fermat-mcp

🚀 Fermat MCP: The Ultimate Math Engine - Unifying SymPy, NumPy & Matplotlib in one powerful server! Perfect for devs & researchers.

20 stars PythonOthers Updated Jul 29, 2026

mathematicsmatplotlibmcpmcp-servernumerical-computationnumpysymbolic-computationsympy

Documentation

Fermat MCP

This project provides a FastMCP server for mathematical computations, including numerical and symbolic calculations, as well as plotting.

Modules

1. mpl_mcp - Matplotlib Integration

Feature	Description
`plot_barchart`	Plots bar charts of given data values
`plot_scatter`	Creates scatter plots from data points
`plot_chart`	Plots line, scatter, or bar charts
`plot_stem`	Creates stem plots for discrete data
`plot_stack`	Generates stacked area/bar charts
`eqn_chart`	Plots mathematical equations

2. numpy_mcp - NumPy Integration

Category	Operations
Basic Math	add, sub, mul, div, power, abs, exp, log, sqrt
Trigonometric	sin, cos, tan
Statistics	mean, median, std, var, min, max, argmin, argmax, percentile
Linear Algebra	dot, matmul, inv, det, eig, solve, svd
Matrix Operations	create, zeros, ones, full, arange, linspace
Array Manipulation	reshape, flatten, concatenate, transpose, stack

3. sympy_mcp - SymPy Integration

Category	Operations
Algebra	simplify, expand, factor, collect
Calculus	diff, integrate, limit, series
Equations	solve, solveset, linsolve, nonlinsolve
Matrix Operations	create, det, inv, rref, eigenvals

Setup

Requirements

Python 3.12 or higher (To install Python3.12 follow Python Download)

uv (To install uv follow uv Installation)

Clone the repository

bash

git clone https://github.com/abhiphile/fermat-mcp

Visual Studio Code, Windsurf

You can find the `mcp.json` file in the

MCP: Open User Configuration or MCP: Open Workspace Configuration

Add the following to your `mcp.json`:

json

{
  "mcpServers": {
    "fmcp": {
      "command": "bash",
      "args": ["MCP_SERVER_ABSOLUTE_PATH/setup.sh"],
      "description": "fmcp server is for mathematical computations, including numerical and symbolic calculations, as well as plotting."
    }
  }
}

Claude (Anthropic)

If you're using Claude or the Anthropic MCP client, add this working MCP configuration to your `mcp.json` (update the directory path to your local clone):

json

{
  "mcpServers": {
    "fmcp": {
      "command": "uv",
      "args": [
        "--directory",
        "/home/ty/Repositories/fermat-mcp",
        "run",
        "server.py"
      ]
    }
  }
}

Gemini CLI

Open your Gemini settings JSON located in ~/.gemini/settings.json where ~ is your home directory.

Add the following to your settings.json:

json

{
  "mcpServers": {
    "fmcp": {
      "command": "bash",
      "args": ["MCP_SERVER_ABSOLUTE_PATH/setup.sh"],
      "description": "fmcp server is for mathematical computations, including numerical and symbolic calculations, as well as plotting."
    }
  }
}

Installing via Smithery

To install Fermat MCP for local usage automatically via Smithery:

bash

npx -y @smithery/cli install @abhiphile/fermat-mcp --client gemini

Example Usage

Using Gemini CLI

code

╭──────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│  > Can you use fmcp server and using numpy method find the eigen values of this 8*8 matrix,                  |
│    2 1 3 1 1 8 4 2                                                                                           |
│    6 6 0 7 1 4 6 1                                                                                           │
│    9 2 1 8 7 9 9 0                                                                                           │
│    2 5 6 6 9 8 0 1                                                                                           │
│    1 3 6 2 3 8 8 1                                                                                           │
│    9 4 2 2 1 2 2 9                                                                                           │
│    8 6 4 4 2 0 2 8                                                                                           │
│    0 0 0 6 6 7 5 6                                                                                           │
╰──────────────────────────────────────────────────────────────────────────────────────────────────────────────╯

