AI Image Generation Server with MCP Interface

What is AI Image Generation Server with MCP Interface?

This project provides a Model Context Protocol (MCP) server with integrated Stable Diffusion image generation capabilities, enabling AI agents to request and receive generated images. It is built for developers who want to add AI-powered image generation to agent workflows such as those using Goose or FastMCP Inspector.

How to use AI Image Generation Server with MCP Interface?

Set up a Python virtual environment (named .venv), activate it, and install the package with pip install -e .. Then set the IMAGE_GEN_DIR environment variable to your target folder and run image-gen-mcp. For testing, use development mode with the FastMCP Inspector (web UI at http://127.0.0.1:6274). Integrate with Goose by adding a StandardIO extension with the command uv run /full/path/to/your/project/.venv/bin/image-gen-mcp and the same environment variable.

Key features of AI Image Generation Server with MCP Interface

• Integrated Stable Diffusion for image generation
• Model Context Protocol (MCP) interface for AI agents
• Returns image URL, dimensions, MIME type, and metadata
• Customizable port for the image generation service (default 5000)
• Works with Goose extension system
• Supports local inference via Ollama (tested with mistral-small3.1) or remote models like GPT-4o

Use cases of AI Image Generation Server with MCP Interface

• AI agents autonomously generating images from text prompts
• Integrating visual outputs into multi‑step agent workflows (e.g., Goose)
• Debugging and testing MCP tools with the FastMCP Inspector web interface
• Direct API access for standalone image generation via curl or scripts

FAQ from AI Image Generation Server with MCP Interface

What models does it support?

It uses Stable Diffusion for generation. When integrated with Goose, it works immediately with OpenAI GPT-4o and locally with Ollama using the mistral-small3.1 model (slower than GPT-4o).

How do I stop the services?

Press Ctrl+C in the terminal where the server is running. Both the image generation service and MCP server stop together.

Where are generated images stored?

Images are saved to the folder specified by the IMAGE_GEN_DIR environment variable. A direct HTTP endpoint serves them from that folder.

Can I change the default port?

Yes. Use the --port flag when starting the service, for example: image-gen-mcp --port 5001.

What are the runtime dependencies?

Python, a virtual environment, the MCP package (installed via pip install -e .), and the Stable Diffusion backend bundled in the project. For full agent integration, Goose or another MCP client is needed.