Fooocus

A free, open-source local AI image generator that uses Stable Diffusion XL for offline image creation, editing and upscaling.

Alternative to:

Fooocus generates images on your computer for people who want to work from prompts without extensive settings adjustments. It uses Stable Diffusion XL and works offline after the required models are downloaded.

Prompt-based generation: Built-in generation defaults and an offline GPT-2 prompt processor reduce the need for manual tuning.
Image references and variations: Use an existing image to guide generation, produce subtle or stronger variations, or upscale a result.
Image editing: Inpainting replaces selected parts of an image, while outpainting extends it beyond its original borders.
Styles and models: Choose photography or anime presets, use SDXL models from Civitai, and adjust styles, negative prompts and prompt weights.
Image descriptions and faces: Convert an image into a text description or use InsightFace for face swapping.

Fooocus is free and open source under GPL-3.0, with a browser interface. On Windows, the stated minimum is an NVIDIA GPU with 4GB of VRAM and 8GB of system RAM, relying on Windows virtual memory.

The project is in limited long-term support, with maintenance restricted to bug fixes. Its model architecture remains SDXL; it doesn't support Flux.