playbooks

home / mcp / pinecone mcp server

Pinecone MCP Server

Provides read and write access to a Pinecone index via an MCP server enabling semantic search, document read/list, and index stats.

Installation
Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "sirmews-mcp-pinecone": {
      "command": "uv",
      "args": [
        "--directory",
        "{project_dir}",
        "run",
        "mcp-pinecone"
      ]
    }
  }
}

You can run the Pinecone MCP Server to read and write data to a Pinecone index from your MCP client, enabling semantic search and document management directly through the MCP interface.

How to use

You will connect your MCP client (for example, Claude Desktop) to the Pinecone MCP Server. The server exposes tools to perform semantic search, read and list documents, and manage index data. Use the server to upsert processed documents, fetch specific records, and retrieve Pinecone index statistics. Your MCP client will talk to the server using the configured MCP runner, and the server will communicate with Pinecone to perform the requested operations.

How to install

Prerequisites: you need a runtime that can run MCP servers via the MCP runner tooling (uv/uvx) and a working Node.js/npm environment if you install via Smithery. Ensure you have a Pinecone account and an API key for your index.

Install via Smithery to set up the Pinecone MCP Server for Claude Desktop automatically:

npx -y @smithery/cli install mcp-pinecone --client claude

Install the server locally using the MCP runtime. You have two common options depending on your preferred runtime approach.

uvx install mcp-pinecone

Alternatively, install via Python’s pip integration for MCP if you use uvp/pip-based workflows (the exact command shown uses uv and uvx in practice). Use the command that matches your setup.

Configuration snippets you will use

The server configuration is provided as MCP JSON blocks. Use these exact structures to configure how you run the Pinecone MCP Server from your Claude Desktop environment.

"mcpServers": {
  "mcp-pinecone": {
    "command": "uv",
    "args": [
      "--directory",
      "{project_dir}",
      "run",
      "mcp-pinecone"
    ]
  }
}
"mcpServers": {
  "mcp-pinecone": {
    "command": "uvx",
    "args": [
      "--index-name",
      "{your-index-name}",
      "--api-key",
      "{your-secret-api-key}",
      "mcp-pinecone"
    ]
  }
}

Available tools

semantic-search

Search for records in the Pinecone index using semantic embeddings to retrieve relevant items.

read-document

Read a specific document from the Pinecone index by identifier.

list-documents

List all documents currently stored in the Pinecone index.

pinecone-stats

Retrieve statistics about the Pinecone index, including the number of records, dimensions, and namespaces.

process-document

Process a document into chunks, generate embeddings, and upsert the chunks into the Pinecone index.

list_resources

List available MCP resources for this server.

read_resource

Read a specific resource available through the MCP server.

list_tools

List all implemented tools provided by the server.

call_tool

Invoke a specific tool to perform its operation.

get_prompt

Retrieve the current prompt configuration for a request.

list_prompts

List available prompts that can be used with the server.