Skip to main content
Watch: I Built an AI App Without Leaving Databricks

Databases - official Pinecone server

Pinecone MCP server

Local stdio MCP server that lets coding agents search Pinecone documentation and manage and query Pinecone indexes: list, describe, create, upsert, search and rerank. Without an API key it can still search the docs. It runs locally from @pinecone-database/mcp (npm) with API key auth.

Last verified 9 tools6 sources

Install

Terminal
claude mcp add-json pinecone-mcp \
  '{"type": "stdio",
    "command": "npx",
    "args": ["-y", "@pinecone-database/mcp"],
    "env": {"PINECONE_API_KEY": "YOUR_API_KEY"}}'
JSON
{
  "mcpServers": {
    "pinecone": {
      "command": "npx",
      "args": [
        "-y", "@pinecone-database/mcp"
      ],
      "env": {
        "PINECONE_API_KEY": "{{YOUR_API_KEY}}"
      }
    }
  }
}

Generic

JSON
{
  "mcpServers": {
    "pinecone": {
      "command": "npx",
      "args": [
        "-y", "@pinecone-database/mcp"
      ],
      "env": {
        "PINECONE_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

Tools

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 record count and 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 records in an index by text query with integrated embedding, metadata filtering and reranking options.
cascading-search
Searches across multiple indexes, deduplicating and reranking the results.
rerank-documents
Reranks a collection of records or text documents using a reranking model.

As documented on October 17, 2026. A running server reports its current list through tools/list.

Notes and caveats

  • Supports only indexes with integrated embedding (integrated inference); indexes for externally generated vectors, standalone embeddings and Assistants are not supported.
  • Requires Node.js v20+.
  • Pinecone recommends its Claude Code and Cursor plugins (which bundle this server plus skills) for the easiest setup; also installable as a Gemini CLI extension.
  • Latest release v0.3.0 (2026-08-07).
  • Docs give no beta/GA label.

FAQ

What does the Pinecone MCP server do?

Local stdio MCP server that lets coding agents search Pinecone documentation and manage and query Pinecone indexes: list, describe, create, upsert, search and rerank. Without an API key it can still search the docs. The vendor documents 9 tools, including search-docs, list-indexes, describe-index and describe-index-stats.

How do I add the Pinecone MCP server to Claude Code?

Pinecone documents a Claude Code config; copy it from the Claude Code section of this page. Snippets for Cursor and Generic are there too.

Does the Pinecone MCP server need authentication?

Yes, API key. Transport: stdio. Status: Unlabelled, last checked October 17, 2026.

Sources

Last verified against 6 Pinecone pages, then re-checked independently.

Newsletter

New MCP servers, the week they ship

One email per week. New official MCP servers, the coding tools that support them and how developers wire them up.