Fooocus (SayMaven Fork - SDXL & Anima DiT Dual-Architecture)

Fooocus is an image generating software (based on Gradio ).

Fooocus presents a rethinking of image generator designs. The software is offline, open source, and free, while at the same time, similar to many online image generators like Midjourney, manual tweaking is not needed, and users only need to focus on prompts and images. Fooocus has also simplified installation: between pressing "download" and generating the first image, the number of needed mouse clicks is strictly limited to less than 3. Minimal GPU memory requirement is 4GB (Nvidia).

🚀 SayMaven Fork Highlights: SDXL & Anima DiT Support

This repository is an active fork extending Fooocus with Dual-Architecture Support and full compatibility with modern Google Colab runtimes (Python 3.13, CUDA 12.8+, PyTorch 2.x, Transformers 5.x):

Dual-Architecture Backbone:
- SDXL Pipeline (UNet): Preserves complete, flawless SDXL compatibility (CLIP-L + OpenCLIP BigG, 4-channel latent, standard SDXL VAE, refiners, and Fooocus V2 expansion).
- Anima Pipeline (DiT - Diffusion Transformer): Full integration of the 28-block DiT architecture (MiniTrainDIT / CosmosTransformer), Qwen3-0.6B text conditioning with Danbooru tag padding (512 token attention sink), and Wan21 16-channel 3D VAE (WanVAE).
ComfyUI Reference Sampler Bootstrap:
- Direct execution via ComfyUI reference sampler (euler_ancestral + simple, shift 3.0, multiplier 1.0) with custom LoRA support and in-memory caching (_anima_lora_cache).
Memory & VRAM Optimizations for Colab (Tesla T4 16GB):
- Weight Unification: Merges identical DiT model weights on GPU, returning ~4.5 GB VRAM instantly.
- Sequential CFG Batching (batch=1): Evaluates positive and negative conditionings sequentially, slashing RoPE activation peaks by ~50% and enabling High-Res Upscale (1.5x / 2.0x) without memory freeze or CUDA OOM.
- Zero CPU Offloading: Eliminates DiT RAM offloading, completely avoiding Linux OOM Killer terminations.
Zero-VRAM Live Step Preview for Anima:
- Instant live preview during DiT sampling steps using CPU-based linear Wan21 RGB latent projection.
Custom Upscaler Dropdown & Robust Model Loader:
- WebUI dropdown selector in the Upscale or Variation and Enhance tabs supporting any custom super-resolution model (RealESRGAN_x4Plus_Anime_6B, 4x-UltraSharp, Remacri, etc.) placed in models/upscale_models/.
- Smart download prevention: Eliminates redundant Hugging Face downloads when custom models are chosen.
- Magic-byte verification in model loader to automatically detect PyTorch .pt/.pth archives even if mistakenly named .safetensors.
Anima ControlNet-LLLite Integration:
- Seamless structural, lineart, and pose conditioning for DiT models via ControlNet-LLLite (Low-Rank Light Adaptation ControlNet) directly in the Image Prompt tab.
- Lightweight model footprint (~52 MB) with automatic lazy downloading (any-test-like-v2.safetensors from kohyass/anima-pencil-xl-controlnet-lllite).
- Calibrated default parameters (strength: 0.2, stop_at: 0.4) tuned to preserve crisp outlines and structural control without over-saturation across both base diffusion and high-res upscale passes.
- Dynamic WebUI switching: automatically restricts Image Prompt methods to ["Anima-LLLite"] for Anima models while preserving the full suite (ImagePrompt, FaceSwap, PyraCanny, CPDS) for SDXL models.
YOLO Anime Detection & ADetailer-Style Multi-Tab Enhancement:
- Native Anime Detection Models: Integrated ONNX models specifically trained for anime/illustration characters (DeepGHS & Hysts):
  - yolov8n-animeface.onnx (Default mask generation model for both Inpaint & Enhance tabs)
  - yolov8s-animeface.onnx
  - face_yolov8n.onnx
  - yolov8n-eyes.onnx (Anime eye detection)
  - hand_yolov8n.onnx (Anime hand & finger detection)
  - person_yolov8n.onnx (Anime full-body character detection)
- Dynamic Custom Model Auto-Discovery: Drop any custom YOLO .onnx or .pt model into models/detection/ to automatically register it in the WebUI.
- Seam-Free Elliptical Feathering: Replaced harsh rectangular bounding-box cuts with smooth elliptical mask contours (cv2.ellipse), Gaussian feathering (k=31), and a 2D boundary window taper inside InpaintWorker, completely eliminating hard boundary cut lines.
- Calibrated DiT Inpaint Defaults:
  - Default Inpaint Method: Improve Detail (face, hand, eyes, etc.)
  - Default Denoise Strength: 0.3 (was 0.5; preserves 70% of character anatomy/pose while refining fine facial/hand micro-details)
  - Default Respective Field: 0.0 (tight crop for maximum localized detail resolution)
  - Inpaint Engine: None (native DiT crop-and-stitch bypasses incompatible SDXL inpaint heads)
- Sequential Multi-Tab Enhancement Pipeline: Chain detection models across tabs (e.g. #1 Anime Face $\rightarrow$ #2 Hands $\rightarrow$ #3 Eyes) with zero-VRAM ONNX inference and smart auto-skip when target objects are not present.
- Persistent Prompt LoRA Retention: LoRAs defined via prompt syntax (<lora:name:weight>) are preserved across all sequential enhancement passes, preventing character LoRAs from unloading when custom enhancement prompts (e.g. beautiful eyes) are used.
Enhanced 3-Stage Gallery Zoom Viewer:
- Intuitive 3-stage inspection flow: Grid Thumbnails $\rightarrow$ Canvas Focused View $\rightarrow$ Fullscreen Lightbox Modal.
- Centered fullscreen lightbox with keyboard arrow navigation (ArrowLeft / ArrowRight) and quick escape (Esc), eliminating UI cut-offs when previewing high-resolution generations.
Modernized Gradio 6, NumPy 2 & CuPy v14 Stack:
- Fully upgraded UI and tensor ecosystem to Gradio 6.26.0, NumPy 2.5.3, and CuPy 14.0.1 on Python 3.13.
- Zero-drift locked dependencies in requirements_versions.txt for 100% reproducible Colab startups.
- Native ImageEditor adapter supporting dictionary payload structures (background + layers with alpha mask extraction) alongside legacy sketch fallbacks.
- Resilient custom route handlers (/file= and /history_log) directly bound to FastAPI, ensuring seamless access to output images and private history logs.
- Responsive WebUI enhancements: symmetrical 2x2 Image Prompt grid, inline horizontal control buttons (Input Image, Enhance, Advanced), and a 2-column scrollable styles selector.
Fast Startup & Colab Optimization:
- Native NumPy 2.x support completely eliminates legacy 7-minute Meson/Ninja wheel compilation on Python 3.13.
- CLI flag --skip-pip (or SKIP_PIP=1) to bypass requirements verification on warm runtimes.
- Automated Civitai token authentication (CIVITAI_API_TOKEN / CIVITAI_TOKEN).

