "INDEX_HOST")

  // A helper function that breaks an array into chunks of size batchSize
  const chunks = (array, batchSize = 200) => {
    const chunks = [];

    for (let i = 0; i < array.length; i += batchSize) {
      chunks.push(array.slice(i, i + batchSize));
    }

    return chunks;
  };

  // Example data generation function, creates many (id, vector) pairs
  const generateExampleData = () =>
    Array.from({ length: RECORD_COUNT }, (_, i) => {
      return {
        id: `id-${i}`,
        values: Array.from({ length: RECORD_DIMENSION }, (_, i) => Math.random()),
      };
    });

  const exampleRecordData = generateExampleData();
  const recordChunks = chunks(exampleRecordData);

  // Upsert data with 200 records per request asynchronously using Promise.all()
  await Promise.all(recordChunks.map((chunk) => index.upsert(chunk)));
  ```

  ```java Java theme={null}
  import com.google.protobuf.Struct;
  import com.google.protobuf.Value;
  import io.pinecone.clients.Index;
  import io.pinecone.configs.PineconeConfig;
  import io.pinecone.configs.PineconeConnection;
  import io.pinecone.proto.UpsertResponse;
  import io.pinecone.unsigned_indices_model.VectorWithUnsignedIndices;

  import java.util.ArrayList;
  import java.util.Arrays;
  import java.util.concurrent.ExecutorService;
  import java.util.concurrent.Executors;
  import java.util.List;

  public class UpsertExample {
      public static void main(String[] args) {
          PineconeConfig config = new PineconeConfig("YOUR_API_KEY");
          // To get the unique host for an index, 
          // see https://docs.pinecone.io/guides/manage-data/target-an-index
          config.setHost("INDEX_HOST");
          PineconeConnection connection = new PineconeConnection(config);
          Index index = new Index(connection, "INDEX_NAME");

          // Run 5 threads concurrently and upsert data into pinecone
          int numberOfThreads = 5;

          // Create a fixed thread pool
          ExecutorService executor = Executors.newFixedThreadPool(numberOfThreads);

          // Submit tasks to the executor
          for (int i = 0; i < numberOfThreads; i++) {
              // upsertData
              int batchNumber = i+1;
              executor.submit(() -> upsertData(index, batchNumber));
          }

          // Shutdown the executor
          executor.shutdown();
      }

      private static void upsertData(Index index, int batchNumber) {
          // Vector ids to be upserted
          String prefix = "v" + batchNumber;
          List<String> upsertIds = Arrays.asList(prefix + "_1", prefix + "_2", prefix + "_3");

          // List of values to be upserted
          List<List<Float>> values = new ArrayList<>();
          values.add(Arrays.asList(1.0f, 2.0f, 3.0f));
          values.add(Arrays.asList(4.0f, 5.0f, 6.0f));
          values.add(Arrays.asList(7.0f, 8.0f, 9.0f));

          // List of sparse indices to be upserted
          List<List<Long>> sparseIndices = new ArrayList<>();
          sparseIndices.add(Arrays.asList(1L, 2L, 3L));
          sparseIndices.add(Arrays.asList(4L, 5L, 6L));
          sparseIndices.add(Arrays.asList(7L, 8L, 9L));

          // List of sparse values to be upserted
          List<List<Float>> sparseValues = new ArrayList<>();
          sparseValues.add(Arrays.asList(1000f, 2000f, 3000f));
          sparseValues.add(Arrays.asList(4000f, 5000f, 6000f));
          sparseValues.add(Arrays.asList(7000f, 8000f, 9000f));

          List<VectorWithUnsignedIndices> vectors = new ArrayList<>(3);

          // Metadata to be upserted
          Struct metadataStruct1 = Struct.newBuilder()
                  .putFields("genre", Value.newBuilder().setStringValue("action").build())
                  .putFields("year", Value.newBuilder().setNumberValue(2019).build())
                  .build();

          Struct metadataStruct2 = Struct.newBuilder()
                  .putFields("genre", Value.newBuilder().setStringValue("thriller").build())
                  .putFields("year", Value.newBuilder().setNumberValue(2020).build())
                  .build();

          Struct metadataStruct3 = Struct.newBuilder()
                  .putFields("genre", Value.newBuilder().setStringValue("comedy").build())
                  .putFields("year", Value.newBuilder().setNumberValue(2021).build())
                  .build();
          List<Struct> metadataStructList = Arrays.asList(metadataStruct1, metadataStruct2, metadataStruct3);

          // Upsert data
          for (int i = 0; i < metadataStructList.size(); i++) {
              vectors.add(buildUpsertVectorWithUnsignedIndices(upsertIds.get(i), values.get(i), sparseIndices.get(i), sparseValues.get(i), metadataStructList.get(i)));
          }

