Google Drive

Overview

Google Drive is a Model Context Protocol server that gives AI systems direct access to Google Drive. With it, any MCP-compatible AI assistant can list, search, upload, download, and manage files and folders in your Google Drive—the model decides when and how to use these capabilities in conversation or agent workflows.

The server runs locally via the npm package google-drive-mcp, which communicates over stdio. You'll need to provide three environment variables to authenticate and configure it: GOOGLE_ACCESS_TOKEN, MCP_TRANSPORT, and GOOGLE_CLIENT_ID. The latest release is v1.3.0, maintained at domdomegg/google-drive-mcp on GitHub.

Google Drive belongs in the File System category on mcp.site and works with MCP clients including Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, Zed, and others across the ecosystem. If you maintain this server, claim your listing to verify ownership and keep the information current.

Installation

// claude_desktop_config.json
{
  "mcpServers": {
    "google_drive": {
      "command": "npx",
      "args": [
        "-y",
        "google-drive-mcp"
      ],
      "env": {
        "GOOGLE_ACCESS_TOKEN": "<google-access-token>",
        "MCP_TRANSPORT": "<mcp-transport>",
        "GOOGLE_CLIENT_ID": "<google-client-id>",
        "GOOGLE_CLIENT_SECRET": "<google-client-secret>"
      }
    }
  }
}

Requires GOOGLE_ACCESS_TOKEN, MCP_TRANSPORT, GOOGLE_CLIENT_ID