Features

Below is a quick list using Midjourney's examples:

Midjourney	Fooocus
High-quality text-to-image without needing much prompt engineering or parameter tuning. (Unknown method)	High-quality text-to-image without needing much prompt engineering or parameter tuning. (Fooocus has an offline GPT-2 based prompt processing engine and lots of sampling improvements so that results are always beautiful, no matter if your prompt is as short as “house in garden” or as long as 1000 words)
V1 V2 V3 V4	Input Image -> Upscale or Variation -> Vary (Subtle) / Vary (Strong)
U1 U2 U3 U4	Input Image -> Upscale or Variation -> Upscale (1.5x) / Upscale (2x)
Inpaint / Up / Down / Left / Right (Pan)	Input Image -> Inpaint or Outpaint -> Inpaint / Up / Down / Left / Right (Fooocus uses its own inpaint algorithm and inpaint models so that results are more satisfying than all other software that uses standard SDXL inpaint method/model)
Image Prompt	Input Image -> Image Prompt (Fooocus uses its own image prompt algorithm so that result quality and prompt understanding are more satisfying than all other software that uses standard SDXL methods like standard IP-Adapters or Revisions)
--style	Advanced -> Style
--stylize	Advanced -> Advanced -> Guidance
--niji	Multiple launchers: "run.bat", "run_anime.bat", and "run_realistic.bat". Fooocus support SDXL models on Civitai (You can google search “Civitai” if you do not know about it)
--quality	Advanced -> Quality
--repeat	Advanced -> Image Number
Multi Prompts (::)	Just use multiple lines of prompts
Prompt Weights	You can use " I am (happy:1.5)". Fooocus uses A1111's reweighting algorithm so that results are better than ComfyUI if users directly copy prompts from Civitai. (Because if prompts are written in ComfyUI's reweighting, users are less likely to copy prompt texts as they prefer dragging files) To use embedding, you can use "(embedding:file_name:1.1)"
--no	Advanced -> Negative Prompt
--ar	Advanced -> Aspect Ratios
InsightFace	Input Image -> Image Prompt -> Advanced -> FaceSwap
Describe	Input Image -> Describe