          UpsertResponse upsertResponse = index.upsert(vectors, "example-namespace");
      }
  }
  ```

  ```go Go theme={null}
  package main

  import (
      "context"
      "fmt"
      "log"
      "math/rand"
      "sync"
      
      "github.com/pinecone-io/go-pinecone/v4/pinecone"
  )

  func main() {
      ctx := context.Background()

      pc, err := pinecone.NewClient(pinecone.NewClientParams{
          ApiKey: "YOUR_API_KEY",
      })
      if err != nil {
          log.Fatalf("Failed to create Client: %v", err)
      }

      // To get the unique host for an index, 
      // see https://docs.pinecone.io/guides/manage-data/target-an-index
      idxConn, err := pc.Index(pinecone.NewIndexConnParams{Host: "INDEX_HOST"})
      if err != nil {
          log.Fatalf("Failed to create IndexConnection for Host: %v", err)
  	  }

      // Generate a large number of vectors to upsert
      vectorCount := 10000
      vectorDim := idx.Dimension

      vectors := make([]*pinecone.Vector, vectorCount)
      for i := 0; i < int(vectorCount); i++ {
          randomFloats := make([]float32, vectorDim)

          for i := int32(0); i < vectorDim; i++ {
              randomFloats[i] = rand.Float32()
          }

          vectors[i] = &pinecone.Vector{
              Id:     fmt.Sprintf("doc1#-vector%d", i),
              Values: randomFloats,
          }
      }

      // Break the vectors into batches of 200
      var batches [][]*pinecone.Vector
      batchSize := 200

      for len(vectors) > 0 {
          batchEnd := batchSize
          if len(vectors) < batchSize {
              batchEnd = len(vectors)
          }
          batches = append(batches, vectors[:batchEnd])
          vectors = vectors[batchEnd:]
      }

      // Use channels to manage concurrency and possible errors
      maxConcurrency := 10
      errChan := make(chan error, len(batches))
      semaphore := make(chan struct{}, maxConcurrency)
      var wg sync.WaitGroup

      for i, batch := range batches {
          wg.Add(1)
          semaphore <- struct{}{}

          go func(batch []*pinecone.Vector, i int) {
              defer wg.Done()
              defer func() { <-semaphore }()

              upsertResp, err := idxConn.UpsertVectors(context.Background(), batch)
              if err != nil {
                  errChan <- fmt.Errorf("batch %d failed: %v", i, err)
                  return
              }

              fmt.Printf("upserted %d vectors (%v of %v batches)\n", upsertResp, i+1, len(batches))
          }(batch, i)
      }

      wg.Wait()
      close(errChan)

      for err := range errChan {
          if err != nil {
              fmt.Printf("Error while upserting batch: %v\n", err)
          }
      }
  }
  ```
</CodeGroup>

### Python SDK with gRPC

Using the Python SDK with gRPC extras can provide higher upsert speeds. Through multiplexing, gRPC is able to handle large amounts of requests in parallel without slowing down the rest of the system (HoL blocking), unlike REST. Moreover, you can pass various retry strategies to the gRPC SDK, including [exponential backoff](/guides/production/error-handling#implement-retry-logic).

To install the gRPC version of the SDK:

```Shell Shell theme={null}
pip install "pinecone[grpc]"
```

To use the gRPC SDK, import the `pinecone.grpc` subpackage and target an index as usual:

```Python Python theme={null}
from pinecone.grpc import PineconeGRPC as Pinecone

# This is gRPC client aliased as "Pinecone"
pc = Pinecone(api_key='YOUR_API_KEY')  

# To get the unique host for an index, 
# see https://docs.pinecone.io/guides/manage-data/target-an-index
index = pc.Index(host="INDEX_HOST")
```

To launch multiple read and write requests in parallel, pass `async_req` to the `upsert` operation:

```Python Python theme={null}
def chunker(seq, batch_size):
  return (seq[pos:pos + batch_size] for pos in range(0, len(seq), batch_size))

async_results = [
  index.upsert(vectors=chunk, async_req=True)
  for chunk in chunker(data, batch_size=200)
]

# Wait for and retrieve responses (in case of error)
[async_result.result() for async_result in async_results]
```

<Note>
  It is possible to get write-throttled faster when upserting using the gRPC SDK. If you see this often, [implement retry logic with exponential backoff](/guides/production/error-handling#implement-retry-logic) while upserting.

  The syntax for upsert, query, fetch, and delete with the gRPC SDK remain the same as the standard SDK.
</Note>

## Upsert limits

| Metric                                                             | Limit                                                         |
| :----------------------------------------------------------------- | :------------------------------------------------------------ |
| Max [batch size](/guides/index-data/upsert-data#upsert-in-batches) | 2 MB or 1000 records with vectors <br /> 96 records with text |
| Max metadata size per record                                       | 40 KB                                                         |
| Max length for a record ID                                         | 512 characters                                                |
| Max dimensionality for dense vectors                               | 20,000                                                        |
| Max non-zero values for sparse vectors                             | 2048                                                          |
| Max dimensionality for sparse vectors                              | 4.2 billion                                                   |

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/upsert.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Upsert vectors

> Upsert vectors into a namespace. If a new value is upserted for an existing vector ID, it will overwrite the previous value.

For guidance, examples, and limits, see [Upsert data](https://docs.pinecone.io/guides/index-data/upsert-data).

<Tip>
  To control costs when ingesting large datasets (10,000,000+ records), use [import](/guides/index-data/import-data) instead of upsert.
</Tip>

<RequestExample>
  ```shell curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  PINECONE_API_KEY="YOUR_API_KEY"
  INDEX_HOST="INDEX_HOST"

  curl "https://$INDEX_HOST/vectors/upsert" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H 'Content-Type: application/json' \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{
      "vectors": [
        {
          "id": "vec1",
          "values": [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1],
          "metadata": {"genre": "comedy", "year": 2020}
        },
        {
          "id": "vec2",
          "values": [0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2],
          "metadata": {"genre": "documentary", "year": 2019}
        }
      ],
      "namespace": "example-namespace"
    }'
  ```
</RequestExample>

<ResponseExample />

```
```


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml post /vectors/upsert
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /vectors/upsert:
    post:
      tags:
        - Vector Operations
      summary: Upsert vectors
      description: >-
        Upsert vectors into a namespace. If a new value is upserted for an
        existing vector ID, it will overwrite the previous value.


        For guidance, examples, and limits, see [Upsert
        data](https://docs.pinecone.io/guides/index-data/upsert-data).
      operationId: upsertVectors
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/UpsertRequest'
        required: true
      responses:
        '200':
          description: A successful response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/UpsertResponse'
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    UpsertRequest:
      description: The request for the `upsert` operation.
      type: object
      properties:
        vectors:
          description: >-
            An array containing the vectors to upsert. Recommended batch limit
            is up to 1000 vectors.
          type: array
          items:
            $ref: '#/components/schemas/Vector'
          minLength: 1
          maxLength: 1000
        namespace:
          example: example-namespace
          description: The namespace where you upsert vectors.
          type: string
      required:
        - vectors
    UpsertResponse:
      description: The response for the `upsert` operation.
      type: object
      properties:
        upsertedCount:
          example: 2
          description: The number of vectors upserted.
          type: integer
          format: int64
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    Vector:
      type: object
      properties:
        id:
          example: example-vector-1
          description: This is the vector's unique id.
          type: string
          required:
            - id
          minLength: 1
          maxLength: 512
        values:
          example:
            - 0.1
            - 0.2
            - 0.3
            - 0.4
            - 0.5
            - 0.6
            - 0.7
            - 0.8
          description: This is the vector data included in the request.
          type: array
          required:
            - values
          items:
            type: number
            format: float
          minLength: 1
          maxLength: 20000
        sparseValues:
          $ref: '#/components/schemas/SparseValues'
        metadata:
          example:
            genre: documentary
            year: 2019
          description: This is the metadata included in the request.
          type: object
      required:
        - id
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
    SparseValues:
      description: >-
        Vector sparse data. Represented as a list of indices and a list of 
        corresponded values, which must be with the same length.
      type: object
      properties:
        indices:
          example:
            - 1
            - 312
            - 822
            - 14
            - 980
          description: The indices of the sparse data.
          type: array
          required:
            - indices
          items:
            type: integer
            format: int64
          minLength: 1
          maxLength: 1000
        values:
          example:
            - 0.1
            - 0.2
            - 0.3
            - 0.4
            - 0.5
          description: >-
            The corresponding values of the sparse data, which must be with the
            same length as the indices.
          type: array
          required:
            - values
          items:
            type: number
            format: float
          minLength: 1
          maxLength: 1000
      required:
        - indices
        - values
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/reference/api/2025-10/data-plane/upsert_records.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Upsert text

> Upsert text into a namespace. Pinecone converts the text to vectors automatically using the hosted embedding model associated with the index.

Upserting text is supported only for [indexes with integrated embedding](https://docs.pinecone.io/reference/api/2025-01/control-plane/create_for_model).

For guidance, examples, and limits, see [Upsert data](https://docs.pinecone.io/guides/index-data/upsert-data).

<RequestExample>
  ```shell curl theme={null}
  # To get the unique host for an index,
  # see https://docs.pinecone.io/guides/manage-data/target-an-index
  INDEX_HOST="INDEX_HOST"
  NAMESPACE="YOUR_NAMESPACE"
  PINECONE_API_KEY="YOUR_API_KEY"

