Google Analytics MCP is the official Model Context Protocol server from the Google Analytics team. It lets LLM-powered assistants query your Google Analytics data in natural language, turning questions about traffic, funnels and realtime activity into proper GA API reports.
The server is open source on GitHub with about 3,520 stars and updates in 2026, and it is currently labeled experimental. Running locally, it connects MCP clients such as AI assistants and agent frameworks to the Google Analytics Admin API and Data API. Its tools retrieve account summaries and property details, list Google Ads links, fetch custom dimensions and metrics, and run core reports, funnel reports and realtime reports. Authentication uses standard Google credentials, and because the server runs on your own machine, your analytics access tokens never leave your environment. Setup instructions and a community Discord channel are provided in the repository. It works with any MCP-compatible client, including desktop AI assistants and agent frameworks, and pairs with the Analytics team’s other tooling for property management, making it a practical bridge between conversational AI and day-to-day analytics work. Detailed setup guides cover both local credential files and application default credentials.
Google Analytics MCP is free and open source under the Apache-2.0 license. It uses the standard Google Analytics APIs, which are free within their normal quotas.
Marketers, analysts and product teams who use MCP-compatible AI assistants and want to ask questions about their Google Analytics data conversationally instead of building reports by hand.