Below is a quick list using LeonardoAI's examples:

LeonardoAI	Fooocus
Prompt Magic	Advanced -> Style -> Fooocus V2
Advanced Sampler Parameters (like Contrast/Sharpness/etc)	Advanced -> Advanced -> Sampling Sharpness / etc
User-friendly ControlNets	Input Image -> Image Prompt -> Advanced

Also, click here to browse the advanced features.

Download

Windows

You can directly download Fooocus with:

>>> Click here to download <<<

After you download the file, please uncompress it and then run the "run.bat".

The first time you launch the software, it will automatically download models:

It will download default models to the folder "Fooocus\models\checkpoints" given different presets. You can download them in advance if you do not want automatic download.
Note that if you use inpaint, at the first time you inpaint an image, it will download Fooocus's own inpaint control model from here as the file "Fooocus\models\inpaint\inpaint_v26.fooocus.patch" (the size of this file is 1.28GB).

After Fooocus 2.1.60, you will also have run_anime.bat and run_realistic.bat. They are different model presets (and require different models, but they will be automatically downloaded). Check here for more details.

After Fooocus 2.3.0 you can also switch presets directly in the browser. Keep in mind to add these arguments if you want to change the default behavior:

Use --disable-preset-selection to disable preset selection in the browser.
Use --always-download-new-model to download missing models on preset switch. Default is fallback to previous_default_models defined in the corresponding preset, also see terminal output.

If you already have these files, you can copy them to the above locations to speed up installation.

Note that if you see "MetadataIncompleteBuffer" or "PytorchStreamReader", then your model files are corrupted. Please download models again.

Below is a test on a relatively low-end laptop with 16GB System RAM and 6GB VRAM (Nvidia 3060 laptop). The speed on this machine is about 1.35 seconds per iteration. Pretty impressive – nowadays laptops with 3060 are usually at very acceptable price.

Note that the minimal requirement is 4GB Nvidia GPU memory (4GB VRAM) and 8GB system memory (8GB RAM). This requires using Microsoft’s Virtual Swap technique, which is automatically enabled by your Windows installation in most cases, so you often do not need to do anything about it. However, if you are not sure, or if you manually turned it off (would anyone really do that?), or if you see any "RuntimeError: CPUAllocator", you can enable it here:

Click here to see the image instructions.

And make sure that you have at least 40GB free space on each drive if you still see "RuntimeError: CPUAllocator" !

Please open an issue if you use similar devices but still cannot achieve acceptable performances.

Note that the minimal requirement for different platforms is different.

See also the common problems and troubleshoots here.

Colab (Modern Runtime Support - SayMaven Fork)

This fork is actively maintained on branch anima-support (Dual-Architecture: SDXL + Anima DiT) and colab-support (Stable SDXL) for Google Colab's modern runtime (Python 3.13+, CUDA 12.8+, Transformers 5.x, PyTorch 2.x).

