Leonardo AI MCP Server

An MCP (Model Context Protocol) server that provides access to Leonardo AI's image generation API with 150 free daily credits.

Features

• Generate high-quality images using various AI models
• 150 free credits daily (refreshes at midnight)
• Support for multiple art styles and models
• Batch generation (1-4 images at once)
• Save images locally or get URLs
• Check remaining credits

Available Models

• Leonardo Diffusion XL: High-quality general purpose model
• Leonardo Vision XL: Photorealistic and artistic styles
• Leonardo Anime XL: Anime and manga style images
• DreamShaper v7: Fantasy and dreamlike images
• Realistic Vision v5: Highly realistic photographs

Installation

1. Clone this repository:

git clone https://github.com/nikolausm/leonardo-ai-mcp-server.git
cd leonardo-ai-mcp-server

2. Install dependencies:

npm install

3. Get your Leonardo AI API key:

   • Sign up at https://leonardo.ai (free)
   • Go to API settings
   • Generate an API key
   • Copy the key

4. Set up environment variable:

export LEONARDO_API_KEY="your_api_key_here"

Configuration in Claude Desktop

Add to your Claude Desktop configuration (~/Library/Application Support/Claude/claude_desktop_config.json on macOS):

{
  "mcpServers": {
    "leonardo-ai": {
      "command": "node",
      "args": ["/path/to/leonardo-ai-mcp-server/index.js"],
      "env": {
        "LEONARDO_API_KEY": "your_api_key_here"
      }
    }
  }
}

Usage

Generate an Image

// Basic usage
generate_image({
  prompt: "a majestic castle on a hilltop at sunset"
})

// Advanced usage
generate_image({
  prompt: "cute cartoon guinea pig with big eyes",
  model: "leonardo-anime-xl",
  negative_prompt: "realistic, scary, dark",
  num_images: 2,
  width: 768,
  height: 768,
  guidance_scale: 10,
  save_path: "/tmp/guinea_pig.png"
})

Check Remaining Credits

check_credits()

List Available Models

list_models()

Free Credits System

• 150 credits daily (resets at midnight UTC)
• Each image costs approximately 2 credits
• That's about 75 images per day for free!
• No credit card required

Tips for Better Results

1. Use specific models:

   • leonardo-anime-xl for anime/cartoon style
   • realistic-vision-v5 for photorealistic images
   • dreamshaper-v7 for fantasy art

2. Optimize dimensions:

   • Use multiples of 8 (512, 768, 1024)
   • Square images often work best

3. Negative prompts are powerful:

   • Use them to avoid unwanted elements
   • Example: "blurry, low quality, distorted"

Examples

Children's Book Illustration

generate_image({
  prompt: "friendly guinea pig reading a book, children's illustration, soft pastel colors",
  model: "leonardo-diffusion-xl",
  negative_prompt: "realistic, dark, scary",
  width: 768,
  height: 768
})

Educational Poster

generate_image({
  prompt: "educational poster showing guinea pig care, clean design, infographic style",
  model: "leonardo-vision-xl",
  num_images: 3,
  guidance_scale: 12
})

Error Handling

The server handles common errors:

• Invalid API key
• Insufficient credits
• Generation failures
• Timeout issues

License

MIT

Author

Michael Nikolaus

Links

• Leonardo AI
• API Documentation