minimax-mcp: Instructions file for Claude Code

MiniMax MCP Server

A Model Context Protocol (MCP) server that provides tools for web search, image understanding, text-to-speech, and image generation using the MiniMax API.

Tools

minimax_web_search

Search the web using MiniMax's search API.

• Parameters: query (string), response_format (markdown|json)
• Returns: Search results with titles, snippets, and links

minimax_understand_image

Analyze images using MiniMax's vision model. Calls /v1/coding_plan/vlm directly via HTTP — no subprocess, no external dependencies.

• Parameters: prompt (string), image_source (URL or local path), response_format
• Returns: Detailed image analysis
• Supported formats: JPEG, PNG, WebP

minimax_text_to_speech

Generate speech audio from text.

• Parameters: text (string, max 5000), voice_id (enum), speed (0.5-2.0)
• Returns: Path to generated MP3 file
• Recommended voice: English_expressive_narrator (others may not be available on all plans)

minimax_generate_image

Generate images from text descriptions.

• Parameters: prompt (string, max 1500), aspect_ratio (enum), n (1-9)
• Returns: Local path(s) to generated image file(s)

Setup

1. Install dependencies

cd /path/to/minimax-mcp
pip install -r requirements.txt

2. Configure API key

Set the MINIMAX_API_KEY environment variable:

export MINIMAX_API_KEY="your_api_key_here"

Or add it to your shell profile (~/.zshrc or ~/.bashrc).

Windows Installation

On Windows, use the same pip command but set the environment variable differently:

1. Install dependencies

pip install -r requirements.txt

2. Set your API key

set MINIMAX_API_KEY=your_api_key_here

Or set permanently via System Properties → Environment Variables.

3. Add to Claude Code Edit %APPDATA%\Claude\settings.json and add:

{
  "mcpServers": {
    "minimax": {
      "command": "python",
      "args": ["C:\\path\\to\\minimax-mcp\\minimax_mcp.py"],
      "env": {
        "MINIMAX_API_KEY": "your_api_key_here"
      }
    }
  }
}