Google's new managed servers make AI integration way easier
Google just dropped its fully managed Model Context Protocol (MCP) servers, letting AI tools plug straight into Google Cloud services—no messy setup needed.
Right now, you can use them with Maps, BigQuery, Compute Engine, and Kubernetes Engine.
If you're an enterprise customer, the public preview is free for now and gives you real-time access to Google's data and infrastructure.
Seamless connections for your favorite AI apps
Built on Anthropic's open-source standard, these MCP servers let popular AI clients like Gemini CLI, Claude, ChatGPT, and AI Studio talk directly to Google tools.
Security is front and center: access is controlled by Google Cloud IAM and Model Armor keeps threats like prompt injection at bay.
Extra security with Apigee integration—and more coming soon
The servers also work with Apigee API management so businesses keep their usual governance and audit controls in place.
Google says support for storage, databases, monitoring, and security services will roll out next.
Making it simple (and safe) to scale up AI at work
With these MCP servers—free during preview for existing enterprise customers who already pay for Google services—companies get a secure way to connect APIs and ramp up their use of real-time data in business processes.
General release is expected very soon in the new year.