Together AI Image Server

An MCP server for generating images using Together AI’s image generation models.

Setup

1. Install dependencies:

npm install

2. Build the server:

npm run build

Configuration

1. Together AI API Key

The server requires a Together AI API key. You can get one from Together AI’s platform.

2. Cline Configuration

Add the server to your Cline MCP settings file:

For macOS/Linux: ~/Library/Application Support/Windsurf/User/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json

{
  "mcpServers": {
    "togetherai-image": {
      "command": "node",
      "args": ["/path/to/togetherai-image-server/build/index.js"],
      "env": {
        "TOGETHER_API_KEY": "your-api-key-here"
      },
      "disabled": false,
      "autoApprove": []
    }
  }
}

Usage in Cline

The server provides a generate_image tool with the following parameters:

{
  prompt: string;           // Required: Text description of the image to generate
  model?: string;          // Optional: Model to use (default: 'black-forest-labs/FLUX.1.1-pro')
  width?: number;          // Optional: Image width in pixels (default: 1024)
  height?: number;         // Optional: Image height in pixels (default: 768)
  steps?: number;         // Optional: Number of inference steps (default: 28)
  n?: number;             // Optional: Number of images to generate (default: 1)
  outputDir?: string;     // Optional: Directory where images will be saved (default: './output')
}

Example Usage

You can use the server in Cline like this:

Generate an image of a cat and save it to my desktop:

togeherai-image
generate_image

{
  "prompt": "A cute cat sitting on a windowsill",
  "outputDir": "/Users/username/Desktop/generated-images"
}

Output

The tool returns a JSON response containing:

• filepath: Full path to the saved image
• filename: Name of the saved image file
• dimensions: Original and final dimensions of the image

Features

• Supports Together AI’s latest FLUX.1.1-pro model
• Automatically resizes images if dimensions are below 256 pixels
• Maintains aspect ratio during resizing
• High-quality JPEG output (90% quality)
• Creates output directories recursively
• Returns detailed metadata including image dimensions