Fooocus
Stable Diffusion image generator with minimal setup
48.1k|stable|v2.5.5
Install via command line
Download and extract Fooocus_win64_2-5-0.7z, then run run.bat
docker-compose up
git clone https://github.com/lllyasviel/Fooocus && cd Fooocus && python launch.py
Prompt-focused generation — Generates high-quality images without manual parameter tuning using offline GPT-2 prompt processing.
Inpainting and outpainting — Custom inpaint algorithm for image editing with directional pan and expansion.
Image prompt and face swap — Uses custom image prompt algorithm and InsightFace integration for variation and transformation.
Multi-model presets — Switch between standard, anime, and realistic model presets directly in the browser.
Low VRAM support — Minimum 4GB GPU memory requirement with optimized sampling for efficient inference.
Fooocus strips away complexity from Stable Diffusion by automating prompt engineering and sampling improvements, letting you focus on describing what you want. Installation takes under three clicks—download, extract, run—with automatic model management. It's built on SDXL and designed for offline use on consumer laptops with modest GPUs.
Minimum 4GB VRAM (Nvidia), 16GB system RAM recommended. Python 3.8+ for source installation.
