Your AI agents can now control your Google Home devices

TechCrunch reports that Google has opened early access to a Model Context Protocol (MCP) server for the Google Home ecosystem, enabling AI agents such as Claude, ChatGPT, Hermes, OpenClaw and Google Antigravity to interact with users' smart home devices. The integration lets supported agents control connected hardware, review camera summaries and query smart home event history using natural language.

By AI NewsroomPublished about 12 hours agoUpdated about 12 hours ago0 views
Your AI agents can now control your Google Home devices

Why It Matters

By linking third-party AI agents to Google Home, users can delegate everyday smart home tasks and monitoring to conversational systems, potentially simplifying device control and automation. Google is starting the rollout with a paid tier, which will shape who gains immediate access and how broadly the feature is tested.

Key Facts

  • Feature: Model Context Protocol (MCP) server for Google Home early access
  • Supported AI agents: Claude, Hermes, OpenClaw, ChatGPT, Google Antigravity (agents that implement MCP)
  • Capabilities: Control connected devices, review camera summaries, monitor smart home activity, access event history, build custom dashboards
  • Setup steps: Create a Google Cloud project, configure it for Home MCP, provide MCP configuration to chosen agent, sign in and grant permissions
  • Device support: Any device in Google Home ecosystem, including Google Nest doorbells and thermostats and 'Works with Google Home'/Matter devices

Google has begun offering early access to a Model Context Protocol (MCP) server for the Google Home platform, allowing AI agents that implement MCP to work directly with users' smart home setups. According to the report, agents such as Claude, ChatGPT, Hermes, OpenClaw and Google Antigravity will be able to perform tasks like controlling devices, pulling up camera summaries and querying a home's event history through conversational commands. The aim is to let people use natural language to manage routine smart-home activities and construct personalized dashboards.

To enable the connection, users must create and configure a Google Cloud project to use Home MCP, then share those MCP configuration details with their chosen AI agent. The agent takes over the rest of the setup flow, prompting the user to sign in and grant the necessary permissions. Google plans to publish a setup guide in the Google Home Developer Center to assist users during the process.

The MCP integration is designed to work with the full range of Google Home-compatible equipment, from Google Nest doorbells and thermostats to devices that carry the 'Works with Google Home' designation or use the Matter standard. Google already uses MCP across other offerings such as Google Cloud services, developer tools and Google Workspace; extending it to Google Home is positioned as a consumer-facing move to support experimentation with AI agents for everyday tasks.

Access to the Google Home MCP will begin rolling out immediately to U.S. subscribers of Google Home Premium Advanced, the platform's $20-per-month tier that includes expanded video history and enhanced alerts. Google declined to say whether MCP access will be extended to other subscription levels or regions. During the early access period, the company is requesting feedback from participants via its Smart Home for Developers Community.

Keep Reading