Pinecone

by pinecone-io

Not rated
GitHub

About

Connect AI tools with Pinecone projects to search, configure indexes, generate code, and manage data.

Details

Author
pinecone-io
Categories
Database, Other, Knowledge Base, Search, AI

Setup

Install Pinecone in your MCP client (Claude Desktop, Cursor, Windsurf, and others).

Repository: https://github.com/pinecone-io/pinecone-mcp

Follow the installation instructions in the repository README, then restart your MCP client.

TheModel Context Protocol(MCP) is a standard that allows coding assistants and other AI tools to interact with platforms like Pinecone. The Pinecone Developer MCP Server allows you to connect these tools with Pinecone projects and documentation.

- SearchPinecone documentationto answer questions accurately.
- Help you configure indexes based on your application's needs.
- Generate code informed by your index configuration and data, as well as Pinecone documentation and examples.
- Upsert and search for data in indexes, allowing you to test queries and evaluate results within your dev environment.

See thedocsfor more detailed information.

This MCP server is focused on improving the experience of developers working with Pinecone as part of their technology stack. It is intended for use with coding assistants. Pinecone also offers theAssistant MCP, which is designed to provide AI assistants with relevant context sourced from your knowledge base.

To configure the MCP server to access your Pinecone project, you will need to generate an API key using theconsole. Without an API key, your AI tool will still be able to search documentation. However, it will not be able to manage or query your indexes.

The MCP server requiresNode.jsv20 or later. Ensure thatnodeandnpxare available in yourPATH.

Next, you will need to configure your AI assistant to use the MCP server.

To add the Pinecone MCP server to a project, create a.cursor/mcp.jsonfile in the project root (if it doesn't already exist) and add the following configuration:

{ "mcpServers": { "pinecone": { "command": "npx", "args": [ "-y", "@pinecone-database/mcp" ], "env": { "PINECONE_API_KEY": "<your pinecone api key>" } } } }

You can check the status of the server inCursor Settings > MCP.

To enable the server globally, add the configuration to the.cursor/mcp.jsonin your home directory instead.

It is recommended to use rules to instruct Cursor on proper usage of the MCP server. Check out thedocsfor some suggestions.

Use Claude desktop to locate theclaude_desktop_config.jsonfile by navigating toSettings > Developer > Edit Config. Add the following configuration:

{ "mcpServers": { "pinecone": { "command": "npx", "args": [ "-y", "@pinecone-database/mcp" ], "env": { "PINECONE_API_KEY": "<your pinecone api key>" } } } }

Restart Claude desktop. On the new chat screen, you should see a hammer (MCP) icon appear with the new MCP tools available.

To install this as aGemini CLIextension, run the following command:

gemini extensions install https://github.com/pinecone-io/pinecone-mcp

You will need to provide your Pinecone API key in thePINECONE_API_KEYenvironment variable.

export PINECONE_API_KEY=<your pinecone api key>

When you rungeminiand pressctrl+t,pineconeshould now be shown in the list of installed MCP servers.

Once configured, your AI tool will automatically make use of the MCP to interact with Pinecone. You may be prompted for permission before a tool can be used.

Here are some prompts you can try with your AI assistant:

- "Search the Pinecone docs for information about metadata filtering"
- "List all my Pinecone indexes and describe their configurations"
- "Create a new index called 'my-docs' using the multilingual-e5-large model"
- "Upsert these documents into my index:
[paste your documents]"
- "Search my index for records related to 'authentication best practices'"
- "What namespaces exist in my index, and how many records are in each?"

Pinecone Developer MCP Server provides the following tools for AI assistants to use:

- search-docs: Search the official Pinecone documentation.
- list-indexes: Lists all Pinecone indexes.
- describe-index: Describes the configuration of an index.
- describe-index-stats: Provides statistics about the data in the index, including the number of records and available namespaces.
- create-index-for-model: Creates a new index that uses an integrated inference model to embed text as vectors.
- upsert-records: Inserts or updates records in an index with integrated inference.
- search-records: Searches for records in an index based on a text query, using integrated inference for embedding. Has options for metadata filtering and reranking.
- cascading-search: Searches for records across multiple indexes, deduplicating and reranking the results.
- rerank-documents: Reranks a collection of records or text documents using a specialized reranking model.

Only indexes with integrated inference are supported. Assistants, indexes without integrated inference, standalone embeddings, and vector search are not supported.

MCP server not appearing in your AI tool

- Ensure Node.js v20 or later is installed:node --version
- Verifynpxis available in your PATH:which npx
- Check that your configuration file is in the correct location and has valid JSON syntax
- Restart your AI tool after making configuration changes

"Invalid API key" or authentication errors

- Verify your API key is correct in thePinecone console
- Check that thePINECONE_API_KEYenvironment variable is set correctly in your MCP configuration
- Ensure there are no extra spaces or quotes around the API key value

- The MCP server only supports indexes with integrated inference. If you're trying to use a serverless index without integrated inference, you'll need to create a new index with an embedding model
- Check the MCP server logs for error messages. In Cursor, view logs inCursor Settings > MCP

- If using a corporate network, ensure your firewall allows connections toapi.pinecone.io
- Try running the server manually to see detailed error output:PINECONE_API_KEY=<your-key> npx @pinecone-database/mcp

We welcome your collaboration in improving the developer MCP experience. Please submit issues in theGitHub issue tracker. Information about contributing can be found inCONTRIBUTING.md.

Search, Query and interact with data in your Milvus Vector Database.

Provides AI assistants with persistent memory using ChromaDB vector storage.

An MCP server for vector storage and retrieval using ChromaDB.

Interact with on-disk documents using agentic RAG and hybrid search via LanceDB.

A toolkit for loading data into the Qdrant vector database, supporting AI-powered development workflows.

Implement semantic memory layer on top of the Qdrant vector search engine

Semantic search using the Qdrant vector database.

A server for interacting with the Astra DB database-as-a-service built on Apache Cassandra.

A vector database server powered by Chroma, enabling semantic document search, metadata filtering, and document management.

An MCP server for the Chroma embedding database, providing persistent, searchable working memory for AI-assisted development with features like automated context recall and codebase indexing.

No reviews yet — be the first

Sign in to leave a review

Use Google, GitHub, or an email account so ratings stay tied to real people.

Email sign in

No reviews posted yet.