Skip to main content

Guides

Every guide in Pinecone Docs.

Get started

OverviewPinecone Nexus is the knowledge engine for agents. It compiles your data into queryable knowledge once, then serves grounded, cited answers to agents on every call.QuickstartDeploy Pinecone Nexus with BYOC, curate a context from your own documents, and run KnowQL queries that return grounded, cited multi-document answers.Key conceptsLearn the core Pinecone Nexus concepts: sources, workspaces, contexts, manifests, artifacts, tasks, sessions, and queries, and how they connect.OverviewOverview of Pinecone Assistant, a managed service for building production-grade RAG chat and agent applications grounded in your data.OverviewPinecone is the vector database for AI agents and applications, built for semantic search, knowledge retrieval, and long-term memory at scale.

Quickstart

OverviewGet started with Pinecone in minutes. Pick the path that matches what you're searching for and run your first search.Ingest your own filesTurn a folder of documents into a searchable index: extract text, chunk it, embed the chunks, and upsert.Bring your own vectorsCreate an index with a dense-vector field, upsert embeddings you already have, and run semantic search.Try full-text searchCreate an index, load documents, and run your first full-text (keyword) search in about five minutes.SDK quickstartQuickstart for Pinecone Assistant using the Python or Node.js SDK to create an assistant, upload documents, and chat with your data via RAG.n8n quickstartBuild an n8n workflow with Pinecone Assistant and OpenAI to download files via HTTP, upload documents, and chat with them from an automation.

Pricing and limitsUnderstand Pinecone Assistant pricing for ingestion units, chat tokens, and storage, plus plan-based service limits for file uploads and knowledge size.

The knowledge engine

How Nexus worksPinecone Nexus distills curated artifacts for retrieval, then answers with code over a retrieval SDK.

Context design

OverviewLearn how a Pinecone Nexus context's manifest turns your sources into queryable knowledge.Design your own manifestAuthor a Pinecone Nexus context's manifest through the API by defining custom artifact and edge types, then curating.Configure artifact formatsConfigure Pinecone Nexus artifact types as markdown prose or queryable SQLite tables through the manifest API.

CurationCuration is the write path that turns sources into chunks and artifacts incrementally, driven by the manifest.Nexus model guidanceLearn how Pinecone Nexus uses models for generation, curation, and retrieval, and how to choose the right one.

Queries and sessions

QueriesPinecone Nexus queries gather evidence and compose a grounded, cited answer.Query tracingInspect a Pinecone Nexus query trace to see reasoning steps, retrieval tool calls, token usage, latency, cost breakdown, and cache savings.

Tools

MCP serverQuery your Pinecone Nexus contexts from any Model Context Protocol client, such as Claude Desktop, over Streamable HTTP.

Bring Your Own Cloud

OverviewLearn what Nexus BYOC is, how it relates to Database BYOC, and its architecture.Residency and limitsSee where your data lives and travels in Nexus BYOC, plus its authentication, encryption, cluster footprint, and limitations.DeployInstall and operate Pinecone Nexus in your own cloud account.

Build an assistant

Create an assistantCreate a Pinecone Assistant with custom instructions, metadata, and region settings using the API, Python SDK, Node.js SDK, or console.Manage assistantsList, describe, update, and delete Pinecone assistants and check assistant status using the API, Python SDK, Node.js SDK, or Pinecone console.

Upload your data

OverviewOverview of Pinecone Assistant files: supported file types (PDF, DOCX, JSON, MD, TXT), metadata filters, storage, and signed URL access.Upload filesUpload local files to a Pinecone assistant with Python, JavaScript, or curl, track ingestion operations, and see how uploads are billed as ingestion units.Multimodal contextEnable multimodal PDF context in Pinecone Assistant to analyze charts, images, and diagrams using OCR and visual understanding for RAG.Manage filesList files in your Pinecone assistant, check individual file ingestion status by ID, view metadata, and delete files using the API, SDKs, or console.

Chat with an assistant

Standard interfaceChat with Pinecone Assistant through the standard interface with default, streaming, or JSON responses, plus citations and chat history support.OpenAI-compatible interfaceChat with Pinecone Assistant using the OpenAI-compatible Chat Completion API for inline citations, streaming responses, and easy integration.

Evaluate answers

OverviewOverview of Pinecone Assistant response evaluation: measure correctness, completeness, and alignment scores to benchmark RAG system quality.Evaluate answersEvaluate RAG system answers with Pinecone Assistant using correctness, completeness, and alignment metrics against a ground truth answer.

Retrieve context snippets

OverviewLearn how Pinecone Assistant retrieves context snippets with relevancy scores and references for RAG applications and agentic workflows.Retrieve context snippetsRetrieve context snippets and citations from a Pinecone Assistant to power your own LLM, RAG application, or agentic workflow with signed URLs.

Integrate with AI agents

Use an MCP serverConnect AI agents to a Pinecone Assistant MCP server via remote or local Model Context Protocol endpoints, including Cursor and Claude Desktop.

Core concepts

Key termsThe key terms and objects in Pinecone and how they relate to each other.ArchitectureLearn how Pinecone is built, from organizations, projects, indexes, and namespaces down to the infrastructure that stores and serves your data.

Index data

OverviewLearn how indexing works in Pinecone: serverless indexes, schemas, namespaces, integrated embedding, and metadata filtering.

Create and configure

Create an indexCreate a Pinecone serverless index for full-text (BM25), semantic (dense vector), sparse-vector, or hybrid search with a document schema.Configure metadata indexingLimit metadata indexing in Pinecone to the fields you filter on, to keep index building and query execution fast.

Data modelingModel your data in Pinecone using documents with dense_vector, sparse_vector, full-text string, and metadata fields for efficient retrieval.

Data ingestion

OverviewCompare data ingestion options in Pinecone: bulk import from object storage, upsert operations, and hosted embedding via the Inference API.Upsert recordsUpsert dense, sparse, and text records into Pinecone indexes, batch upserts for higher throughput, and partition data with namespaces.Import recordsImport large datasets efficiently from Amazon S3, Google Cloud Storage, or Azure Blob Storage into Pinecone serverless indexes using object storage.Migrate from pgvectorMigrate vector data from pgvector to Pinecone Database, validate search results, synchronize incremental changes, and cut over application traffic.Check data freshnessCheck data freshness in Pinecone serverless indexes using log sequence numbers (LSNs) and vector counts to verify recent upserts and deletes.

Multitenancy

DesignDesign multitenancy in Pinecone by choosing between namespace-per-tenant isolation and metadata filtering, with their cost and performance tradeoffs.ImplementImplement multitenancy in Pinecone with one namespace per tenant on a serverless index to isolate customer data for SaaS RAG or semantic search apps.

Dedicated read nodes

OverviewDedicated read nodes give a Pinecone index its own provisioned read hardware for predictable, low-latency performance at high query volumes.ConceptsNode types, shards, replicas, and index fullness are the building blocks of a Pinecone dedicated read nodes index.CreateCreate a Pinecone dedicated read nodes index from scratch or from a backup of an existing index.MigrateMigrate an existing on-demand or pod-based Pinecone index to dedicated read nodes.Size and testCalculate how many shards and replicas a Pinecone dedicated read nodes index needs, and load-test your workload to validate the configuration.Tune queriesTune scan_factor and max_candidates on a Pinecone dedicated read nodes index to trade recall for lower latency and higher throughput.ScaleScale a Pinecone dedicated read nodes index by adding replicas for query throughput and shards for storage capacity.ManageAdd a hosted embedding model, monitor fullness, change node types, pause, or convert a Pinecone dedicated read nodes index back to on-demand.

