richardbaxterseo/gemini-mcp

Gemini MCP Server

A Model Context Protocol (MCP) server that provides seamless integration with Google's Gemini AI models. This server allows AI assistants like Claude to interact with Gemini models for text generation, image analysis, and more.

Features

• 💬 Chat with Gemini Models - Send messages to various Gemini models with customizable parameters
• 📝 List Available Models - Get detailed information about all available Gemini models
• 🖼️ Image Analysis - Analyze images using Gemini's vision capabilities
• 🔧 Configurable Parameters - Control temperature, max tokens, and system prompts
• 🛡️ Enhanced Error Handling - Robust server version with better content filtering management
• 🚀 Easy Setup - Simple installation and configuration process

Installation

Prerequisites

• Python 3.7 or higher
• A Google AI Studio API key (Get one here)

Setup

1. Clone the repository:

git clone https://github.com/richardbaxterseo/gemini-mcp.git
cd gemini-mcp

2. Install dependencies:

pip install -r requirements.txt

3. Set up your API key:

# Windows
set GEMINI_API_KEY=your-api-key-here

# macOS/Linux
export GEMINI_API_KEY=your-api-key-here

Configuration

For Claude Desktop

Add the following to your Claude Desktop configuration file:

Windows: %APPDATA%\Claude\claude_desktop_config.json macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "gemini": {
      "command": "python",
      "args": ["C:\\path\\to\\gemini-mcp\\server.py"],
      "env": {
        "GEMINI_API_KEY": "your-api-key-here"
      }
    }
  }
}

Using the Robust Server (Recommended)

For better error handling and content filtering management, use server_robust.py:

{
  "mcpServers": {
    "gemini": {
      "command": "python",
      "args": ["C:\\path\\to\\gemini-mcp\\server_robust.py"],
      "env": {
        "GEMINI_API_KEY": "your-api-key-here"
      }
    }
  }
}

The robust server provides:

• Enhanced error messages when content is filtered
• Relaxed safety settings to minimize false positives
• Better handling of various Gemini API response formats
• Actionable suggestions when responses are blocked

Available Tools

gemini_chat

Chat with Gemini models to generate text responses.

Parameters:

• message (required): The message to send to Gemini
• model (optional): Model to use (default: "gemini-2.5-flash")
• temperature (optional): Controls randomness 0-2 (default: 0.7)
• max_tokens (optional): Maximum tokens in response (default: 2048)
• system_prompt (optional): System instruction to guide the model

Example:

"Can you explain quantum computing in simple terms?"

gemini_list_models

List all available Gemini models and their capabilities.

Example:

"Show me all available Gemini models"

gemini_analyze_image

Analyze images using Gemini's vision capabilities.

Parameters:

• image_url (required): URL of the image to analyze
• prompt (optional): Question about the image (default: "What's in this image?")
• model (optional): Model to use (default: "gemini-2.5-flash")

Example:

"Analyze this image: https://example.com/image.jpg"

Troubleshooting

Common Issues

1. "GEMINI_API_KEY environment variable not set"

   • Make sure you've set your API key in the environment or in the Claude configuration

2. "No module named 'mcp'"

   • Run pip install -r requirements.txt to install all dependencies

3. Server not appearing in Claude

   • Restart Claude Desktop after updating the configuration
   • Check that the path in your configuration is correct
   • Ensure Python is in your system PATH

4. "Content was filtered" errors

   • Use server_robust.py for better handling of content filtering
   • Try rephrasing your query or using different parameters
   • Break complex queries into smaller, more specific questions

Debug Mode

To see detailed logs, you can run the server manually:

set GEMINI_API_KEY=your-api-key-here
python server.py

Testing the Server

To ensure the server is working correctly, try these test queries:

1. Basic conversation: "What is machine learning?"
2. Complex technical query: "Explain the differences between transformer and LSTM architectures"
3. Creative writing: "Write a short story about a robot learning to paint"
4. Code generation: "Create a Python function to calculate fibonacci numbers"
5. Content filtering test: If you encounter filtered content, the robust server will provide helpful feedback

Development

To contribute or modify the server:

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Test thoroughly
5. Submit a pull request

License

MIT License - see LICENSE file for details

Acknowledgments

• Built on the Model Context Protocol
• Uses Google's Gemini API
• Inspired by the MCP ecosystem

Support

For issues, questions, or contributions:

• Open an issue on GitHub
• Check existing issues for solutions
• Contribute improvements via pull requests