 ╭─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
 │ ✔  numpy_mcp_numerical_operation (fmcp MCP Server) {"a":[[2,1,3,1,1,8,4,2],[6,6,0,7,1,4,6,1],[9,2,1,8,7,9,9,0],[2,5,6,6,9,8,0,1],[1,3,… │
 │                                                                                                                                         │
 │    {"eigenvalues":["32.077244457548815+0j","-11.531090644775198+0j","-6.6653982146786195+0j","0.6715984762411508+3.37024850             │
 │    10270413j","0.6715984762411508-3.3702485010270413j","4.541270555490195+2.776364664923869j","4.541270555490195-2.77636466             │
 │    4923869j","3.6935063384423428+0j"],"eigenvectors":[["-0.23263835483680192+0j","-0.2264723575289234+0j","-0.4308391916391             │
 │    0195+0j","-0.012346573390129022+0.17748655663058255j","-0.012346573390129022-0.17748655663058255j","-0.21221572277027187             │
 │    +0.3524396218277479j","-0.21221572277027187-0.3524396218277479j","0.3451499664861578+0j"],["-0.31955742545335186+0j","-0             │
 │    .2569860493445581+0j","0.05691886770041556+0j","-0.35591013681869693-0.2242364092694275j","-0.35591013681869693+0.224236             │
 │    4092694275j","0.1932161673963751-0.39527849111641133j","0.1932161673963751+0.39527849111641133j","-0.7979681696063214+0j             │
 │    "],["-0.46626263247473404+0j","-0.4684914620112376+0j","0.5469400556350749+0j","0.34325164099973565+0.06607019711949293j             │
 │    ","0.34325164099973565-0.06607019711949293j","0.21312270185159682+0.28822307710358636j","0.21312270185159682-0.288223077             │
 │    10358636j","0.42707422750984786+0j"],["-0.41589316441674523+0j","0.2291771012892302+0j","0.09410792992600435+0j","0.6375             │
 │    92441360358+0j","0.637592441360358+-0j","0.46446646137729414+0j","0.46446646137729414+-0j","0.08171661775583623+0j"],["-             │
 │    0.35812884189789035+0j","-0.26551071423139044+0j","-0.649979374400915+0j","-0.2999153430497845+0.20110182336747695j","-0             │
 │    .2999153430497845-0.20110182336747695j","0.04424367733961273-0.020400357434930304j","0.04424367733961273+0.0204003574349             │
 │    30304j","-0.037276545325188024+0j"],["-0.3042416233100104+0j","0.4027405071320744+0j","0.2343447346668298+0j","-0.119224             │
 │    87219398595+0.10317726668781271j","-0.11922487219398595-0.10317726668781271j","-0.3387759141362585+0.15790241128043445j"             │
 │    ,"-0.3387759141362585-0.15790241128043445j","0.11473144194899779+0j"],["-0.35389732755448017+0j","0.5359362131480812+0j"             │
 │    ,"0.13989902586074168+0j","-0.16008513226461316-0.26111873108374495j","-0.16008513226461316+0.26111873108374495j","0.055             │
 │    729749650881726-0.22748184206919145j","0.055729749650881726+0.22748184206919145j","-0.19838514696566653+0j"],["-0.327615             │
 │    5213691367+0j","-0.30122843153433887+0j","0.07858579556736099+0j","0.00042251379361638556-0.11670182367986544j","0.00042             │
 │    251379361638556+0.11670182367986544j","-0.18590228871569267-0.2479130526139182j","-0.18590228871569267+0.247913052613918             │
 │    2j","-0.03374593856285834+0j"]]}                                                                                                     │
 ╰─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
✦ Of course, I can do that. Here are the eigenvalues for the given matrix:

  32.077244457548815
  -11.531090644775198
  -6.6653982146786195
  0.6715984762411508 + 3.3702485010270413j
  0.6715984762411508 - 3.3702485010270413j
  4.541270555490195 + 2.776364664923869j
  4.541270555490195 - 2.776364664923869j
  3.6935063384423428

Using Visual Studio Code, Windsurf

Here the task was to plot both x**2 and sin(x) in the same plot.

Frequently asked questions

What is fermat-mcp?

fermat-mcp is 🚀 Fermat MCP: The Ultimate Math Engine - Unifying SymPy, NumPy & Matplotlib in one powerful server! Perfect for devs & researchers.

How do I install fermat-mcp?

Open the GitHub repository and follow its README. Most MCP servers are added to your client's MCP config, then called by your agent.

Is fermat-mcp open source?

Yes — it is hosted on GitHub at https://github.com/abhiphile/fermat-mcp and has 20 stars.

Related MCP tools

semble

Fast and Accurate Code Search for Agents. Uses 99% fewer tokens than grep+read

5,993 Python

agentscode-searchembeddings+4

google_workspace_mcp

Control Gmail, Google Calendar, Docs, Sheets, Slides, Chat, Forms, Tasks, Search & Drive with AI - Comprehensive Google Workspace MCP Server & CLI Tool

3,117 Python

aigmailgoogle-calendar+17

jcodemunch-mcp

Cut AI token costs 95%+ on code exploration. The leading MCP server for precise, symbol-level GitHub code retrieval via tree-sitter AST. Works with Claude Code, Cursor & any MCP client. 313B+ tokens saved.

2,651 Python

claudeclaude-codeai-coding+17

paperbanana

Open source implementation and extension of Google Research’s PaperBanana for automated academic figures, diagrams, and research visuals, expanded to new domains like slide generation.

2,302 Python

academic-researcharxivgemini+17

pg-aiguide

MCP server and Claude plugin for Postgres skills and documentation. Helps AI coding tools generate better PostgreSQL code.

1,834 Python

aiai-agentsai-coding+13

MiniMax-MCP

Official MiniMax Model Context Protocol (MCP) server that enables interaction with powerful Text to Speech, image generation and video generation APIs.

1,576 Python

image-generationimage-to-videomcp+7

Run your own MCP server? See who uses it and what to fix.

Measure it with TrackMCP