Adopt the Documents APILearn what changes in API version 2026-07, whether your existing code is affected, and how to move to schema-based indexes.

Overview

OverviewCompare Pinecone search types and choose the right retrieval approach: full-text (BM25), semantic (dense vector), sparse-vector, and hybrid.OverviewUpsert and search typed JSON documents in Pinecone with BM25 scoring, Lucene query syntax, dense and sparse vector ranking, and metadata filters.Query syntaxWrite Pinecone full-text search queries with Lucene query_string syntax, including boolean, phrase, prefix, boosting, and fuzzy operators.Text processingControl how Pinecone full-text search tokenizes and analyzes text, including tokens and analyzers, stemming, language options, and substring (n-gram) matching.

Semantic search

Semantic searchSearch a Pinecone index of dense vectors to find semantically similar records using text or vector queries, top_k results, and nearest neighbor lookup.OverviewCombine keyword and semantic retrieval in Pinecone with a text-match filter on a dense search, or by fusing separate searches with reciprocal rank fusion.Use a single indexRun hybrid search on one Pinecone index storing a dense and a sparse vector per record, weighting the two signals client-side with an alpha value.Use separate indexesStore dense and sparse vectors in two Pinecone indexes linked by ID, query each, and merge the results client-side with reciprocal rank fusion.Reciprocal rank fusionCombine results from separate Pinecone searches into a single ranking with reciprocal rank fusion (RRF), a client-side method that fuses rankings, not scores.

Sparse-vector searchRun sparse-vector search in Pinecone with custom sparse encoders like pinecone-sparse-english-v0 for token-weighted keyword retrieval.Filter by metadataNarrow Pinecone search results by adding metadata filter expressions to your query, using operators like $eq, $in, $gt, and $and for precise retrieval.Rerank resultsImprove retrieval quality by reranking initial search results with a hosted or external model to surface the most relevant matches for RAG.

Optimize

Increase relevanceImprove Pinecone search quality with reranking models, metadata filtering, and hybrid search that combines full-text and vector retrieval for RAG.Increase throughputIncrease Pinecone throughput with bulk import from object storage, batch upserts, parallel requests, and the Python gRPC SDK for faster ingestion.Decrease latencyReduce query and upsert latency in Pinecone using namespaces, metadata filters, targeting indexes by host, and regional colocation strategies.Save on Pinecone costsSave on Pinecone costs by using bulk import over upsert, namespaces for multitenancy, and query patterns that reduce read unit consumption.Test at scaleBenchmark Pinecone at production scale by importing 10M vectors and measuring semantic search throughput, query latency, and costs.

Manage data

Target an indexTarget a Pinecone index by host URL (recommended for production) or by name for data operations like upsert, query, and fetch across SDKs.Manage indexesList, describe, configure, and delete serverless indexes in Pinecone, including tags, deletion protection, and metadata index configuration.Manage namespacesCreate, describe, list, and delete namespaces in Pinecone serverless indexes, including defining filterable metadata fields and schemas ahead of upsert.

Manage backups

Backups overviewLearn how serverless index backups work in Pinecone, including scheduled backups, retention policies, and use cases for restoring or copying data.Back up an indexCreate backups of serverless indexes to protect data, copy indexes, or experiment with configurations using the Pinecone SDK, API, or console.Restore an indexRestore a Pinecone serverless index from a backup, change the index name, tags, or deletion protection, and preserve embedding model configuration.

Update dataPatch fields on Pinecone documents by ID, update record metadata and vector values by ID, or update metadata across documents and records with a filter.Delete dataDelete Pinecone documents or records from a namespace by ID or metadata filter, or delete every document or record the namespace holds.Fetch dataRetrieve Pinecone documents or records from a namespace by ID or metadata filter to inspect their fields, vector values, and metadata.List IDsList the IDs of Pinecone documents or records in a namespace of a serverless index, filter them by ID prefix, and paginate through the results.

Manage cost

Understanding Pinecone costUnderstand how Pinecone bills read units, write units, storage, egress, and embedding for full-text search, semantic search, and hybrid search.Manage costReduce Pinecone spend with strategies like spend alerts, ID-prefix listing, multitenant namespaces, prepaid credits, and support cost optimization help.Monitor usage and costsMonitor Pinecone usage and costs at the organization, index, and operation level by tracking read units, write units, and storage consumption.

Move to production

OverviewPrepare Pinecone indexes for production with best practices for project structure, security, scaling, API keys, and reliability across your workloads.Bring Your Own Cloud (BYOC)Deploy Pinecone BYOC in your own AWS, GCP, or Azure account for data sovereignty, network isolation, and regional data residency requirements.

Enforce security

OverviewOverview of Pinecone security features for production: API keys, SSO, service accounts, audit logs, CMEK encryption, backups, and Private Endpoints.Manage roles and accessAssign and manage roles for users, service accounts, and API keys using the Pinecone console or the Admin API.Manage roles with OktaAutomatically assign Pinecone organization and project roles from SAML attributes with Okta.SCIM provisioning with OktaAutomatically provision members and roles from Okta to Pinecone over SCIM.Configure CMEKSet up customer-managed encryption keys (CMEK) with AWS KMS to encrypt Pinecone data with keys you control, using IAM roles and key policies.Configure Private EndpointsConfigure Pinecone Private Endpoints with AWS PrivateLink or Azure Private Link to keep index traffic off the public internet and secure VPCs.Data deletion policyLearn how Pinecone permanently deletes customer data, including soft deletion, the 90-day retention window, and secure erasure of records and indexes.Configure audit logsEnable Pinecone audit logging to an Amazon S3 bucket to track user, service account, and API actions for compliance, security review, and WORM retention.

Error handlingHandle Pinecone API errors with retry logic, exponential backoff, and best practices for 4xx client errors, 5xx server errors, and rate limit responses.Monitor performanceMonitor Pinecone index performance metrics (query latency, throughput, and errors) in the Pinecone console or via Prometheus and Datadog.CI/CDBuild a GitHub Actions CI/CD workflow with Pinecone Local to run automated integration tests against an in-memory emulator without touching production.

Admin

Manage billing

Account deactivation for non-paymentLearn what happens when a Pinecone account is deactivated for non-payment: 30-day data retention, permanent deletion, and how to reactivate.Pinecone Standard plan trialEvaluate the Pinecone Standard plan with $300 in credits over 21 days, including bulk import, backup and restore, RBAC, and higher scale limits.Upgrade your planLearn how Pinecone Assistant admins upgrade to a paid plan to unlock higher limits, more files, evaluations, and advanced access controls.Change your payment methodUpdate the credit card or payment method on file for your Pinecone organization to keep billing details current and avoid failed charges.Downgrade your planDowngrade your Pinecone subscription from a paid tier back to the free Starter plan, including steps to review usage limits before switching.Download a usage reportExport a detailed Pinecone usage and cost report for your organization to analyze index consumption, credits, and monthly billing charges.Access your invoicesView, download, and manage your Pinecone billing invoices from the console, including past invoice history, payment status, and PDF exports.