  # Upsert records into a namespace
  # `chunk_text` fields are converted to dense vectors
  # `category` fields are stored as metadata
  curl "https://$INDEX_HOST/records/namespaces/$NAMESPACE/upsert" \
    -H "Content-Type: application/x-ndjson" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10" \
    -d '{"_id": "rec1", "chunk_text": "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut.", "category": "digestive system"}
        {"_id": "rec2", "chunk_text": "Apples originated in Central Asia and have been cultivated for thousands of years, with over 7,500 varieties available today.", "category": "cultivation"}
        {"_id": "rec3", "chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases.", "category": "immune system"}
        {"_id": "rec4", "chunk_text": "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes.", "category": "endocrine system"}'
  ```
</RequestExample>

<ResponseExample />


## OpenAPI

````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-10/db_data_2025-10.oas.yaml post /records/namespaces/{namespace}/upsert
openapi: 3.0.3
info:
  title: Pinecone Data Plane API
  description: >-
    Pinecone is a vector database that makes it easy to search and retrieve
    billions of high-dimensional vectors.
  contact:
    name: Pinecone Support
    url: https://support.pinecone.io
    email: support@pinecone.io
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
  version: 2025-10
servers:
  - url: https://{index_host}
    variables:
      index_host:
        default: unknown
        description: host of the index
security:
  - ApiKeyAuth: []
tags:
  - name: Vector Operations
  - name: Bulk Operations
  - name: Namespace Operations
externalDocs:
  description: More Pinecone.io API docs
  url: https://docs.pinecone.io/introduction
paths:
  /records/namespaces/{namespace}/upsert:
    post:
      tags:
        - Vector Operations
      summary: Upsert text
      description: >-
        Upsert text into a namespace. Pinecone converts the text to vectors
        automatically using the hosted embedding model associated with the
        index.


        Upserting text is supported only for [indexes with integrated
        embedding](https://docs.pinecone.io/reference/api/2025-01/control-plane/create_for_model).


        For guidance, examples, and limits, see [Upsert
        data](https://docs.pinecone.io/guides/index-data/upsert-data).
      operationId: upsertRecordsNamespace
      parameters:
        - in: header
          name: X-Pinecone-Api-Version
          description: Required date-based version header
          required: true
          schema:
            default: 2025-10
            type: string
          style: simple
        - in: path
          name: namespace
          description: The namespace to upsert records into.
          required: true
          schema:
            type: string
          style: simple
      requestBody:
        description: >
          Each record in the request body must include an `_id` field and a
          field that matches your index's `field_map` configuration (such as
          `chunk_text` or `data`). All other fields are stored as metadata.
        content:
          application/x-ndjson:
            schema:
              type: array
              items:
                $ref: '#/components/schemas/UpsertRecord'
        required: true
      responses:
        '201':
          description: A successful response.
        '400':
          description: Bad request. The request body included invalid request parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        4XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
        5XX:
          description: An unexpected error response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/rpcStatus'
components:
  schemas:
    UpsertRecord:
      example:
        _id: example-record-1
      description: The request for the `upsert` operation.
      type: object
      properties:
        _id:
          description: >-
            The unique ID of the record to upsert. Note that `id` can be used as
            an alias for `_id`.
          type: string
      required:
        - _id
    rpcStatus:
      type: object
      properties:
        code:
          type: integer
          format: int32
        message:
          type: string
        details:
          type: array
          items:
            $ref: '#/components/schemas/protobufAny'
    protobufAny:
      type: object
      properties:
        typeUrl:
          type: string
        value:
          type: string
          format: byte
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: Api-Key
      description: >-
        An API Key is required to call Pinecone APIs. Get yours from the
        [console](https://app.pinecone.io/).

````

---

# Source: https://docs.pinecone.io/integrations/vercel.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Vercel

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

Vercel is a platform for developers that provides the tools, workflows, and infrastructure you need to build and deploy your web apps faster, without the need for additional configuration. Vercel supports popular frontend frameworks out-of-the-box, and its scalable, secure infrastructure is globally distributed to serve content from data centers near your users for optimal speeds.

Pinecone provides the long-term memory for your Vercel AI projects. Using Pinecone with Vercel enables you to quickly set up and authenticate a connection to your Pinecone data/indexes, and then easily scale to support billions of data points. The integration is designed to be self-serve with strong defaults for a smooth setup, with optional advanced settings.

<PrimarySecondaryCTA primaryLabel={"Get started"} primaryHref={"https://vercel.com/integrations/pinecone"} />

---

# Source: https://docs.pinecone.io/reference/api/versioning.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# API versioning

Pinecone's APIs are versioned to ensure that your applications continue to work as expected as the platform evolves. Versions are named by release date in the format `YYYY-MM`, for example, `2025-10`.

## Release schedule

On a quarterly basis, Pinecone releases a new **stable** API version as well as a **release candidate** of the next stable version.

* **Stable:** Each stable version remains unchanged and supported for a minimum of 12 months. Since stable versions are released every 3 months, this means you have at least 9 months to test and migrate your app to the newest stable version before support for the previous version is removed.

* **Release candidate:** The release candidate gives you insight into the upcoming changes in the next stable version. It is available for approximately 3 months before the release of the stable version and can include new features, improvements, and [breaking changes](#breaking-changes).

Below is an example of Pinecone's release schedule:

<img className="block max-w-full dark:hidden" noZoom src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=be4366853edee83e003c3863e31fa1ce" data-og-width="2120" width="2120" data-og-height="960" height="960" data-path="images/api-versioning.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=66d593bf5bc7c842b28392daebb93c4a 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=c241a0f2c0e6e3ec3b1415e10a8516c8 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=0b88fe499a3c01dbd424ee943e9ab0c9 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=fcdffc7f6d2b907681018e42cdbe0920 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=bc180b4a9a96edc502c5b49470ca3881 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=70bbd09f29afc200f6c9111495cd5f05 2500w" />

<img className="hidden max-w-full dark:block" noZoom src="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning-dark.png?fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=47a32d4f515a118a7500a0e2aec3beed" data-og-width="2120" width="2120" data-og-height="960" height="960" data-path="images/api-versioning-dark.png" data-optimize="true" data-opv="3" srcset="https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning-dark.png?w=280&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=deb21035b92a663a67b5cfe9420d999d 280w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning-dark.png?w=560&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=8519ab97f6f889abeaab6105ea681e31 560w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning-dark.png?w=840&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=dd4ef1e1e99cb42da31c83a7cd0d5e5d 840w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning-dark.png?w=1100&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=927c7d86e73e85f417c832a96e98fdbb 1100w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning-dark.png?w=1650&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=eb0c5f538f0b3422c475280f09476a62 1650w, https://mintcdn.com/pinecone/9qvQp5nU6duJvusQ/images/api-versioning-dark.png?w=2500&fit=max&auto=format&n=9qvQp5nU6duJvusQ&q=85&s=eacb669a6469064133034bdc1e30add1 2500w" />

## Specify an API version

<Warning>
  When using the API directly, it is important to specify an API version in your requests. If you don't, requests default to the oldest supported stable version. Once support for that version ends, your requests will default to the next oldest stable version, which could include breaking changes that require you to update your integration.
</Warning>

To specify an API version, set the `X-Pinecone-Api-Version` header to the version name.

For example, based on the version support diagram above, if it is currently October 2025 and you want to use the latest stable version to describe an index, you would set `"X-Pinecone-Api-Version: 2025-10"`:

```shell curl theme={null}
PINECONE_API_KEY="YOUR_API_KEY"

curl -i -X GET "https://api.pinecone.io/indexes/movie-recommendations" \
    -H "Api-Key: $PINECONE_API_KEY" \
    -H "X-Pinecone-Api-Version: 2025-10"
```

To use an older version, specify that version instead.

## SDK versions

Official [Pinecone SDKs](/reference/pinecone-sdks) provide convenient access to Pinecone APIs. SDK versions are pinned to specific API versions. When a new API version is released, a new version of the SDK is also released.

For the mapping between SDK and API versions, see [SDK versions](/reference/pinecone-sdks#sdk-versions).

## Breaking changes

Breaking changes are changes that can potentially break your integration with a Pinecone API. Breaking changes include:

* Removing an entire operation
* Removing or renaming a parameter
* Removing or renaming a response field
* Adding a new required parameter
* Making a previously optional parameter required
* Changing the type of a parameter or response field
* Removing enum values
* Adding a new validation rule to an existing parameter
* Changing authentication or authorization requirements

## Non-breaking changes

Non-breaking changes are additive and should not break your integration. Additive changes include:

* Adding an operation
* Adding an optional parameter
* Adding an optional request header
* Adding a response field
* Adding a response header
* Adding enum values

## Get updates

To ensure you always know about upcoming API changes, follow the [Release notes](/release-notes/).

---

# Source: https://docs.pinecone.io/integrations/voltagent.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# VoltAgent

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

[VoltAgent](https://voltagent.dev) is a TypeScript-based, AI-agent framework for building production-ready applications with retrieval-augmented generation (RAG) capabilities. It supports two retrieval patterns: automatic search on every interaction, or LLM-decides-when-to-search, with built-in observability and tracking.

This integration connects VoltAgent agents to Pinecone's managed vector database, automatically generating embeddings with OpenAI's API. It provides semantic search with similarity scoring, source tracking, metadata filtering, and namespace organization, and it supports serverless deployment with automatic scaling.

The integration provides:

* A complete RAG setup with sample data
* Two pre-configured agent types
* Automatic index creation and population
* Source references and similarity scores
* Production-ready architecture

Use this integration to quickly build AI agents that can intelligently search and retrieve information from Pinecone vector databases, while maintaining full observability and control over the retrieval process.

<PrimarySecondaryCTA primaryHref={"https://voltagent.dev/docs/rag/pinecone/"} primaryLabel={"Get started"} />

---

# Source: https://docs.pinecone.io/integrations/voyage.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Voyage AI

> Using Voyage AI and Pinecone to generate and index high-quality vector embeddings

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

[Voyage AI](https://www.voyageai.com) provides cutting-edge embedding and rerankers. Voyage AI's generalist [embedding models](https://docs.voyageai.com/docs/embeddings) continually top the [MTEB leaderboard](https://huggingface.co/spaces/mteb/leaderboard), and the [domain-specific embeddings](https://blog.voyageai.com/2024/01/23/voyage-code-2-elevate-your-code-retrieval/) enhance the retrieval quality for enterprise use cases significantly.

<PrimarySecondaryCTA secondaryHref={"#setup-guide"} secondaryLabel={"View setup guide"} />

## Setup guide

In this guide, we use the [Voyage Embedding API endpoint](https://docs.voyageai.com/docs/embeddings) to generate text embeddings for terms of service and consumer contract documents, and then index those embeddings in the Pinecone vector database.

This is a powerful and common combination for building retrieval-augmented generation (RAG), semantic search, question-answering, code assistants, and other applications that rely on NLP and search over a large corpus of text data.

### 1. Set up the environment

Start by installing the Voyage and Pinecone clients and HuggingFace *Datasets* for downloading the *LegalBench: Consumer Contracts QA* ([`mteb/legalbench_consumer_contracts_qa`](https://huggingface.co/datasets/mteb/legalbench_consumer_contracts_qa)) dataset used in this guide:

```shell Shell theme={null}
pip install -U voyageai pinecone[grpc] datasets
```

### 2. Create embeddings

Sign up for an API key at [Voyage AI](https://dash.voyageai.com) and then use it to initialize your connection.

```Python Python theme={null}
import voyageai

vc = voyageai.Client(api_key="<YOUR_VOYAGE_API_KEY>")
```

Load the *LegalBench: Consumer Contracts QA*  dataset, which contains 154 consumer contract documents and 396 labeled queries about these documents.

```Python Python theme={null}
from datasets import load_dataset

# load the documents and queries of legalbench consumer contracts qa dataset
documents = load_dataset('mteb/legalbench_consumer_contracts_qa', 'corpus', cache_dir = './', split='corpus')
queries = load_dataset('mteb/legalbench_consumer_contracts_qa', 'queries', cache_dir = './', split='queries')

```

Each document in `mteb/legalbench_consumer_contracts_qa` contains a `text` field by which we will embed using the Voyage AI client.

```Python Python theme={null}
num_documents = len(documents['text'])
voyageai_batch_size = 128  # Please check the restrictions of number of examples and number of tokens per request here https://docs.voyageai.com/docs/embeddings
embeds = []
while len(embeds) < num_documents:
    embeds.extend(vc.embed(
        texts=documents['text'][len(embeds):len(embeds)+voyageai_batch_size],
        model='voyage-law-2',  # Please check the available models here https://docs.voyageai.com/docs/embeddings
        input_type='document',
        truncation=True
    ).embeddings)
```

Check the dimensionality of the returned vectors. You will need to save the embedding dimensionality from this to be used when initializing your Pinecone index later.

```Python Python theme={null}
import numpy as np

shape = np.array(embeds).shape
print(shape)
```

```
[Out]:
(154, 1024)
```

In this example, you can see that for each of the `154` documents, we created a `1024`-dimensional embedding with the Voyage AI `voyage-law-2` model.

### 3. Store the Embeddings

Now that you have your embeddings, you can move on to indexing them in the Pinecone vector database. For this, you need a Pinecone API key. [Sign up for one here](https://app.pinecone.io).

You first initialize our connection to Pinecone and then create a new index called `voyageai-pinecone-legalbench` for storing the embeddings. When creating the index, you specify that you would like to use the cosine similarity metric to align with Voyage AI's embeddings, and also pass the embedding dimensionality of `1024`.

