Turn REST APIs into AI-agent tools without building another MCP server
A major challenge in enterprise Agentic AI is connecting agents to existing business capabilities locked behind REST APIs.
Google Cloud’s API Gateway MCP support (Public Preview) offers a practical approach:
REST API → OpenAPI → API Gateway → MCP → AI Agent
Instead of creating and maintaining a separate MCP server, teams can annotate an existing OpenAPI 3.x specification and expose selected REST operations as MCP tools.
The gateway handles:
🔹 MCP JSON-RPC requests
🔹 REST request translation
🔹 Authentication and authorization
🔹 Quotas and policies
🔹 Logging and observability
🔹 MCP response translation
This creates a shared policy path for REST and MCP traffic, helping enterprises expose existing APIs to AI agents while preserving gateway controls.
The key architectural insight: MCP doesn’t necessarily mean another service to build and operate. API Gateway can act as the bridge between enterprise APIs and agentic applications.
The result is a simpler path from API-first architectures to agent-ready enterprises.
📌 Source: https://developers.googleblog.com/turn-your-rest-apis-into-mcp-tools-with-google-cloud-api-gateway/
A major challenge in enterprise Agentic AI is connecting agents to existing business capabilities locked behind REST APIs. Google Cloud’s API Gateway MCP support (Public Preview) offers a practical approach: REST API → OpenAPI → API Gateway → MCP → AI Agent Instead of creating and maintaining a separate MCP server, teams can annotate an…