Manage security

OverviewOverview of Pinecone Assistant admin security features including API keys, single sign-on, service accounts, and audit logs across your organization.Manage roles and accessAssign organization and project roles to users, service accounts, and API keys in Pinecone Assistant, with common role combinations for each scenario.Configure SSO with OktaIntegrate Okta with Pinecone to enable single sign-on, configure SAML settings, and manage secure user authentication for your organization.Configure audit logsSet up Pinecone audit logs streamed to an Amazon S3 bucket to track user, service account, and API actions across the control plane.

Manage organizations

OverviewLearn how Pinecone organizations group projects, share billing, and use organization roles to control member permissions and access to resources.Manage membersAdd, invite, and manage members in your Pinecone organization, including assigning organization roles, changing permissions, and removing users.Manage service accountsCreate and manage service accounts at the organization level in Pinecone for programmatic Admin API access, including client secrets and role assignment.OverviewLearn how Pinecone organizations group projects under shared billing, and how organization owners, members, and roles control access and permissions.Monitor usage and costMonitor Pinecone Assistant usage and cost, set monthly spend alerts, and track token usage across chat, context retrieval, and evaluation.Manage organization membersInvite new members to your Pinecone organization, change their organization roles, and remove users to control access across all projects.Manage service accountsCreate and manage organization-level Pinecone service accounts, retrieve access tokens, and grant programmatic access to the Admin API.

Manage projects

OverviewLearn how Pinecone projects organize indexes, cloud environments, API keys, and members, including project roles and per-role permission summaries.Create a projectCreate a new Pinecone project in your organization using the console or Admin API, including project name, tags, and customer-managed encryption keys (CMEK).Manage projectsView project details, rename projects, and delete projects in your Pinecone organization using the console or Admin API with owner permissions.Manage project membersAdd, remove, and update Pinecone project members using role-based access control (RBAC) to manage permissions and project owner roles.Manage API keysCreate, view, update, and delete Pinecone API keys with custom permissions per project, including scoped roles and data plane access controls.Manage service accountsAdd service accounts to a Pinecone project and assign project roles to enable programmatic API access for automated workflows and CI pipelines.OverviewUnderstand how Pinecone projects contain assistants, indexes, and users, and review project owner and project user roles and their permissions.Create a projectCreate a new Pinecone project in the console or through the Admin API, add tags, and optionally enable a customer-managed encryption key (CMEK).Manage projectsView project details, rename projects, add project tags, and delete Pinecone projects using the console or the Admin API with an access token.Manage project membersAdd users to a Pinecone project, assign project owner or user roles, edit permissions, and remove members to control access to indexes and API keys.Manage API keysCreate, view, update, and delete Pinecone API keys for a project, plus assign custom permission roles to control data plane and control plane access.Manage service accountsAdd and manage project-level service accounts in Pinecone Assistant to enable programmatic Admin API access, roles, and permissions.

Operations

Integrate with cloud storage

Amazon S3Set up a Pinecone storage integration with an Amazon S3 bucket using an IAM role to bulk import data into indexes and export audit logs.Google Cloud Storage (GCS)Set up a Pinecone storage integration with a Google Cloud Storage bucket using a service account key to bulk import data into your indexes.Azure Blob StorageSet up a Pinecone storage integration with an Azure Blob Storage container using a service principal to bulk import data into your indexes.Manage storage integrationsUpdate or delete existing Amazon S3, Google Cloud Storage, and Azure Blob storage integrations for your Pinecone project in the console.

Integrate with AI agents

MCP serverConnect AI agents to Pinecone through the MCP server to search docs, manage indexes, and query data from Claude, Cursor, Antigravity, or Claude Code.Build a knowledge retrieval agentBuild an AI agent that uses Pinecone to retrieve knowledge and answer questions accurately, with Pinecone as a tool inside the agent.

Local developmentRun Pinecone Local, an in-memory Docker emulator, to develop and test apps offline without an account or usage fees.

Using pods

OverviewUnderstand Pinecone pod-based indexes, including pod types and sizing. Pod indexes are legacy and unavailable to new customers, and serverless is recommended.Migrate a pod-based index to serverlessMigrate a Pinecone pod-based index to serverless for automatic scaling, better performance, and usage-based pricing with no minimum spend commitment.Choose a pod typeChoose a Pinecone pod type and size (s1, p1, p2). Pod indexes are legacy and unavailable to new customers, and serverless needs no capacity planning.Create a pod-based indexCreate a Pinecone pod-based index. Pod indexes are legacy and unavailable to new customers, and serverless is the current option for new indexes.Manage pod-based indexesManage Pinecone pod-based indexes. Pod indexes are legacy and unavailable to new customers, and serverless is recommended for new projects.Scale pod-based indexesScale Pinecone pod-based indexes by adding pods or replicas. Pod indexes are legacy and unavailable to new customers, and serverless scales automatically.

Back up and restore

Understanding collectionsLegacy documentation for Pinecone collections, a pod-only feature for creating static index snapshots. Collections aren't available for serverless indexes.Back up a pod-based indexLegacy guide for backing up Pinecone pod-based indexes using collections. Collections are a pod-only feature not available for serverless indexes.Restore a pod-based indexLegacy guide for restoring Pinecone pod-based indexes from collections. Pod indexes are no longer available to new customers as of August 2025.

Reference

IntroductionProgrammatic access to Nexus workspaces, contexts, sources, curation, and KnowQL queries.AuthenticationAuthenticate to the Nexus API with your Pinecone API key, sent directly or exchanged for a short-lived session token.

APIs

IntroductionAssistant API reference overview covering document upload, chat, and RAG endpoints supported by the Pinecone Python and Node.js SDKs.AuthenticationAssistant API authentication with API keys, HTTP headers, and SDK client initialization for Pinecone Assistant requests and RAG apps.LimitsReference for Pinecone Assistant limits, including file size, storage, chat token, and rate limits across the Starter, Builder, Standard, and Enterprise plans.IntroductionPinecone API reference overview for Database, Inference, and Admin endpoints with supported SDKs including Python, Node.js, Java, and Go.AuthenticationPinecone API authentication guide covering API key generation, SDK client setup, and HTTP header configuration for authorized requests.VersioningLearn how Pinecone versions its APIs by release date, how long each stable version is supported, and how to set the X-Pinecone-Api-Version header.

Database limits

OverviewReference for Pinecone Database rate, object, operation, and identifier limits, plus known index and serverless limitations.Rate limitsRequest-per-second, monthly usage, and model throughput limits for Pinecone Database serverless indexes, and the 429 errors returned when you exceed them.Object limitsLimits on the number and size of Pinecone Database projects, indexes, namespaces, storage, backups, and collections, by plan.Operation limitsFixed limits on Pinecone Database operations, including upsert, update, import, query, fetch, and delete batch sizes and metadata filter expressions.Identifier limitsMaximum length and allowed characters for Pinecone identifiers, including organization, project, index, namespace, and record names.Known limitationsKnown limitations and feature restrictions for Pinecone indexes, including upsert consistency, metadata rules, and serverless caveats.