```Python Python theme={null}
from pinecone.grpc import PineconeGRPC as Pinecone
from pinecone import ServerlessSpec

# initialize connection to pinecone (get API key at app.pinecone.io)
pc = Pinecone(api_key="<YOUR_PINECONE_API_KEY>")

index_name = 'voyageai-pinecone-legalbench'

# if the index does not exist, we create it
if not pc.has_index(index_name):
    pc.create_index(
        index_name,
        dimension=shape[1],
        spec=ServerlessSpec(
            cloud='aws',
            region='us-east-1'
        ),
        metric='cosine'
    )

# connect to index
index = pc.Index(index_name)
```

Now you can begin populating the index with your embeddings. Pinecone expects you to provide a list of tuples in the format (`id`, `vector`, `metadata`), where the `metadata` field is an optional extra field where you can store anything you want in a dictionary format. For this example, you will store the original text of the embeddings.

While uploading your data, you will batch everything to avoid pushing too much data in one go.

```Python Python theme={null}
batch_size = 128

ids = [str(i) for i in range(shape[0])]
# create list of metadata dictionaries
meta = [{'text': text} for text in documents['text']]

# create list of (id, vector, metadata) tuples to be upserted
to_upsert = list(zip(ids, embeds, meta))

for i in range(0, shape[0], batch_size):
    i_end = min(i+batch_size, shape[0])
    index.upsert(vectors=to_upsert[i:i_end])

# let's view the index statistics
print(index.describe_index_stats())
`

`[Out]:
{'dimension': 1024,
 'index_fullness': 0.0,
 'namespaces': {'': {'vector_count': 154}},
 'total_vector_count': 154}
```

You can see from `index.describe_index_stats` that you have a *1024-dimensionality* index populated with *154* embeddings. The `indexFullness` metric tells you how full your index is. At the moment, it is empty. Using the default value of one *p1* pod, you can fit around 750K embeddings before the `indexFullness` reaches capacity. The [Usage Estimator](https://www.pinecone.io/pricing/) can be used to identify the number of pods required for a given number of *n*-dimensional embeddings.

### 4. Semantic search

Now that you have your indexed vectors, you can perform a few search queries. When searching, you will first embed your query using `voyage-law-2`, and then search using the returned vector in Pinecone.

```Python Python theme={null}
# get a sample query from the dataset, "Will Google help me if I think someone has taken and used content Ive created without my permission?" 
query = queries['text'][0]
print(f"Query: {query}")

# create the query embedding
xq = vc.embed(
    texts=[query],
    model='voyage-law-2',
    input_type="query",
    truncation=True
).embeddings

# query, returning the top 3 most similar results
res = index.query(vector=xq, top_k=3, include_metadata=True)
```

The response from Pinecone includes your original text in the `metadata` field. Let's print out the `top_k` most similar questions and their respective similarity scores.

```Python Python theme={null}
for match in res['matches']:
    print(f"{match['score']:.2f}: {match['metadata']['text']}")
```

```[Out]:  theme={null}
0.59: Your content
Some of our services give you the opportunity to make your content publicly available  for example, you might post a product or restaurant review that you wrote, or you might upload a blog post that you created.
See the Permission to use your content section for more about your rights in your content, and how your content is used in our services
See the Removing your content section to learn why and how we might remove user-generated content from our services
If you think that someone is infringing your intellectual property rights, you can send us notice of the infringement and well take appropriate action. For example, we suspend or close the Google Accounts of repeat copyright infringers as described in our Copyright Help Centre.


0.47: Google content
Some of our services include content that belongs to Google  for example, many of the visual illustrations that you see in Google Maps. You may use Googles content as allowed by these terms and any service-specific additional terms, but we retain any intellectual property rights that we have in our content. Dont remove, obscure or alter any of our branding, logos or legal notices. If you want to use our branding or logos, please see the Google Brand Permissions page.

Other content
Finally, some of our services gives you access to content that belongs to other people or organisations  for example, a store owners description of their own business, or a newspaper article displayed in Google News. You may not use this content without that person or organisations permission, or as otherwise allowed by law. The views expressed in the content of other people or organisations are their own, and dont necessarily reflect Googles views.


0.45: Taking action in case of problems
Before taking action as described below, well provide you with advance notice when reasonably possible, describe the reason for our action and give you an opportunity to fix the problem, unless we reasonably believe that doing so would:
cause harm or liability to a user, third party or Google
violate the law or a legal enforcement authoritys order
compromise an investigation
compromise the operation, integrity or security of our services

Removing your content
If we reasonably believe that any of your content (1) breaches these terms, service-specific additional terms or policies, (2) violates applicable law, or (3) could harm our users, third parties or Google, then we reserve the right to take down some or all of that content in accordance with applicable law. Examples include child pornography, content that facilitates human trafficking or harassment, and content that infringes someone elses intellectual property rights.

