jankutschera/flux2-mcp-server
If you are the rightful owner of flux2-mcp-server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.
The Flux2 MCP Server is a Model Context Protocol server that allows Claude and other MCP clients to generate images using the advanced FLUX.2 models from Black Forest Labs.
Flux2 MCP Server
An MCP (Model Context Protocol) server that enables Claude and other MCP clients to generate images using fal.ai's FLUX.2 models - the latest generation of image generation from Black Forest Labs.
Features
- FLUX.2 [pro] - Studio-grade, production-optimized images
- FLUX.2 [dev] - Development version with full parameter control
- FLUX.2 [flex] - Flexible generation with prompt expansion
- FLUX.2 [schnell] - Ultra-fast generation (4 steps)
- Image Editing - Edit existing images with text instructions
Installation
Quick Install (Recommended)
# Add to Claude Code
claude mcp add flux2 --env "FAL_KEY=your-fal-api-key" -- npx -y flux2-mcp-server
Manual Installation
npm install -g flux2-mcp-server
Or run directly with npx:
FAL_KEY=your-api-key npx flux2-mcp-server
Setup
1. Get your fal.ai API Key
2. Configure Claude Desktop
Add to your Claude Desktop config:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"flux2": {
"command": "npx",
"args": ["-y", "flux2-mcp-server"],
"env": {
"FAL_KEY": "your-fal-api-key"
}
}
}
}
3. Restart Claude Desktop
Available Tools
generate_image_flux2_pro
Generate studio-grade images with FLUX.2 [pro].
"Generate a photorealistic portrait of a woman in a futuristic city"
Parameters:
prompt(required): Image descriptionimage_size: square_hd, square, portrait_4_3, portrait_16_9, landscape_4_3, landscape_16_9seed: For reproducible resultsoutput_format: jpeg or pngsafety_tolerance: 1-5 (1=strict, 5=permissive)
generate_image_flux2_dev
Development model with full control over inference parameters.
Additional Parameters:
num_inference_steps: 1-50 (default: 28)guidance_scale: 1-20 (default: 3.5)
generate_image_flux2_flex
Flexible model with automatic prompt expansion.
Additional Parameters:
enable_prompt_expansion: Auto-expand prompts (default: true)num_inference_steps: 2-50 (default: 28)guidance_scale: 1.5-10 (default: 3.5)
generate_image_flux2_schnell
Ultra-fast generation for rapid prototyping.
Parameters:
num_inference_steps: 1-12 (default: 4)
edit_image_flux2
Edit existing images with text instructions.
"Add sunglasses to the person in this image"
Parameters:
prompt(required): Editing instructionsimage_url(required): URL of image to edit
list_flux2_models
List all available FLUX.2 models with pricing info.
Usage Examples
In Claude, you can say:
- "Generate a cyberpunk cityscape at sunset using Flux 2 Pro"
- "Create a quick sketch of a cat using the schnell model"
- "Edit this image and add a rainbow in the background"
- "What Flux 2 models are available?"
Pricing (fal.ai)
| Model | Price |
|---|---|
| FLUX.2 [pro] | $0.03/MP (first), $0.015/MP (additional) |
| FLUX.2 [dev] | ~$0.025/image |
| FLUX.2 [flex] | $0.06/MP |
| FLUX.2 [schnell] | ~$0.003/image |
MP = Megapixel
Development
# Clone and install
git clone https://github.com/yourusername/flux2-mcp-server.git
cd flux2-mcp-server
npm install
# Build
npm run build
# Run in development
FAL_KEY=your-key npm run dev
Environment Variables
| Variable | Required | Description |
|---|---|---|
FAL_KEY | Yes | Your fal.ai API key |
License
MIT
Links
- fal.ai - AI model platform
- FLUX.2 Announcement - Black Forest Labs
- MCP Documentation - Model Context Protocol