Image Generation MCP Server
@sarthakkimtani
MCP server enabling Image Generation for LLMs, built in Python and integrated with Together AI.
Overview
What is Image Generation MCP Server?
A Model Context Protocol (MCP) server that enables seamless generation of high-quality images via Together AI. It provides a standardized interface to specify image generation parameters for MCP-compatible clients.
How to use Image Generation MCP Server?
Install using uv and configure in your MCP client (e.g., Claude Desktop) by adding a server entry with the command uv --directory /ABSOLUTE/PATH/TO/image-gen/ run image-gen and the required TOGETHER_AI_API_KEY environment variable. Once connected, call the generate_image tool with a descriptive prompt and optional width, height, and model parameters.
Key features of Image Generation MCP Server
- High-quality image generation via the Flux.1 Schnell model
- Customizable output dimensions (width and height)
- Clear error handling for prompt validation and API issues
- Easy integration with MCP-compatible clients
- Support for specifying an exact Together AI model with automatic fallback
Use cases of Image Generation MCP Server
- Generate images from descriptive text prompts for content creation
- Specify custom dimensions to match project layout requirements
- Integrate image generation into existing MCP-based workflows and tools
FAQ from Image Generation MCP Server
What model does the server use by default?
The default model is black-forest-labs/FLUX.1-schnell. You can override it with any Together AI model name via the model parameter.
What are the prerequisites for running the server?
You need Python 3.12 or higher, the httpx and mcp packages, and a valid Together AI API key.
How do I configure the API key?
Set the TOGETHER_AI_API_KEY environment variable in your MCP client configuration (e.g., claude_desktop_config.json).
What tools does the server provide?
It provides one tool: generate_image, which accepts a prompt (required), and optional width, height, and model parameters.
Can I specify the image dimensions?
Yes, you can optionally set width and height in pixels when calling generate_image.
