FooocusPlus – AI Art Made Simple

FooocusPlus is a community-developed AI image generation program that makes creating stunning works of art easier than ever. FooocusPlus runs offline on your computer and is completely free! Now you can have your very own Stable Diffusion or Midjourney that helps turn your creative ideas into beautiful images without any technical skills.

Features

• Generate stunning images with meticulously prepared presets – just describe your new masterpiece and watch it come to life
• Get inspiration for your art – choose from an extensive list of styles, or use our integrated AI assistant to help you write the perfect prompt
• Enhance existing images using AI – banish Photoshop editing with advanced image modification tools, or even use an existing picture as inspiration for a new creation!
• Bring the power of community home – use LoRA models from websites like Civitai to add new concepts to your images
• all required models are auto-downloaded but only if you actually use them, preventing local drive clutter
• to provide worldwide access, Civitai is not used as a download source
• Full control at your fingertips – simple one-click access to advanced generation options, letting you use custom settings like a pro without ever leaving our friendly user interface
• Simple image adjustments and background transparency functions using the built-in Image Editor
• Multilingual – full support for English, French, Portuguese and Spanish out of the box, partial support for Simplified Chinese and limited support for over 40 other languages
• Private – runs on your computer and does not send information about your identity or AI art generation to third parties without your consent

Supported Models

FooocusPlus gives you access to a diversity of AI image generation base models, including:

• Stable Diffusion XL (SDXL) and a large number of SDXL variants
• Pony Diffusion XL
• Flux from Black Forest Labs
• Hunyuan-DiT
• Kwai Kolors
• Playground 2.5
• Segmind-Vega (coming soon)
• Stable Diffusion 1.5 (SD1.5)
• Stable Diffusion 3.5 Large & Medium
• Z-Image Base & Turbo

If you’re new to AI art, don’t worry! Our built-in presets will give you one-click access to all of those exciting AI image generation models without any prior knowledge required. Favorite presets are located in a bar at the top of the FooocusPlus main canvas, and the whole preset collection is available from two dropdown menus located under the Settings tab. You'll be able to choose your own favorite model in no time. Favorites can be added or removed using the Favorite Preset Control accordion located right below the Extras tab.

System Requirements

To run FooocusPlus, you will need a computer with a graphics card capable of running advanced AI software, including at least:

• 6GB of video RAM (8GB recommended for SDXL base models, 12GB for Flux)
• 16GB of system RAM (32GB recommended for SDXL, 48GB for Flux)
• 50GB of hard drive space

While some graphics cards with 4GB of video RAM (VRAM) will work - some of them even supporting Flux - image generation may be slow or erratic. Some 4GB VRAM cards may not support image generation at all.

For those systems that are capable of generating FooocusPlus images with only 4GB of VRAM, we include a unique default base model that is small and fast. In addition, FooocusPlus supports Stable Diffusion 1.5 (SD1.5), which should run in most 4GB VRAM cards.

FooocusPlus supports NVIDIA graphics cards on Windows and Linux, macOS on Silicon, and many AMD graphics cards on Linux. We also provide limited support for macOS on Intel. AMD support under Windows is coming soon.

Torchruntime is integrated into FooocusPlus. Please check their compatibility tables for more details on hardware and software support. But unlike Torchruntime, FooocusPlus does support NVIDIA Blackwell 50xx graphics cards in full native mode with fine tuning parameters and local PyTorch attention.

The Comfy generative system is integrated into FooocusPlus to provide support for non-SDXL models. Comfy no longer supports legacy video cards that require PyTorch 2.4 or earlier software. Due to this incompatibility, Comfy mode is disabled for this legacy hardware. However legacy video cards with at least 11 GB of video RAM (VRAM) do support Comfy using Forced Compatibility Mode.

Installing FooocusPlus

For Linux users, please use the Installation Script for Linux in the FooocusPlus Wiki.

And for Windows users, please use this Installation Procedure, also available in the Wiki.

Language Support

Language and other optional startup files are available in the FooocusPlus\UserDir\batch_startups folder. Please copy whatever you need to the FooocusPlus folder before using them. The following language startups files offer full support - in the user interface, the Prompt Translator, dynamic status reports and the console window:

• Chinese, Simplified: 常规启动.bat and 不更新启动.bat (no update)
• English, UK: run_FooocusPlus_UK.bat
• French: run_FooocusPlus_française.bat
• Portuguese: run_FooocusPlus_Português.bat
• Spanish: run_FooocusPlus_español.bat

A total of 40 languages are supported by FooocusPlus, but the majority only provide Prompt Translator, dynamic status reports, Wildcard and console window translation in the selected language. There are several limited support language files in the batch_startups folder, and details on how to setup other languages in the Wiki. Please start with Language An Introduction.

Getting Help

Have a question or concern? Visit the Discussions page here at GitHub or join our friendly community at the Facebook Pure Fooocus group. If you think you have found a bug or want to suggest a missing feature, visit the GitHub Issues page to file a report so that we can investigate.

Licensing

FooocusPlus is free to download and use, and the FooocusPlus developers will never request money or identifiable information from you in order to access or run the software. We do not accept donations or sponsorships, but instead welcome positive feedback, encouragement, and GitHub stars.

FooocusPlus is published under the GPL-3.0 license, and your use of the software is conditional on your acceptance of both the terms of the GPL-3.0 and the additional terms contained in this section. In particular, you acknowledge that some AI image generation models included in FooocusPlus are licensed for non-commercial community use only, and you agree to purchase all necessary licenses from the relevant developers prior to commercially distributing artwork generated with these models.

You also agree that you will not use FooocusPlus to:

• Violate the laws of your country, state, or locality
• Exploit minors or create child exploitation content
• Create non-consensual or illegal pornographic content
• Create material that may be used to harm or harass others
• Violate the license agreements of included software packages or AI models

Credits

FooocusPlus is Copyright © 2024-2026 David Sage and contributors.

This software was made possible by integrating open-source technologies from projects including:

• Fooocus by illyasviel and mashb1t
• ComfyUI by comfyanonymous and the Comfy-Org team
• SimpleSDXL2 by metercai
• diffusers by the huggingface team
• torchruntime by cmdr2 and the easydiffusion team
• OneButtonPrompt by AIrjen, via SimpleSDXL2 and RuinedFooocus
• Superprompter by sammcj
• Z-Image Turbo DIT and Stable Diffusion 1.5 (SD1.5) support by Irmagon (Thomas Gaud)
• the "All 276 Styles in Fooocus (Google table)", also by Irmagon (Thomas Gaud)
• the members of the Pure Fooocus Facebook group for their encouragement, enthusiasm, Beta testing and the Founders Project welcome images