Suspending or terminating your access to Google services
Google reserves the right to suspend or terminate your access to the services or delete your Google Account if any of these things happen:
you materially or repeatedly breach these terms, service-specific additional terms or policies
were required to do so to comply with a legal requirement or a court order
we reasonably believe that your conduct causes harm or liability to a user, third party or Google  for example, by hacking, phishing, harassing, spamming, misleading others or scraping content that doesnt belong to you
If you believe that your Google Account has been suspended or terminated in error, you can appeal.
Of course, youre always free to stop using our services at any time. If you do stop using a service, wed appreciate knowing why so that we can continue improving our services.

```

The semantic search pipeline with Voyage AI and Pinecone is able to identify the relevant consumer contract documents to answer the user query.

---

# Source: https://docs.pinecone.io/troubleshooting/wait-for-index-creation.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Wait for index creation to be complete

Pinecone index creation involves several different subsystems, including one which accepts the job of creating the index and one that actually performs the action. The Python SDK and the REST API are designed to interact with the first system during index creation but not the second.

This means that when a request call to [`create_index()`](/reference/api/latest/control-plane/create_index) is made, what's actually happening is that the job is being submitted to the queue to be completed. We do it this way for several reasons, including enforcing separation between the control and data planes.

If you need your application to wait for the index to be created before continuing to its next step, there is a way to ensure this happens, though. [`describe_index()`](/reference/api/latest/control-plane/describe_index) returns data about the state of the index, including whether it is ready to accept data. You simply call that method until it returns a 200 status code and the status object reports that the index is ready. Because the return is a tuple, we just have to access the slice containing the status object and check the boolean state of the ready variable. This is one possible method of doing so using the Python SDK:

```python  theme={null}
import pinecone
from time import sleep

def wait_on_index(index: str):
  """
  Takes the name of the index to wait for and blocks until it's available and ready.
  """
  ready = False
  while not ready:
    try:
      desc = pinecone.describe_index(index)
      if desc[7]['ready']:
        return True
    except pinecone.core.client.exceptions.NotFoundException:
      # NotFoundException means the index is created yet.
      pass
    sleep(5)
```

Calling `wait_on_index()` would then allow your application to only continue to upsert data once the index is fully online and available to accept data, avoiding potential 403 or 404 errors.

---

# Source: https://docs.pinecone.io/integrations/zapier.md

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Zapier

export const PrimarySecondaryCTA = ({primaryLabel, primaryHref, primaryTarget, secondaryLabel, secondaryHref, secondaryTarget}) => <div style={{
  display: 'flex',
  alignItems: 'center',
  gap: 16
}}>
   {primaryLabel && primaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  background: 'var(--brand-blue)',
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex'
}}>
      <a href={primaryHref} target={primaryTarget} style={{
  paddingLeft: 22,
  paddingRight: 22,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 4,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
        <div style={{
  textAlign: 'justify',
  color: 'var(--text-contrast)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
          {primaryLabel}
        </div>
        <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" style={{
  marginLeft: 2
}}>
          <path d="M9.70492 6L8.29492 7.41L12.8749 12L8.29492 16.59L9.70492 18L15.7049 12L9.70492 6Z" fill="white" style={{
  fille: "var(--text-contrast)"
}} />
        </svg>
      </a>
    </div>}

    {secondaryLabel && secondaryHref && <div style={{
  width: 'fit-content',
  height: 42,
  borderRadius: 4,
  overflow: 'hidden',
  flexDirection: 'column',
  justifyContent: 'center',
  alignItems: 'center',
  display: 'inline-flex',
  textDecoration: 'none'
}}>
        <a href={secondaryHref} target={secondaryTarget} style={{
  paddingLeft: 11,
  paddingRight: 11,
  paddingTop: 8,
  paddingBottom: 8,
  justifyContent: 'center',
  alignItems: 'center',
  gap: 8,
  display: 'inline-flex',
  textDecoration: 'none',
  borderBottom: 'none'
}}>
          <div style={{
  textAlign: 'justify',
  color: 'var(--brand-blue)',
  fontSize: 15,
  fontWeight: '600',
  letterSpacing: 0.46,
  wordWrap: 'break-word'
}}>
            {secondaryLabel}
          </div>
        </a>
      </div>}

  </div>;

[Zapier](https://zapier.com/) lets you connect Pinecone with thousands of the most popular apps, so you can automate your work and have more time for what matters most — no code required.

With this integration, Pinecone can trigger workflows when your index status changes, and other apps can command Pinecone to perform actions. This means you can automatically add or remove data, run searches, or manage indexes based on events happening in your other tools.

For example, you might use this integration to automatically add spreadsheet entries to your Pinecone index, or build a bot that searches for answers and posts them directly to your team chat. It's all about weaving Pinecone's powerful search and data capabilities into your existing workflows, making everything work together automatically.

<PrimarySecondaryCTA primaryHref={"https://zapier.com/apps/pinecone/integrations"} primaryLabel={"Get started"} secondaryHref={"https://help.zapier.com/hc/articles/38950923151117"} secondaryLabel={"View setup guide"} />