MCPHIVE

BigQuery MCP

Local · stdio

BigQuery is Google Cloud's serverless data warehouse for petabyte-scale datasets, one of the databases supported by Google's official MCP Toolbox for Databases — a unified toolkit (PostgreSQL, MySQL, Spanner, AlloyDB) that runs as an MCP server with prebuilt BigQuery tools. It gives agents SQL, schema inspection, catalog search, and forecasting. Auth is via Google Cloud ADC or a service account.

Why this server

  • bigquery-execute-sql runs queries against petabyte-scale datasets right from chat
  • bigquery-list-dataset-ids and bigquery-get-table-info describe the schema with no console visit
  • bigquery-forecast builds a forecast from historical time series in a single tool call
  • The same MCP Toolbox also supports AlloyDB, Cloud SQL, Spanner and Firestore

Usage examples

Prompt: What datasets and tables do we have in the analytics-prod project?

Result: The agent called bigquery-list-dataset-ids, then bigquery-list-table-ids and bigquery-get-table-info — mapping the whole project without opening the BigQuery console.

Prompt: Forecast next quarter's revenue based on 2 years of data

Result: bigquery-forecast built the forecast from the historical revenue series, with bigquery-execute-sql first preparing the aggregated table for the model.

Local server

This MCP runs locally on the user's machine (stdio), so availability isn't tracked — rely on reviews and the rating instead.

Reviews

No reviews yet.