ErrorsLook up the HTTP status codes the Pinecone API returns, what each 2xx success and 4xx or 5xx error means, and where to handle them in production.

Files

List FilesUpload a fileUpsert a fileCreate or replace a file in the specified assistant. If a file with the given assistant_file_id already exists, it will be replaced with the new file. If it doesn't exist, a new file will be created with that identifier.Describe a fileGet the current status and metadata of a file uploaded to an assistant.Delete a file List FilesUpload file to assistantDescribe a file uploadDelete an uploaded fileList FilesUpload file to assistantDescribe a file uploadDelete an uploaded fileList FilesUpload file to assistantDescribe a file uploadDelete an uploaded file

Operations

List operationsList all operations for an assistant. Returns operations that are in progress, as well as recently completed or failed operations.Describe an operationGet the status of an operation.

Chat

Chat with an assistantChat through an OpenAI-compatible interfaceChat with an assistantChat through an OpenAI-compatible interfaceChat with an assistantChat through an OpenAI-compatible interfaceChat with an assistantChat through an OpenAI-compatible interface

Evaluation

Evaluate an answerEvaluate an answerEvaluate an answerEvaluate an answer

Context snippets

Retrieve context from an assistantRetrieve context from an assistantRetrieve context from an assistantRetrieve context from an assistant

Admin

Organizations

List organizationsList all organizations associated with an account.Get organization detailsGet an organization's details.Delete an organizationDelete an organization and all its configuration; delete all its projects first.Update an organizationUpdate an organization's name.

Service accounts

List service accountsList service accounts in the caller's organization.Create a service accountCreate a service account with optional initial role bindings; the client secret is returned only once.Get service account detailsGet a service account by ID; the client secret is returned only from create and rotate-secret requests.Delete a service accountDelete a service account and its role bindings; tokens it minted are revoked within a few seconds.Update a service accountUpdate a service account's name; role bindings are managed through the role-binding endpoints.Rotate a service account's OAuth client secretRotate a service account's OAuth client secret; the previous secret and its tokens are revoked within seconds and the new secret is returned only once.Create a service accountCreate a service account with optional initial role bindings; the client secret is returned only once.List service accountsList service accounts in the caller's organization.Get service account detailsGet a service account by ID; the client secret is returned only from create and rotate-secret requests.Update a service accountUpdate a service account's name; role bindings are managed through the role-binding endpoints.Delete a service accountDelete a service account and its role bindings; tokens it minted are revoked within a few seconds.Rotate a service account's OAuth client secretRotate a service account's OAuth client secret; the previous secret and its tokens are revoked within seconds and the new secret is returned only once.Create an access token Create an access tokenCreate an access tokenGet an access tokenGet an access token

Users

List users in the organizationList users in the caller's organization, optionally filtered by email address.Get user detailsGet a user in the caller's organization by ID.Remove a user from the organizationRemove a user from the organization and revoke their role bindings; their Pinecone account is not deleted.List users in the organizationList users in the caller's organization, optionally filtered by email address.Get user detailsGet a user in the caller's organization by ID.Remove a user from the organizationRemove a user from the organization and revoke their role bindings; their Pinecone account is not deleted.

Invites

List invitesList pending and expired invites in the caller's organization.Invite a user to the organizationInvite a user to the organization by email and grant their initial role bindings.Get invite detailsGet an invite in the caller's organization by ID.Delete an inviteDelete a pending or expired invite and its role bindings; to remove an accepted user, delete the user instead.Resend an invite emailResend the invite email and extend its expiration to 7 days from now; limited to 100 emails per hour per organization.Invite a user to the organizationInvite a user to the organization by email and grant their initial role bindings.List invitesList pending and expired invites in the caller's organization.Get invite detailsGet an invite in the caller's organization by ID.Delete an inviteDelete a pending or expired invite and its role bindings; to remove an accepted user, delete the user instead.Resend an invite emailResend the invite email and extend its expiration to 7 days from now; limited to 100 emails per hour per organization.

Role bindings

List role bindingsList role bindings in the caller's organization, optionally filtered by principal, resource, and role.Create a role bindingGrant a role to a principal at an organization or project scope.Get role binding detailsGet a role binding in the caller's organization by ID.Delete a role bindingDelete a role binding; permissions are revoked when the deletion completes.Create a role bindingGrant a role to a principal at an organization or project scope.List role bindingsList role bindings in the caller's organization, optionally filtered by principal, resource, and role.Get role binding detailsGet a role binding in the caller's organization by ID.Delete a role bindingDelete a role binding; permissions are revoked when the deletion completes.

Architecture

Assistant architecturePinecone Assistant architecture overview covering document ingestion, chunking, vector retrieval, and LLM response generation for RAG.

Database

Indexes

List indexesCreate an indexCreate a Pinecone index. Define the schema for your index — dense vector, sparse vector, and full-text search fields — and, optionally, the deployment infrastructure (managed serverless or BYOC). To create an index with an integrated embedding model, use Create an index with integrated embedding.Create an index with integrated embedding Describe an indexDelete an indexConfigure an indexGet index statsList indexesCreate an indexCreate an index with integrated embeddingDescribe an indexDelete an indexConfigure an indexGet index statsList indexesCreate an indexCreate an index with integrated embeddingDescribe an indexDelete an indexConfigure an indexGet index statsList indexesCreate an indexCreate an index with integrated embeddingDescribe an indexDelete an indexConfigure an indexConfigure an existing index. For serverless indexes, you can configure index deletion protection, tags, and integrated inference embedding settings for the index. For pod-based indexes, you can configure the pod size, number of replicas, tags, and index deletion protection.Get index stats List indexesCreate an indexCreate an index with integrated embeddingDescribe an indexDelete an indexConfigure an indexGet index statsList indexesCreate an indexDescribe an indexDelete an indexConfigure an indexGet index statsList indexesCreate an indexDescribe an indexDelete an indexConfigure an indexGet index statsList indexesCreate an indexDescribe an indexDelete an indexThis operation deletes an existing index.Configure an indexThis operation configures the pod size and number of replicas for a pod-based index.Get index stats

Namespaces

List namespacesCreate a namespaceDescribe a namespaceDelete a namespaceList namespacesCreate a namespaceDescribe a namespaceDelete a namespaceList namespacesCreate a namespaceDescribe a namespaceDelete a namespaceList namespacesDescribe a namespaceDelete a namespace

Vectors

Search with a vectorSearch with textUpsert recordsUpsert records into a namespace. If a new value is upserted for an existing record ID, it will overwrite the previous value.Upsert text Fetch recordsLook up and return records by ID from a single namespace. The returned records include the vector data and/or metadata.Fetch records by metadataLook up and return records by metadata from a single namespace. The returned records include the vector data and metadata.Update a recordUpdate records by ID or by metadata in a namespace. Updating by ID changes the vector and/or metadata of a single record. Updating by metadata changes metadata across multiple records using a metadata filter.Delete recordsDelete records by id or by metadata from a single namespace.List record IDsList the IDs of records in a single namespace of a serverless index. An optional prefix can be passed to limit the results to IDs with a common prefix.Search with a vector Search with textUpsert recordsUpsert records into a namespace. If a new value is upserted for an existing record ID, it will overwrite the previous value.Upsert text Fetch recordsLook up and return records by ID from a single namespace. The returned records include the vector data and/or metadata.Fetch records by metadataLook up and return records by metadata from a single namespace. The returned records include the vector data and metadata.Update a recordUpdate records by ID or by metadata in a namespace. Updating by ID changes the vector and/or metadata of a single record. Updating by metadata changes metadata across multiple records using a metadata filter.Delete recordsDelete records by id or by metadata from a single namespace.List record IDsList the IDs of records in a single namespace of a serverless index. An optional prefix can be passed to limit the results to IDs with a common prefix.Search with a vector Search with textUpsert vectorsUpsert textFetch vectorsLook up and return vectors by ID from a single namespace. The returned vectors include the vector data and/or metadata.Fetch vectors by metadataLook up and return vectors by metadata filter from a single namespace. The returned vectors include the vector data and/or metadata.Update a vector Delete vectorsDelete vectors by id from a single namespace.List vector IDs Search with a vectorSearch with textUpsert vectorsUpsert textFetch vectorsLook up and return vectors by ID from a single namespace. The returned vectors include the vector data and/or metadata.Update a vector Delete vectorsDelete vectors by id from a single namespace.List vector IDs Search with a vectorSearch a namespace with a query vector or record ID and return the IDs of the most similar records, along with their similarity scores.Search with text Upsert vectorsUpsert textFetch vectorsLook up and return vectors, by ID, from a single namespace. The returned vectors include the vector data and/or metadata.Update a vector Delete vectorsDelete vectors, by id, from a single namespace.List vector IDs Upsert vectorsSearch with a vectorThe query operation searches a namespace, using a query vector. It retrieves the ids of the most similar items in a namespace, along with their similarity scores.Fetch vectors Update a vectorDelete vectorsList vector IDsUpsert vectorsSearch with a vectorSearch a namespace with a query vector or record ID and return the IDs of the most similar records, along with their similarity scores.Fetch vectors Update a vectorDelete vectorsList vector IDsUpsert vectorsSearch with a vectorThe query operation searches a namespace, using a query vector. It retrieves the ids of the most similar items in a namespace, along with their similarity scores.Fetch vectors Update a vectorDelete vectorsList vector IDs

Documents

Upsert documentsUpsert documents into a namespace.Search documentsSearch for documents in a namespace using one or more scoring methods (dense vector, sparse vector, text, or query string similarity).Fetch documentsFetch documents from a namespace. Returns the specified fields for each document. Exactly one of ids or filter must be specified.List documentsList documents in a namespace.Update documentsApply partial updates to documents in a namespace. Documents are selected either per ID with documents, or in bulk with filter.Delete documentsDelete documents from a namespace. Exactly one of ids, filter, or delete_all must be specified.

Backups

Create a backup of an indexList backups for all indexes in a projectList backups for an indexWhen include_deleted is false (or omitted), index_name must resolve to an active index in the project. If no active index by that name exists—including the case where only deleted indexes have used the name—the API returns 404, not an empty list.Describe a backup Delete a backupCreate an index from a backupList restore jobsDescribe a restore jobList collectionsCreate a collectionDescribe a collectionDelete a collectionCreate a backup of an indexList backups for all indexes in a projectList backups for an indexWhen include_deleted is false (or omitted), index_name must resolve to an active index in the project. If no active index by that name exists—including the case where only deleted indexes have used the name—the API returns 404, not an empty list.Describe a backup Delete a backupCreate an index from a backupCreate an index from a backup. For serverless backups, you can optionally set read_capacity so the restored index is created with dedicated read nodes (DRN) instead of defaulting to on-demand capacity.List restore jobs Describe a restore jobList collectionsCreate a collectionDescribe a collectionDelete a collectionCreate a backup of an indexList backups for all indexes in a projectList backups for an indexList all backups for an index.Describe a backup Delete a backupCreate an index from a backupList restore jobsDescribe a restore jobList collectionsCreate a collectionDescribe a collectionDelete a collectionCreate a backup of an indexList backups for all indexes in a projectList backups for an indexList all backups for an index.Describe a backup Delete a backupCreate an index from a backupList restore jobsDescribe a restore jobList collectionsCreate a collectionDescribe a collectionDelete a collectionList collectionsCreate a collectionDescribe a collectionDelete a collectionList collectionsCreate a collectionDescribe a collectionDelete a collectionList collectionsCreate a collectionDescribe a collectionDelete a collectionList collectionsCreate a collectionDescribe a collectionDelete a collection

Backup schedules

Create backup scheduleCreate a recurring backup schedule for a Pinecone serverless or BYOC index using the API, including daily, weekly, or monthly frequency and retention policy.List backup schedulesList all backup schedules configured for a Pinecone serverless index, including schedule IDs, frequency, retention, and next scheduled run times.Describe backup scheduleRetrieve details for a Pinecone backup schedule by schedule ID, including frequency, retention days, next scheduled run, and enabled status.Update backup scheduleUpdate a Pinecone backup schedule using the API, including pausing with enabled, changing frequency, and adjusting the retention expire_after_days.Delete backup scheduleDelete a backup schedule for a Pinecone serverless index using the API by schedule ID, while preserving all previously created backup snapshots.List schedule historyList execution history for a Pinecone backup schedule, including scheduled runs, completed backups, statuses, and cursor-based pagination support.Create backup scheduleCreate a recurring backup schedule for a Pinecone serverless index using the API, including daily, weekly, or monthly frequency and retention policy.List backup schedulesList all backup schedules configured for a Pinecone serverless index, including schedule IDs, frequency, retention, and next scheduled run times.Describe backup scheduleRetrieve details for a Pinecone backup schedule by schedule ID, including frequency, retention days, next scheduled run, and enabled status.Update backup scheduleUpdate a Pinecone backup schedule using the API, including pausing with enabled, changing frequency, and adjusting the retention expire_after_days.Delete backup scheduleDelete a backup schedule for a Pinecone serverless index using the API by schedule ID, while preserving all previously created backup snapshots.List schedule historyList execution history for a Pinecone backup schedule, including scheduled runs, completed backups, statuses, and cursor-based pagination support.

SDKs

Pinecone SDKs overviewBrowse official Pinecone SDKs for Python, Node.js, Java, and Go, including the version mappings between SDK releases and API versions.

Python

OverviewInstall and use the Pinecone SDK for Pinecone Python SDK: auth, typed clients, and API operations. For installation instructions, usage examples, and.

Node.js

OverviewInstall and use the Pinecone Node.js and TypeScript SDK to manage indexes, upsert vectors, run semantic search, and call the Admin and Inference APIs.

Java

OverviewInstall and use the Pinecone Java SDK to manage indexes and namespaces, upsert vectors, run semantic search, and call the Admin and Inference APIs from Java.OpenTelemetryMonitor Pinecone Java SDK operations with OpenTelemetry metrics for client latency, server processing time, and errors via Prometheus.

Go

OverviewInstall and use the Pinecone Go SDK to manage indexes and namespaces, upsert vectors, run semantic search, and call the Admin and Inference APIs from Go.

