BigQuery

v1.0.0Databasesstable

Practical MCP server for large BigQuery datasets. Supports vector search. Keep LLM context small while staying fast and allowing only safe read-only actions.

bigquerybigquery-mlclaude-codecursordataform
Share:
8
Stars
0
Downloads
0
Weekly
0/5

What is BigQuery?

BigQuery is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to practical mcp server for large bigquery datasets. supports vector search. keep llm context small while staying fast and allowing only safe read-only actions.

Practical MCP server for large BigQuery datasets. Supports vector search. Keep LLM context small while staying fast and allowing only safe read-only actions.

This server falls under the Databases category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Practical MCP server for large BigQuery datasets. Supports v

Use Cases

Large dataset querying and analysis
Vector search integration
pvoo

Maintainer

LicenseMIT
Languagepython
Versionv1.0.0
UpdatedMar 19, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

NPM

npx -y mcp-server-bigquery

Manual Installation

npx -y mcp-server-bigquery

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use BigQuery

The BigQuery MCP Server is a production-ready, read-only MCP server designed for querying large Google BigQuery datasets without flooding LLM context windows. It features smart token-saving modes, per-query cost caps, vector search via BigQuery ML, and strict write-blocking guardrails — making it safe and efficient for analytical AI workflows.

Prerequisites

  • Python 3.10 or later with uv or uvx installed (`pip install uv`)
  • An active Google Cloud project with BigQuery API enabled
  • Google Cloud authentication configured: run `gcloud auth application-default login`
  • Sufficient BigQuery IAM permissions (BigQuery Data Viewer and BigQuery Job User roles)
  • Claude Desktop or another MCP-compatible client
1

Authenticate with Google Cloud

Set up Application Default Credentials so the server can access your BigQuery project without a service account key file.

gcloud auth application-default login
2

Test the server with uvx

Run the server directly using uvx (no install needed). Replace YOUR_PROJECT with your GCP project ID.

uvx bigquery-mcp --project YOUR_PROJECT --location US
3

Configure environment variables (optional)

Set optional environment variables to restrict datasets, control cost caps, and tune result sizes.

export BIGQUERY_ALLOWED_DATASETS="dataset1,dataset2"
export BIGQUERY_MAX_BYTES_BILLED="5000000000"
export BIGQUERY_LIST_MAX_RESULTS="500"
export BIGQUERY_SAMPLE_ROWS="3"
4

Add the server to Claude Desktop config

Open ~/Library/Application Support/Claude/claude_desktop_config.json and add the bigquery server entry.

5

Restart Claude Desktop and verify

Quit and reopen Claude Desktop. The BigQuery server should appear as a connected MCP tool. Ask Claude to list your datasets to confirm the connection.

BigQuery Examples

Client configuration

Claude Desktop config using uvx to launch the BigQuery MCP server with your project ID.

{
  "mcpServers": {
    "bigquery": {
      "command": "uvx",
      "args": [
        "bigquery-mcp",
        "--project", "your-gcp-project-id",
        "--location", "US"
      ]
    }
  }
}

Prompts to try

Analytical queries you can run via Claude once the server is connected.

- "List all datasets in my BigQuery project and show their descriptions."
- "Show me the schema of the `sales.transactions` table with sample rows."
- "Run a query to count daily orders in the last 30 days from the orders table."
- "Search for rows in the products table semantically similar to 'wireless headphones'."

Troubleshooting BigQuery

Authentication error: Application Default Credentials not found

Run `gcloud auth application-default login` and follow the browser prompt. If running in CI or a container, set GOOGLE_APPLICATION_CREDENTIALS to the path of a service account JSON key file.

Query blocked with write operation error

The server intentionally blocks all non-SELECT statements. Only read queries (SELECT, WITH) are permitted. Rewrite your query to use SELECT instead of INSERT, UPDATE, or DELETE.

Cost cap exceeded error when running queries

Increase BIGQUERY_MAX_BYTES_BILLED in your environment or MCP config env block. The default cap is roughly $0.50 per query. For large tables, add a LIMIT clause to your query to reduce bytes scanned.

Frequently Asked Questions about BigQuery

What is BigQuery?

BigQuery is a Model Context Protocol (MCP) server that practical mcp server for large bigquery datasets. supports vector search. keep llm context small while staying fast and allowing only safe read-only actions. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install BigQuery?

Install via npm with the command: npx -y mcp-server-bigquery. Then add the server configuration to your AI client's JSON config file (e.g., claude_desktop_config.json or .cursor/mcp.json).

Which AI clients work with BigQuery?

BigQuery works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is BigQuery free to use?

Yes, BigQuery is open source and available under the MIT license. You can use it freely in both personal and commercial projects.

Browse More Databases MCP Servers

Explore all databases servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "bigquery": { "command": "npx", "args": ["-y", "mcp-server-bigquery"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use BigQuery?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides