Generate Image

You can run a dedicated MCP server that forwards OpenAI-compatible image generation API calls, enabling you to generate images across multiple models and providers through a single, compatible interface. This server is built to be easy to deploy locally or in your environment and can be integrated with MCP clients to request image generation from OpenAI-compatible endpoints.

How to use

You connect to the server from an MCP client by starting the server process described in the configuration snippets. Once running, you can issue image generation requests through the MCP-compatible interface, and the server will forward those requests to the appropriate underlying image models (such as OpenAI’s DALL·E, Claude, MidJourney-like APIs, and other supported sources). Ensure your client is configured to reach the server’s endpoint (local stdio or a remote HTTP URL) as described in the setup steps.

How to install

Prerequisites you need before starting: Node.js 16+ and either npm or yarn, plus a supported operating system (Windows or macOS). Make sure you have network access to install dependencies.

Install dependencies and prepare the project with these commands:

npm install

Additional configuration and runtime details

Configure the MCP server for your environment by specifying how you want to start the process and which API keys to use. The following examples show two explicit startup configurations you can place in your MCP client configuration.

{
  "mcpServers": {
    "genimg-mcp": {
      "command": "node",
      "arg": ["C:\\path\\to\\genimg-mcp\\dist\\index.js"],
      "env": {
        "OPENAI_API_KEY": "sk-xxxxx",
        "OPENAI_MODEL_ID": "dall-e-3"  
      }
    }
  }
}

Troubleshooting and debugging

Debugging MCP servers can be challenging because communication happens over standard input and output. You can use the built-in inspector tool to aid debugging. Run the inspector to start a debugging session and obtain a browser-based interface for troubleshooting.

npm run inspector

Security and keys

Protect your API keys and only expose the server to trusted clients. Store keys in environment variables as shown in the configuration examples, and avoid hard-coding sensitive values in code or configuration files.

Contact and licensing

If you need support, reach out to the provided contact channel for guidance. This project uses the MIT license; review the LICENSE file for full terms.

Configuration notes

The MCP server supports multiple startup configurations. The Windows-style startup uses a node-based start with the path to the compiled index, while the macOS-style startup uses the full path to the compiled index as the command. Each configuration can define its own environment variables as needed.

Available tools

inspector

MCP Inspector tool for debugging and inspecting MCP server communication via a browser-based interface.