BigQuery MCP Server

Name: BigQuery MCP Server
Author: takuya0206

@takuya0206

Repository

About BigQuery MCP Server

No overview available yet

Basic information

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "bigquery-mcp-server": {
      "command": "bun",
      "args": [
        "run",
        "build"
      ]
    }
  }
}

Tools

No tools detected

We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.

Overview

What is BigQuery MCP Server?

A Model Context Protocol (MCP) server that enables Large Language Models to access Google BigQuery. It supports dataset discovery, schema exploration, and read-only SQL query execution.

How to use BigQuery MCP Server?

Install via local build (bun install, bun run build) or Docker. Configure the server with --project-id, --location, --max-results, --max-bytes-billed, and optionally a --key-file for service account authentication. Add the server to your MCP client configuration with the path to the binary and arguments.

Key features of BigQuery MCP Server

Execute read-only SELECT queries with configurable limits
List all datasets in a project
List tables with schemas for a given dataset
Retrieve table schema and sample data (up to 20 rows)
Dry-run queries to estimate cost without execution
Supports Application Default Credentials or service account key files

Use cases of BigQuery MCP Server

LLMs exploring BigQuery dataset structures and schemas
Running safe, read-only SQL queries via natural language
Estimating query costs before executing large queries
Inspecting partitioned tables and their partitions
Programmatic dataset and table discovery within a project

FAQ from BigQuery MCP Server

What authentication methods are supported?

The server supports Application Default Credentials (ADC) and service account key files. ADC is set up via gcloud auth application-default login.

What BigQuery permissions are required?

The recommended role is roles/bigquery.user. Alternatively, both roles/bigquery.dataViewer and roles/bigquery.jobUser are needed.

Are write queries allowed?

No. Only SELECT queries are permitted; the server performs security checks to reject non-SELECT statements.

What is the default query processing limit?

A default limit of 500 GB is set for query processing to prevent excessive costs. This can be changed via --max-bytes-billed.

Can I run the server in a Docker container?

Yes. Build the Docker image with docker build -t bigquery-mcp-server . and run it with docker run -it --rm bigquery-mcp-server --project-id=your-project-id.