Quick Start in Google Colab (Anima DiT & SDXL Dual-Architecture):

%cd /content
!if [ ! -d "Fooocus" ]; then git clone -b anima-support https://github.com/SayMaven/Fooocus.git; fi
%cd /content/Fooocus
!python entry_with_update.py --share --always-high-vram --preset anima --skip-pip --port 7866

Quick Start for Standard SDXL:

%cd /content
!if [ ! -d "Fooocus" ]; then git clone -b anima-support https://github.com/SayMaven/Fooocus.git; fi
%cd /content/Fooocus
!python entry_with_update.py --share --always-high-vram --skip-pip --port 7866

Tips:

Using --always-high-vram shifts resource allocation from RAM to VRAM and achieves the overall best balance between performance, flexibility, and stability on the default Tesla T4 instance.
Using --skip-pip bypasses package dependency re-checks during subsequent launches, cutting startup time down to seconds.
You can switch presets anytime with --preset anima, --preset anime, --preset realistic, etc.

Linux (Using Anaconda)

If you want to use Anaconda/Miniconda, you can:

git clone -b anima-support https://github.com/SayMaven/Fooocus.git
cd Fooocus
conda env create -f environment.yaml
conda activate fooocus
pip install -r requirements_versions.txt

Then download the models into Fooocus/models/checkpoints, or let Fooocus automatically download the models using the launcher:

conda activate fooocus
python entry_with_update.py

Or, if you want to open a remote port, use:

conda activate fooocus
python entry_with_update.py --listen

Use python entry_with_update.py --preset anima for Fooocus Anima (DiT) Edition, or --preset anime / --preset realistic.

Linux (Using Python Venv)

Using Python venv:

git clone -b anima-support https://github.com/SayMaven/Fooocus.git
cd Fooocus
python3 -m venv fooocus_env
source fooocus_env/bin/activate
pip install -r requirements_versions.txt

Launch the software with:

source fooocus_env/bin/activate
python entry_with_update.py

Or, if you want to open a remote port, use:

source fooocus_env/bin/activate
python entry_with_update.py --listen

Use python entry_with_update.py --preset anima for Fooocus Anima (DiT) Edition, or --preset anime / --preset realistic.

Linux (Using native system Python)

If you have Python (and Pip) installed on your system:

git clone -b anima-support https://github.com/SayMaven/Fooocus.git
cd Fooocus
pip3 install -r requirements_versions.txt

Launch the software with:

python3 entry_with_update.py

Or, if you want to open a remote port, use:

python3 entry_with_update.py --listen

Use python entry_with_update.py --preset anima for Fooocus Anima (DiT) Edition, or --preset anime / --preset realistic.

Linux (AMD GPUs)

Note that the minimal requirement for different platforms is different.

Same with the above instructions. You need to change torch to the AMD version

pip uninstall torch torchvision torchaudio torchtext functorch xformers 
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/rocm5.6

AMD is not intensively tested, however. The AMD support is in beta.

Use python entry_with_update.py --preset anime or python entry_with_update.py --preset realistic for Fooocus Anime/Realistic Edition.

Windows (AMD GPUs)

Note that the minimal requirement for different platforms is different.

Same with Windows. Download the software and edit the content of run.bat as:

.\python_embeded\python.exe -m pip uninstall torch torchvision torchaudio torchtext functorch xformers -y
.\python_embeded\python.exe -m pip install torch-directml
.\python_embeded\python.exe -s Fooocus\entry_with_update.py --directml
pause

Then run the run.bat.

AMD is not intensively tested, however. The AMD support is in beta.

For AMD, use .\python_embeded\python.exe Fooocus\entry_with_update.py --directml --preset anime or .\python_embeded\python.exe Fooocus\entry_with_update.py --directml --preset realistic for Fooocus Anime/Realistic Edition.

Mac

Note that the minimal requirement for different platforms is different.

Mac is not intensively tested. Below is an unofficial guideline for using Mac. You can discuss problems here.

You can install Fooocus on Apple Mac silicon (M1 or M2) with macOS 'Catalina' or a newer version. Fooocus runs on Apple silicon computers via PyTorch MPS device acceleration. Mac Silicon computers don't come with a dedicated graphics card, resulting in significantly longer image processing times compared to computers with dedicated graphics cards.

