Fermat MCP is a mathematical computation server project based on FastMCP, designed to integrate symbolic computation, numerical computation, and plotting into a single service. The project consists of three integrated modules: mpl_mcp handles Matplotlib integration, numpy_mcp handles NumPy integration, and sympy_mcp handles SymPy integration.
In terms of use, this project is suitable for developers and researchers performing mathematical computation and visualization tasks in MCP-capable clients, such as symbolic algebra, numerical operations, and chart plotting.
Main capabilities include: symbolic computation via SymPy, numerical computation via NumPy, and plotting via Matplotlib. The example task given in the project description is to plot x**2 and sin(x) in the same figure.
Usage: the runtime environment requires Python 3.12 or later, as well as uv; clients such as Visual Studio Code and Windsurf are supported. During configuration, the mcp.json file can be found in MCP: Open User Configuration or MCP: Open Workspace Configuration, and the corresponding configuration can be added to it; for Claude (Anthropic), configuration must be added in mcp.json and the directory path updated to the local clone path; for Gemini, the settings file located at ~/.gemini/settings.json must be edited. In addition, automatic installation via Smithery is supported for local use, and a usage method for Gemini CLI is also provided.
Limitations: the material does not provide further information such as fees, pricing, platform support scope, or performance metrics, and none is currently provided.