Manim MCP Server

A Model Context Protocol (MCP) server for compiling and serving Manim animations.

🎯 Two Server Modes

1. HTTP API Server (app/server.py) - For REST API calls, testing, and web integration

2. Standard MCP Server (mcp_server.py) - For Claude Desktop, Dify, and other MCP clients

See MCP_SETUP.md for detailed MCP configuration instructions.

A FastAPI-based MCP (Model Control Protocol) server that provides two main tools:

1. Manim Compile: Compile Manim code and return a video ID

2. Video Download: Download a compiled Manim video by ID

Features

• Secure authentication using JWT tokens

• LangGraph integration for workflow management

• Support for different video qualities and resolutions

• Simple API endpoints for integration

Prerequisites

• Python 3.8+

• Manim Community Edition (v0.19.0 or later)

• FFmpeg

• Required Python packages (see requirements.txt)

Installation

1. Clone the repository:

   git clone <repository-url> cd manim-mcp-server

2. Create a virtual environment and activate it:

   python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate

3. Install the required packages:

   pip install -r requirements.txt

4. Install Manim and its dependencies:

   pip install manim

Configuration

1. Set up environment variables (create a .env file):

   SECRET_KEY=your-secret-key-here ACCESS_TOKEN_EXPIRE_MINUTES=30

Running the Server

Option 1: Using the startup script (recommended)

Option 2: Using uvicorn directly

The server will be available at http://localhost:8000

API Documentation

Once the server is running, you can access the interactive API documentation at:

• Swagger UI: http://localhost:8000/docs

• ReDoc: http://localhost:8000/redoc

API Endpoints

Root

• GET / - Get server information and available tools

Manim Compilation

• POST /tools/manim_compile - Compile Manim code

  { "parameters": { "code": "from manim import *\nclass Example(Scene):\n def construct(self):\n circle = Circle()\n self.play(Create(circle))", "scene_name": "Example" } }

  Parameters:

  • code (required): The Manim Python code to compile

  • scene_name (required): Name of the specific scene class to compile

Video Download

• GET /videos/{file_id} - Download a compiled video by ID

LangGraph Compatible Endpoints

• GET /v1/tools - List all available tools

• POST /v1/tools/call - Call a tool (LangGraph compatible)

  { "tool": "manim_compile", "parameters": { "code": "from manim import *\nclass Example(Scene):\n def construct(self):\n circle = Circle()\n self.play(Create(circle))" } }

Example Usage

1. Check server status

2. Compile Manim code

3. Download the compiled video

4. Compile a specific scene by name

5. List available tools

6. Run the example script

Testing

See TESTING.md for detailed testing instructions.

Quick test:

Security

• Always use HTTPS in production

• Consider adding authentication for production deployments

• Validate and sanitize all user inputs

• Set appropriate CORS policies for your use case

License

This project is licensed under the MIT License - see the LICENSE file for details.