CLI

QuickstartPinecone CLI quickstart for installing pc, authenticating, and managing indexes, vectors, and data imports directly from your terminal.Command referenceComplete Pinecone CLI command reference covering syntax, subcommands, flags, help output, and exit codes for auth, index, and project.AuthenticationAuthenticate the Pinecone CLI using user login, service accounts, or API keys, including auth priority and Admin API access rules.Target contextSet the Pinecone CLI target context with pc target to control which organization and project your commands run against, including in CI/CD and JSON mode.

Tools

Spark-Pinecone connectorPinecone data tools: Use the connector to efficiently create, ingest, and update vector embeddings at scale with Databricks and Pinecone.

Control plane

Manage Workspaces

List workspacesList the workspaces in the project. Results are paginated; the pagination object is omitted when there are no further results.Create a workspaceCreate a Nexus workspace. The workspace name must be unique within the project.Describe a workspaceGet a description of a workspace.Delete a workspaceDelete an existing workspace. Deletion is asynchronous: the workspace transitions to the Terminating state and its contexts are deleted, after which the workspace itself is removed.

Data plane

Auth

Exchange a Pinecone API key for a session tokenExchange a Pinecone API key for the session token every other endpoint expects. Send the token as a bearer credential.Current identity (whoami)Who the current credential authenticates, and which project and workspace it is scoped to.

Project

Get the active projectThe project the session is scoped to, plus whether this deployment requires a workspace host.Project-wide task statsAggregate task counters across the whole project (all contexts), broken down by state and workflow. The per-context equivalent is GET /contexts/{slug}/tasks/stats.

Connectors

List the project's linked connectorsThe project's linked source providers. Filter by kind to those valid for a context kind.Get a connectorOne linked connector, secrets stripped.Unlink a connectorUnlink the connector from the project. Sources it already imported stay in place.Start the OAuth link flow for a providerReturns an authorize_url for the user to open in a browser. The connector is persisted by the provider's OAuth redirect callback once the user approves access.Link a connector with an API keyLink a key-based provider by supplying its API key directly, instead of running the OAuth flow.Browse a connector's folders/filesBrowse the provider's folders and files to find the ids to import. Omit folder for the root.List a context's connectors (with per-context enabled flag)The project's connectors, each with whether this context may import from it.Enable or disable a connector for a contextEnable or disable one connector for this context.

Contexts

List contexts (newest first, with stats)Every context in the active project, newest first, each with its aggregate task counters.Create a contextCreated empty, and not queryable until you import sources and curate them — curate is explicit. Optionally seed a manifest. A work context is the exception: queryable from day zero, built from traces of work rather than documents.Get a contextOne context and its derived lifecycle flags.Update a contextAll fields optional; an absent one is left untouched. An empty string clears description/guide, and {} clears the manifest back to defaults.Delete a context (and its Pinecone indexes)Delete the context, its sources, its curated knowledge, and its Pinecone indexes. Not reversible.Per-context task statisticsAggregate counters over every task this context has run. The project-wide equivalent is GET /stats.Trigger an on-demand self-tuning optimizeTunes the manifest from real query traffic, then chains a forced re-curate. Every field of the required body is optional, so {} is valid — and a no-op without candidate_queries. The context must be curated, and only one optimize runs at a time.Run the Design-flow explore agent (propose a manifest)Inspects the uploaded source and proposes manifest-template matches in the task's output (see ExploreOutput). Persists nothing — review the proposal, then apply it with a manifest update and a forced curate. Body is optional.Estimate the token + time cost of curating the in-progress manifestEstimates the token and time cost of curating the sources under the manifest in the request body, reported as the task's output (see ProfileEstimateOutput). Persists nothing. Body is optional.

Manifest

Get the context's manifestThe context's pinned manifest document. A context that pins nothing returns null — the runtime fills defaults. Writes go through the manifest field on PUT /contexts/{slug}.List the built-in manifest templatesThe templates a new context can be seeded from.

Curation

Get the curation ledgerWhat the last curate recorded: the per-source hash and timestamp map, artifact edges, corpus groups, the live version pin, delete tombstones, artifact-reclaim intents, and the pointer and glossary resume cursors. Chunk ids are tracked too but not served — an internal retrieval unit.Build or rebuild the context's index from its sourcesCurate the staged sources under the active manifest. Required before querying — there is no auto-curate. Body is optional; force: true rebuilds fully, which is what a manifest edit needs. Search contexts only: a work context builds via work/groom.Drop named sources from the index + ledgerDrop named sources from the index and the curation ledger without re-reading the rest of the corpus.

Source Files

Import source documents from a project connectorPull source documents from a linked project connector (Box, ...) into the context. Supply either folder or files. Stages the source only — curate explicitly before querying.Upload a source file or archiveOne file per request, max 2 GiB; archives (.zip, .tar, .tar.gz, .tgz) are expanded by the import runtime. Stages the source without indexing it — curate explicitly before querying.List the source file tree (root)The root of the source tree — what has been staged but not necessarily curated.Source file/dir/size countsAggregate counts. Where source limits apply, five more fields carry the caps and usage against them (max_files_per_context, max_bytes_per_context, max_bytes_per_file, used_files, used_bytes); they are absent otherwise.Flat manifest of source files (path, size, last_modified)Flat source-object inventory. Excludes the inbox/ staging directory.List the source file tree under a pathOne directory of the source tree.Read a source file (raw bytes)The raw bytes of one staged source file.Delete a source fileRemove one file from the source tree. Its index entries survive until the next curate.

Knowledge

List curated knowledge (root)The root of the curated knowledge tree — chunks and artifacts the last curate wrote.Knowledge file/dir/size countsFile, directory, and byte counts over the curated knowledge tree.List curated knowledge under a pathOne directory of the curated knowledge tree.Read a curated knowledge file (raw bytes)The raw bytes of one curated knowledge file.

Query

Run one KnowQL query turnSend ask and, on a new session, a scope of 1–10 contexts; continue an existing one with Fetch one query (turn)One turn, including its steps, citations, and usage. This is what a background run is polled with.Cancel an in-progress query turnIdempotent — an already-terminal query is returned unchanged. On success the turn's status becomes cancelled.Fetch the recorded trace for a query turnThe per-turn debug trace (steps, tool calls, strategy, cost, rollup). Trace persistence is unconditional — every turn lands a trace blob — so this is available once the turn is terminal.Stream a query turn's events (SSE)The turn's events as they land — the same type-named events POST /query emits with stream: true, as a standalone resumable subscription. Reconnect with Last-Event-ID to replay from where the stream dropped.Record thumbs-up/down feedback on a query turnRecord a thumbs-up or thumbs-down on a turn. The rating is returned on the turn thereafter.List the selectable query modelsWhat the model field on POST /query draws from: the default, every catalog entry, the tier → model-id map, per-phase defaults, and the curate-capable ids. Only entries with available: true are selectable; the rest are rejected with 400.List project-wide query sessions (newest first)Every conversation thread in the project, newest first.Get a session with its queries (conversation order)One session with its turns in conversation order.Delete a session and its queriesDelete a session and every turn belonging to it.

Tasks

List tasks (project-wide, paginated)Task records across the project, newest first. Filter by workflow, state, or context, and page with limit plus page or offset.Get a task (with steps)Unlike the listing, this carries the task's steps — every one by default, or the newest N with steps_limit, where steps_total reports the true count.Cancel/terminate a taskBody is optional. Max body 1 MiB.

