Pinecone MCP Server

Example Use Cases

• 1

  List and describe available Pinecone indexes and collections to understand available vector data.

• 2

  Query vectors (with support for namespaces, filters, and adjustable top-K) and retrieve nearest neighbors programmatically.

• 3

  Fetch specific vectors by ID or obtain index statistics as part of decision-making or retrieval workflows.

Description

This project implements a Model Context Protocol (MCP) server that exposes Pinecone vector database operations as tools usable by AI agents. It supports listing and describing indexes and collections, querying vectors with namespaces and filters, fetching vectors by ID, and retrieving index statistics. The server is designed for easy integration with agent configurations (e.g., Claude Desktop) and can be run via npx or installed globally, passing the Pinecone API key via environment variables.