Fooocus - AI Image Tools Tool

Overview

Fooocus is an open-source, offline image-generation UI built on Stable Diffusion XL with a Gradio front-end. It emphasizes minimal manual tuning: the project bundles prompt-processing, sampling improvements and opinionated presets so short, human prompts produce high-quality images without heavy parameter tweaking. According to the repository, Fooocus includes an offline GPT‑2–based prompt-expansion engine, built-in inpaint/outpaint models, upscaling/refiner presets and multiple launch presets (anime/realistic), and attempts to keep the first-run experience to a few clicks by auto-downloading required models. ([github.com](https://github.com/lllyasviel/Fooocus)) Fooocus targets lower-VRAM setups compared with many SDXL workflows — the README documents a minimal NVIDIA GPU requirement of about 4 GB VRAM (though performance and speed improve with larger GPUs) and offers AMD (ROCm) and Apple Silicon (PyTorch MPS) guidance. The project is distributed under GPL‑3.0 and is delivered as a Windows bundle (run.bat), a pip-installable Python setup for Linux, and additional Docker instructions. The README also notes that Fooocus is now maintained in a limited long-term-support (LTS) mode with primarily bug fixes; active forks and community forks provide alternative features and updates. ([github.com](https://github.com/lllyasviel/Fooocus))

GitHub Statistics

• Stars: 47,462
• Forks: 7,738
• Contributors: 58
• License: GPL-3.0
• Primary Language: Python
• Last Updated: 2025-09-02T20:28:40Z
• Latest Release: v2.5.5

The GitHub repository shows strong community interest and adoption: the project has roughly 47.5k stars and ~7.7k forks, with over 1,100 commits and a few hundred open issues and pull requests, indicating substantial past activity and community contribution. The repo lists dozens of contributors and uses an active Issues/Discussions area for support and announcements. ([github.com](https://github.com/lllyasviel/Fooocus)) Activity profile: although the project attracted broad attention and many forks, the maintainer marked Fooocus as moving to limited long-term support (LTS) focused on bug fixes, and recommends forks for experimental feature development. That means the core repo is relatively stable, but many users now follow forks (and community-maintained variants) for new features and model-format support. Community discussion also highlights third-party paid websites falsely claiming to be official Fooocus — the project and maintainers repeatedly warn users to use the official GitHub source. ([github.com](https://github.com/lllyasviel/Fooocus))

Installation

Install via pip:

git clone https://github.com/lllyasviel/Fooocus.git

cd Fooocus

pip3 install -r requirements_versions.txt

python3 entry_with_update.py

python3 entry_with_update.py --listen

Key Features

• Prompt auto-expansion using an offline GPT-2–style engine for richer prompts without manual editing.
• Low-VRAM support—project documents minimal NVIDIA VRAM around 4 GB for basic use.
• Built-in inpaint/outpaint workflow with a dedicated inpaint control model (first-run auto-download ~1.28 GB).
• Inline LoRAs, wildcards and array processing for programmatic prompt variants and batch outputs.
• Presets and launchers (run.bat, run_anime.bat, run_realistic.bat) for one-click style selection and model auto-download.
• Image prompt tools: upscaling, variation, image-to-image, describe/inspect, and face-swap (InsightFace) helpers.
• Windows bundle plus pip/Docker options and platform notes for AMD (ROCm) and Apple Silicon (MPS).

Community

The project has a large user base and many forks (47.5k stars, ~7.7k forks) with active discussions and hundreds of issues and PRs on GitHub. The maintainer describes the repo as in LTS (bug‑fix) mode, so new experimental features are often developed in forks. Community feedback (Reddit/GitHub discussions) praises Fooocus's usability and SDXL quality but calls out slow performance on low‑end hardware, occasional accuracy issues, and warnings about third‑party websites selling access — the maintainers consistently advise using the official GitHub release or community Colab instances. ([github.com](https://github.com/lllyasviel/Fooocus))

Key Information

• Category: Image Tools
• Type: AI Image Tools Tool