Task Files

Read a task file (raw bytes)Serves a task's files while it runs and after it finishes. This is how a finished pack's .context.zip is downloaded.Stat a task fileStat one file in the task container without reading its bytes.List a task's files under a pathOne directory inside the task's working directory.List a task's working-directory filesThe root of the task's working directory.Delete a task fileDelete one file from a running task container.

Examples

NotebooksRunnable Colab notebooks covering semantic search, lexical search, hybrid search, RAG, embeddings, reranking, and data ingestion with Pinecone.Sample appsSample apps and tools built with Pinecone, including semantic search, multi-tenant RAG, multimodal search, Assistant chat, and pgvector migration.Reference architecturesOfficial AWS reference architecture for building high-scale production systems with Pinecone, including Pulumi IaC, documentation, and video tutorial.

Connect an integration

View allBrowse Pinecone integrations across vector embedding providers, data ingestion tools, frameworks, and infrastructure to ship AI apps faster.

IDEs & CLIs

OverviewUse Pinecone with agentic IDEs and CLIs like Claude Code, Gemini CLI, and Cursor via MCP server, plugins, and agent skills for vector search.Agent SkillsInstall the Pinecone Agent Skills library in any agentic IDE to manage indexes, run semantic search, and build RAG assistants with natural language.Claude Code PluginInstall the official Pinecone plugin for Claude Code to manage indexes, run vector search, and build RAG assistants from the terminal with slash commands.Cursor PluginInstall the official Pinecone plugin for Cursor to manage indexes, run vector search, and build RAG apps from the editor using MCP and slash commands.Gemini CLI ExtensionInstall the official Pinecone extension for Gemini CLI to manage indexes, run vector search, and build RAG assistants with natural-language commands.

Data sources

AirbyteUse the Airbyte Pinecone destination connector to build no-code ETL pipelines that embed source data and upsert vectors for semantic search and RAG.ApifyUse the Apify Pinecone integration to crawl and scrape websites, embed the results, and upsert vectors for RAG and semantic search over web content.ArynUse Aryn Sycamore and the Partitioning Service with Pinecone to extract, chunk, and embed complex PDFs and documents for higher-accuracy RAG pipelines.BoxConnect Box folders to Pinecone Database to embed documents with Pinecone Inference and build RAG agents and semantic search over Box content.ConfluentUse the Pinecone Sink Connector for Confluent Kafka to stream events, embed them with LLMs, and upsert vectors for real-time semantic search and RAG.DatabricksUse Databricks and the Pinecone Spark connector to distribute embedding jobs across a cluster and upsert vectors at scale for semantic search and RAG.DatavoloUse Datavolo pipelines to source, transform, and enrich unstructured data into Pinecone for retrieval augmented generation and multimodal AI apps.EstuaryUse the Estuary Pinecone connector to build real-time, millisecond-latency pipelines that incrementally embed source data for LLM search and RAG.FleakUse Fleak's low-code workflows with Pinecone to combine SQL, LLM inference, and vector search in serverless APIs for RAG and data enrichment pipelines.FlowiseAIUse FlowiseAI with Pinecone to build low-code LLM apps that upsert documents and query vector indexes for RAG chatbots and semantic search flows.GathrUse Gathr's no-code data-to-outcome platform with Pinecone to build enterprise RAG apps, chatbots, and vector pipelines from ingestion through deployment.MatillionUse Matillion Data Productivity Cloud to chunk, embed, and upsert data into Pinecone with low-code AI pipelines for RAG, ETL, and GenAI apps.NexlaUse Nexla no-code data pipelines to ingest from SharePoint, OneDrive, and 500+ sources into Pinecone for enterprise RAG and vector search apps.RedpandaStream events into Pinecone with Redpanda Connect's declarative YAML pipelines for real-time vector ingestion, at-least-once delivery, and RAG ETL.SnowflakeDeploy Pinecone on Snowpark Container Services to run vector search and GenAI apps inside Snowflake without moving data out of your warehouse.StreamNativePipe Apache Pulsar topics to Pinecone via the StreamNative sink connector for real-time vector ingestion, event streaming, and Kafka-compatible RAG.UnstructuredUse Unstructured to parse, chunk, and load PDFs, Word, and 25+ document types into Pinecone for LLM-ready RAG ETL pipelines and semantic search apps.

Frameworks

AI EngineUse the AI Engine WordPress plugin with Pinecone to power chatbots, semantic search, and RAG over site content directly from your WordPress dashboard.Amazon BedrockSelect Pinecone as a Knowledge Base for Amazon Bedrock to ground LLMs with your enterprise data and build accurate, low-latency RAG applications on AWS.Amazon SageMakerCombine Amazon SageMaker with Pinecone to host LLMs and embedding models for scalable retrieval augmented generation and vector search workloads.Cloudera AIIntegrate Cloudera AI with Pinecone to run distributed Spark and Python embedding pipelines and power scalable RAG and vector search on enterprise data.Context DataUse Context Data with Pinecone to build no-code flows from Postgres, S3, Salesforce, and more, embedding data and querying indexes in Query Studio.GenkitUse the Genkit Pinecone plugin with Firebase to build AI features with type-safe indexers, embedders, and retrievers for RAG and semantic search apps.HaystackUse Deepset Haystack's PineconeDocumentStore to build production NLP pipelines that index, embed, and query documents for question answering and RAG.Instill AIWire Pinecone into Instill AI no-code pipelines to upsert, query, and power RAG agents, chatbots, and knowledge bases with vector similarity.LangChainBuild LangChain RAG apps, agents, and chatbots on Pinecone: manage embeddings, vector stores, retrievers, and chains for LLM-powered search.LlamaIndexBuild LlamaIndex RAG pipelines on Pinecone: ingest documents, structure private data, and run semantic search and question-answering over LLMs.n8nAutomate Pinecone vector store and Assistant workflows in n8n: build RAG pipelines, semantic search, and no-code AI automations with 400+ apps.NucliaPower Nuclia RAG-as-a-Service with Pinecone: index documents, customize retrieval and chunking strategies, and deploy LLM knowledge boxes at scale.OctoAICombine OctoAI's GTE Large embeddings and open-source LLMs with Pinecone and Canopy for cost-effective RAG, semantic search, and generative AI apps.VoltAgentBuild TypeScript AI agents with VoltAgent and Pinecone: automatic embeddings, semantic search, metadata filtering, and observable RAG in production.

Infrastructure

AWS MarketplaceSubscribe to Pinecone through AWS Marketplace for centralized procurement, pay-as-you-go billing, and consolidated invoicing on your AWS account.GitHub CopilotInstall the Pinecone GitHub Copilot Extension to query your indexes, get coding assistance for the Pinecone API, and debug RAG apps from your editor.Google Cloud MarketplaceSubscribe to Pinecone through Google Cloud Marketplace for centralized procurement, pay-as-you-go billing, and consolidated GCP invoicing.Microsoft MarketplaceBuy and manage Pinecone through Microsoft Marketplace with pay-as-you-go billing, centralized SaaS procurement, and simplified Azure licensing.PulumiProvision Pinecone indexes and collections as code with Pulumi in Python, TypeScript, Go, or C# for infrastructure-as-code vector database DevOps.TerraformManage Pinecone indexes, collections, API keys, and projects with the Terraform provider for repeatable infrastructure-as-code and DevOps workflows.VercelDeploy AI apps on Vercel with Pinecone as long-term memory: one-click setup, scalable vector search, and RAG for Next.js and edge-deployed frontends.ZapierAutomate Pinecone with Zapier: trigger workflows on index changes, upsert data from spreadsheets, and connect vector search to 6,000+ no-code apps.

