September 24, 2025

Google has announced the public release of the Data Commons Model Context Protocol (MCP) Server, a major step in making Data Commons’ extensive public datasets instantly accessible for AI developers, data scientists, and organizations worldwide.

The MCP Server is designed to bring real-world statistical data directly into AI workflows, helping reduce Large Language Model (LLM) hallucinations by anchoring responses in verified, structured data.

Key Benefits for Developers

• Native integration: AI agents can now consume Data Commons directly, without requiring developers to navigate complex APIs.

• Accelerated development: Enables faster creation of data-rich, agentic applications that deliver trustable, sourced information.

• End-to-end support: Agents can handle exploratory, analytical, and generative queries seamlessly.

Real-World Use Case: ONE Data Agent

In partnership with the ONE Campaign, Google’s Data Commons powered the ONE Data Agent, an interactive tool for health financing data.

• Users can search tens of millions of health financing data points in seconds using plain language.

• The tool simplifies discovery, visualization, and dataset downloads for advocacy, reporting, and policy-making.

Getting Started

The MCP Server is designed for seamless integration into Google Cloud’s agent development ecosystem:

• Works with the Agent Development Kit (ADK) and Gemini CLI.

• Can integrate into other MCP-compatible workflows.

Resources available to developers:

• [PyPi package] to install MCP for Gemini CLI or other clients

• [Colab notebook] with an ADK sample agent

• [GitHub repository] with code samples

With this release, Google is making trustable, sourced public data readily available to developers, opening the door to more reliable, impactful AI agents across industries.