BigQuery MCP Server

This project provides a server that enables Large Language Models (LLMs) like Claude to interact directly with BigQuery data. It acts as a translator between the LLM and the database, facilitating secure and efficient communication using the Model Context Protocol (MCP).

The server allows users to query BigQuery data using natural language, eliminating the need to write SQL queries manually. It supports accessing tables and materialized views, exploring dataset schemas, and analyzing data within a 1GB query limit. Data access is read-only for security.

The server can be installed via Smithery or manually configured. It requires Node.js, a Google Cloud project with BigQuery enabled, and Claude Desktop. Configuration involves authenticating with Google Cloud and adding project details to the Claude Desktop config file. Command-line arguments are available for specifying the project ID, location, and service account key file. The project is under the MIT License and sponsored by OREDATA.