Models

AnyscalePair Anyscale Endpoints with Canopy and Pinecone to build RAG apps on managed open-source LLMs, with vector search and semantic retrieval out of the box.CohereConnect Pinecone and Cohere to ship vector search and RAG applications: generate Cohere embeddings, index them in Pinecone, and rerank results.Voyage AIGenerate top-ranked Voyage AI embeddings and rerankers, then index them in Pinecone for high-accuracy RAG, semantic search, and code retrieval apps.Hugging Face Inference EndpointsGenerate embeddings with Hugging Face Inference Endpoints and index them in Pinecone for semantic search, RAG, and transformer model deployment.Jina AIUse Jina AI long-context embeddings with Pinecone for multilingual semantic search, RAG chatbots, and domain-specific retrieval up to 8K tokens.OpenAIPair OpenAI embeddings and completion models with Pinecone for semantic search, long-term memory, RAG, and context-aware LLM question-answering.Twelve LabsStore Twelve Labs multimodal video embeddings in Pinecone to power video search, recommendations, and content moderation with fast similarity retrieval.

Observability

DatadogMonitor Pinecone with the Datadog integration to track request latency, index fullness, and usage trends, and alert on anomalies in vector workloads.LangtraceTrace Pinecone API calls with Langtrace's OpenTelemetry SDK to debug RAG pipelines, monitor vector DB latency, and ship LLM observability.New RelicMonitor Pinecone with New Relic's Prometheus quickstart: dashboards, alerts, and AI observability for vector database latency and RAG performance.TraceloopInstrument Pinecone with Traceloop's OpenLLMetry SDK to emit OpenTelemetry traces and metrics for LLM observability in Datadog, Grafana, and more.TruLensEvaluate and track Pinecone RAG apps with TruLens: measure grounding, relevance, and hallucinations to iterate on vector search configurations fast.

Build an integration

Integration ecosystemHow Pinecone integrations are built with public SDKs and APIs and how listings on the Integrations hub are reviewed for partner support and quality.Attribute usage to your integrationAttribute Pinecone SDK and REST usage to your integration with source tags and User-Agent values so support and analytics can trace traffic to your product.Connect your users to PineconeEmbed a Connect to Pinecone flow in your app or notebook so users can sign in, choose a project, and receive an API key without leaving your integration.

Changelog

2026Every change to Pinecone in 2026, including new features, API updates, deprecations, and pricing and plan changes.2025Every change to Pinecone in 2025, including new features, API updates, deprecations, and pricing and plan changes.2024Every change to Pinecone in 2024, including new features, API updates, deprecations, and pricing and plan changes.2023Every change to Pinecone in 2023, including new features, API updates, deprecations, and pricing and plan changes.2022Every change to Pinecone in 2022, including new features, API updates, deprecations, and pricing and plan changes.Feature availabilityPinecone feature availability across public preview, general availability, and limited availability releases, with links to the relevant changelog entries.

Contact support

Contact SupportContact Pinecone Support from the console Help center, check which billing and support plans include it, and find business hours and Sev-1 coverage.How to work with SupportGet faster answers from Pinecone Support: try the AI support chatbot, use your account email, open tickets in the console, and pick an accurate severity.Pinecone Support SLAsLook up Pinecone Support first-response SLAs by support plan and ticket severity, from 30 minutes on a Premium Sev-1 to business days at lower severities.

Account management

Login code issuesFix a rejected Pinecone login code, which can expire after 10 hours or be invalidated by a shared inbox, a newer code request, or an offset system clock.Custom data processing agreementsRequest a custom data processing agreement (DPA) with Pinecone if your team requires terms beyond the standard agreement available on the Pinecone website.Delete your organizationDelete a Pinecone organization by first deleting its indexes, collections, and projects, then downgrading to the Starter plan. This can't be undone.Delete your accountDelete your Pinecone account by removing your user from every organization and deleting any organization where you are the sole member. This is permanent.Billing disputes and refundsUnderstand Pinecone's policy of not refunding unused indexes, and why migrating a pod-based index to serverless can lower a bill you'd otherwise dispute.

Indexes

Wait for index creation to be completeWait for Pinecone index creation to finish using describe_index polling in the Python SDK before upserting to avoid 403 or 404 errors.Restrictions on index namesLearn the rules for Pinecone index names: lowercase alphanumeric Latin characters and dashes only, no dots, and 52 characters total with your project ID.Return all vectors in an indexLearn why a single Pinecone query can't return every vector in an index, and how to page through record IDs with the list operation instead.

Data

Embedding values changed when upsertedDiagnose why Pinecone embedding values appear changed after upsert, including float32 precision rounding and how the API serializes numeric vector data.Limitations of querying by IDUnderstand why querying by record ID can omit that ID under ANN search, and when to use fetch or metadata filters to retrieve a specific vector reliably.

Common errors

Index creation error - missing spec parameterFix the Pinecone create_index TypeError for a missing spec parameter by passing the spec argument that sets the index's deployment model, cloud, and region.Serverless index creation error - max serverless indexesResolve the Pinecone serverless index creation error caused by hitting the 20-index project limit by deleting unused indexes or upgrading your plan.Serverless index connection errorsFix Pinecone serverless connection errors that fail to resolve a controller hostname by upgrading to a current version of the Python or Node.js SDK.Error: Handshake read failed when connectingFix the Pinecone 'Handshake read failed' connection error by checking your firewall and network, then verifying your SDK setup and API key are correct.PineconeAttribute errors with LangChainResolve PineconeAttribute errors in LangChain caused by outdated packages by upgrading langchain-pinecone and the Pinecone Python SDK to current releases.Error: Cannot import name 'Pinecone' from 'pinecone'Fix the Python ImportError 'cannot import name Pinecone from pinecone' by upgrading the Pinecone Python SDK to 3.0.0 or later, with or without gRPC.Python AttributeError: module pinecone has no attribute initFix the Python AttributeError 'module pinecone has no attribute init' by upgrading to Pinecone Python SDK 3.0 or later, which serverless indexes require.

Miscellaneous

Node.js TroubleshootingTroubleshoot Pinecone Node.js SDK issues that work locally but fail in deployment, including environment, network, and serverless runtime configuration.CORS IssuesTroubleshoot Pinecone CORS errors and Access-Control-Allow-Origin issues when calling the API from localhost or browser-based web apps.Debug model vs. Pinecone recall issuesDistinguish an embedding model problem from a Pinecone recall problem using a six-step evaluation that compares brute-force search against your index.Unable to pip installResolve install issues for the Pinecone Python SDK: pick the right Python 3.x command, install pinecone or pinecone with gRPC, and upgrade to the latest.