stable-diffusion-mcp-server by hkhkkh - MCP Server

🎨 Stable Diffusion MCP Server

An enhanced Model Context Protocol (MCP) server for Stable Diffusion, providing comprehensive AI art generation capabilities through a simple, intuitive interface.

✨ Features

🎯 Core AI Art Generation

Text-to-Image (txt2img): Generate images from text descriptions
Image-to-Image (img2img): Transform existing images with AI
Prompt Analysis: Reverse-engineer prompts from images (interrogate)

🔧 Advanced Capabilities

Model Management: List, switch, and manage SD models dynamically
LoRA Support: Apply LoRA models for enhanced artistic effects
Smart Recommendations: AI-driven model and parameter suggestions
Auto-Launch: Automatic detection and startup of SD WebUI
Multi-language Support: Works with Chinese and English prompts

🛠️ MCP Tools Available

Tool Name	Description	Key Parameters
`generate_image`	Text-to-image generation	prompt, model, lora, dimensions
`img2img`	Image-to-image transformation	image_path, prompt, denoising_strength
`interrogate_image`	Extract prompts from images	image_path, model
`list_models`	List available SD models	-
`list_loras`	List available LoRA models	-
`switch_model`	Change current SD model	model_name
`check_sd_status`	Check SD WebUI status	-
`start_sd_webui`	Launch SD WebUI	-

🚀 Quick Start

Prerequisites

Python 3.8+
Stable Diffusion WebUI (AUTOMATIC1111 or compatible)
SD WebUI running on http://127.0.0.1:7860 (default)

Installation

Clone the repository

git clone https://github.com/hkhkkh/stable-diffusion-mcp-server.git
cd stable-diffusion-mcp-server

Install dependencies
```
pip install -r requirements.txt
```
Configure your SD WebUI path (optional) Edit sd_mcp_server.py and update the SD_LAUNCHER_PATH variable to point to your SD WebUI executable.
Start the MCP server
```
python sd_mcp_server.py
```

Usage Examples

Basic Image Generation

Generate a cute cat sitting in a garden
Create a cyberpunk cityscape at night

Advanced Generation with Model/LoRA

Using realistic model, draw a portrait with lora:detailed_face:0.8
Generate anime style artwork using anything model

Image-to-Image

Transform this image: C:\path\to\image.jpg into anime style
Based on this photo, create a fantasy version

Model Management

List all available SD models
Switch to majicmix_v7 model
Show current model information

🎨 Supported Models

The server includes intelligent model recommendations based on content type:

Realistic Models: majicmix (v2, v7), chilloutmix, realistic series
Anime Models: anything, animevae, cartoon styles
Specialized Models: Architecture, landscapes, portraits
LoRA Support: Facial details, style enhancements, character-specific

📁 Project Structure

stable-diffusion-mcp-server/
├── sd_mcp_server.py          # Main MCP server
├── requirements.txt          # Python dependencies
├── generated_images/         # Output directory
├── demo_ai_prompt_writing.py # AI prompt enhancement demo
├── test_*.py                # Test scripts
└── docs/                    # Additional documentation

🔧 Configuration

SD WebUI Settings

Default API URL: http://127.0.0.1:7860
Output directory: generated_images/
Auto-launch support for Windows SD WebUI

Model Database

The server includes a comprehensive model knowledge base for intelligent recommendations:

Model strengths and weaknesses
Optimal use cases
Tag associations
Quality ratings

🧪 Testing

Run the included test scripts to verify functionality:

python test_smart_generate.py    # Test smart generation
python test_current_model.py     # Test model info
python test_recommendations.py   # Test model recommendations

📖 Documentation

🤝 Contributing

Contributions are welcome! Please read the for licensing terms.

For Non-Commercial Use

Fork the repository
Create a feature branch
Submit a pull request

For Commercial Use

Contact the author for commercial licensing
Revenue sharing agreement required

📄 License

This project uses a dual licensing model:

Non-Commercial: CC BY-NC-SA 4.0 (free for personal/educational use)
Commercial: Custom license with revenue sharing requirement

See for full details.

📞 Contact

Author: Yang Haokang
GitHub: @hkhkkh
Email: 0221123105@mail.imu.edu.cn

For commercial licensing inquiries, please open an issue or contact directly.

🌟 Acknowledgments

Built on the Model Context Protocol (MCP)
Compatible with AUTOMATIC1111 Stable Diffusion WebUI
Inspired by the open-source AI art community

Made with ❤️ for the AI art community