Pinecone Developer MCP Server
Official Pinecone MCP server: documentation search, index creation, record upserts and search from your agent
- 72 ★ on the repository
- Last commit
- Apache-2.0
Broad access
Why this level
- Creates indexes and writes data to an external service
- Uses a secret API key
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add mcp/pinecone-developer-mcp-serverDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- Creates indexes and writes data to an external service
- Uses a secret API key
This is third-party code. Review the repository files before installing.
What it does
The server connects the agent to a Pinecone vector database project and its documentation. The agent can search the docs, list indexes, read their configuration and per-namespace statistics. With an API key it creates indexes with an integrated embedding model, upserts records and searches them with metadata filters and reranking. It also offers search across several indexes and standalone document reranking.
Who it is for. For developers building search or RAG on Pinecone who want to configure and test indexes from their editor.
Good fit when
- You need to choose an index configuration for your use case and create it right away
- You want to test search quality on your data without leaving the editor
- You need Pinecone code grounded in current documentation
Not a fit when
- Your indexes lack integrated inference, which the server does not support
- You want agent access to a knowledge base via Pinecone Assistant, which has a separate Assistant MCP
Example request
List all my Pinecone indexes and search my-docs for records about authenticationMCP
- Transport
- stdio
- Authentication
- key is optional
Environment variables
PINECONE_API_KEYsecret- API key from the Pinecone console. Without it only documentation search works.
Limitations
Only indexes with integrated inference are supported; plain vector search and standalone embeddings are not. Working with indexes needs a Pinecone API key; without it only documentation search works. Requires Node.js 20 or newer. Access to Pinecone from Russia without a VPN is unverified.
How to disable. Remove the pinecone block from mcp.json or claude_desktop_config.json. For Gemini CLI remove the extension with gemini extensions uninstall.
FAQ
Does it work without a key?
Yes, but only for Pinecone documentation search. Managing and querying indexes needs a key.
Why does search on my index fail?
Most likely the index has no integrated embedding model. The server supports only such indexes, so create a new one with a model.
Discussion
No questions or reviews yet.
Related
MCP serversThe official MCP server for the Chroma vector database: collections, documents, semantic and full-text search
Official MongoDB MCP server: queries and aggregations, index management, Atlas clusters and local deployments
A local MCP server that turns a folder of documents into a knowledge base with hybrid search for Claude Code and Cursor
Neon's official remote MCP server: projects, branches, SQL queries and Postgres migrations from your agent
Redis's official MCP server: strings, hashes, streams, JSON and vector search from your agent
An MCP server with a toolbox for global briefs: markets, SEC filings, conflict, climate and news from open sources
README in short
The README lists the server's capabilities and tools: docs search, index listing and description, stats, model-based index creation, record upserts and search, cascading search and reranking. It gives npx configurations for Cursor and Claude Desktop and installation as a Gemini CLI extension. Limitations and common Node.js, key and network issues are covered separately. Apache 2.0 licensed.