Glm Vision

A Model Context Protocol (MCP) server that integrates GLM-4.5V from Z.AI with Claude Code.

Features

• Image Analysis: Analyze images using GLM-4.5V's vision capabilities
• Local File Support: Analyze local image files or URLs
• Configurable: Easy setup with environment variables

Installation

Prerequisites

• Python 3.10 or higher
• GLM API key from Z.AI
• Claude Code installed

Setup

1. Clone or create the project directory:

   cd /path/to/your/project
   

2. Create and activate virtual environment:

   python3 -m venv env
   source env/bin/activate  # On Windows: env\Scripts\activate
   

3. Install dependencies:

   pip install -r requirements.txt
   # or with uv (recommended)
   uv pip install -r requirements.txt
   

4. Set up environment variables:

   cp .env.example .env
   # Edit .env with your GLM API key from Z.AI
   

5. Add the server to Claude Code:

   # Using uv (recommended)
   uv run mcp install -e . --name "GLM Vision Server"
   
   # Or manually add to Claude Desktop configuration:
   claude mcp add-json --scope user glm-vision '{
     "type": "stdio",
     "command": "/path/to/your/project/env/bin/python",
     "args": ["/path/to/your/project/glm-vision.py"],
     "env": {"GLM_API_KEY": "your_api_key_here"}
   }'
   

Configuration

Set these environment variables in your .env file:

Variable	Description	Default
GLM_API_KEY	Your GLM API key from Z.AI	(required)
GLM_API_BASE	GLM API base URL	https://api.z.ai/api/paas/v4
GLM_MODEL	Model name to use	glm-4.5v

Usage

Available Tools

glm-vision

Analyze an image file using GLM-4.5V's vision capabilities. Supports both local files and URLs.

Parameters:

• image_path (required): Local file path or URL of the image to analyze
• prompt (required): What to ask about the image
• temperature (optional): Response randomness (0.0-1.0, default: 0.7)
• thinking (optional): Enable thinking mode to see model's reasoning process (default: false)
• max_tokens (optional): Maximum tokens in response (max 64K, default: 2048)

Example:

Use the glm-vison tool with:
- image_path: "/path/to/your/image.jpg"
- prompt: "Describe what you see in this image"

Testing

Test the server using the MCP Inspector:

# With uv
uv run python glm-vision.py

# Or with python
python glm-vision.py

Development

Running Tests

# Install development dependencies
pip install -e ".[dev]"

# Run tests
pytest

# Format code
black .
isort .

# Type checking
mypy glm-vision.py

Troubleshooting

1. API Key Issues: Make sure your GLM_API_KEY is correctly set in the environment
2. Connection Problems: Check your internet connection and API endpoint
3. Model Errors: Verify that the model name (GLM_MODEL) is correct and available

License

MIT License - see LICENSE file for details.

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

Support

For issues related to the GLM API, contact Z.AI support. For MCP server issues, please create an issue in the repository.