tadata-org/fastapi_mcp is a Python library that exposes FastAPI endpoints as Model Context Protocol (MCP) tools. MCP is the protocol that lets AI applications connect to external tools and data under a defined set of rules. The library can either mount an MCP server inside a FastAPI app or be deployed on its own, and it uses the ASGI interface directly. Mounting means attaching the server to a specific path inside the app so that both are handled in the same process. The latest release of the repository is v0.4.0, and the source is the tadata-org/fastapi_mcp repository on GitHub.
The key point is that it reuses existing FastAPI dependencies as they are. Authentication and authorization follow FastAPI's own conventions, and request/response schemas and the Swagger endpoint documentation are preserved as well. In other words, it exposes the list of functions an AI application can call as tools, without requiring you to redefine schemas and authentication rules you already use. The license is MIT, the primary language is Python, and the GitHub star count is 12,022.
That said, the release notes for v0.4.0 are not included in this material. To know what changed in this version, you have to check the repository's releases page directly. No independent performance measurements or real-world adoption cases are presented either. Considering that the repository was created on March 8, 2025 and is a relatively recent project, it is worth a look first for teams that already have an API, but there is not yet enough basis to apply it straight away to large-scale production environments.