Install the conda package manager and pytorch nightly. Read the Accelerated PyTorch training on Mac Apple Developer guide for instructions. Make sure pytorch recognizes your MPS device.
Open the macOS Terminal app and clone this repository with git clone -b anima-support https://github.com/SayMaven/Fooocus.git.
Change to the new Fooocus directory, cd Fooocus.
Create a new conda environment, conda env create -f environment.yaml.
Activate your new conda environment, conda activate fooocus.
Install the packages required by Fooocus, pip install -r requirements_versions.txt.
Launch Fooocus by running python entry_with_update.py. (Some Mac M2 users may need python entry_with_update.py --disable-offload-from-vram to speed up model loading/unloading.) The first time you run Fooocus, it will automatically download the Stable Diffusion SDXL models and will take a significant amount of time, depending on your internet connection.

Use python entry_with_update.py --preset anime or python entry_with_update.py --preset realistic for Fooocus Anime/Realistic Edition.

Docker

See docker.md

Download Previous Version

See the guidelines here.

Minimal Requirement

Below is the minimal requirement for running Fooocus locally. If your device capability is lower than this spec, you may not be able to use Fooocus locally. (Please let us know, in any case, if your device capability is lower but Fooocus still works.)

Operating System	GPU	Minimal GPU Memory	Minimal System Memory	System Swap	Note
Windows/Linux	Nvidia RTX 4XXX	4GB	8GB	Required	fastest
Windows/Linux	Nvidia RTX 3XXX	4GB	8GB	Required	usually faster than RTX 2XXX
Windows/Linux	Nvidia RTX 2XXX	4GB	8GB	Required	usually faster than GTX 1XXX
Windows/Linux	Nvidia GTX 1XXX	8GB (* 6GB uncertain)	8GB	Required	only marginally faster than CPU
Windows/Linux	Nvidia GTX 9XX	8GB	8GB	Required	faster or slower than CPU
Windows/Linux	Nvidia GTX < 9XX	Not supported	/	/	/
Windows	AMD GPU	8GB (updated 2023 Dec 30)	8GB	Required	via DirectML (* ROCm is on hold), about 3x slower than Nvidia RTX 3XXX
Linux	AMD GPU	8GB	8GB	Required	via ROCm, about 1.5x slower than Nvidia RTX 3XXX
Mac	M1/M2 MPS	Shared	Shared	Shared	about 9x slower than Nvidia RTX 3XXX
Windows/Linux/Mac	only use CPU	0GB	32GB	Required	about 17x slower than Nvidia RTX 3XXX

* AMD GPU ROCm (on hold): The AMD is still working on supporting ROCm on Windows.

* Nvidia GTX 1XXX 6GB uncertain: Some people report 6GB success on GTX 10XX, but some other people report failure cases.

Note that Fooocus is only for extremely high quality image generating. We will not support smaller models to reduce the requirement and sacrifice result quality.

Troubleshoot

See the common problems here.

Default Models

Given different goals, the default models and configs of Fooocus are different:

Task	Windows	Linux args	Main Model	Refiner	Config
General (SDXL)	run.bat		juggernautXL_v8Rundiffusion	not used	here
Realistic (SDXL)	run_realistic.bat	--preset realistic	realisticStockPhoto_v20	not used	here
Anime (SDXL)	run_anime.bat	--preset anime	animaPencilXL_v500	not used	here
Anima (DiT)	run_anima.bat	--preset anima	Ob_animaV4	not used	here
Anima Base (DiT)		--preset anima_base_v1	anima-base-v1.0	not used	here
Hassaku Anima (DiT)		--preset hassaku_anima_v01	hassakuAnima_v01	not used	here
Wai Anima (DiT)		--preset wai_anima	waiANIMA_v10Base10	not used	here

Note that the download is automatic - you do not need to do anything if the internet connection is okay. However, you can download them manually if you (or move them from somewhere else) have your own preparation.

🔑 Civitai API Token Authentication

Many checkpoints and LoRAs on Civitai (such as early-access, age-gated, or restricted models) require account authentication. Without a token, downloading them directly will trigger an HTTP 401 Unauthorized error.

Fooocus supports automated Civitai token authentication via the CIVITAI_API_TOKEN (or CIVITAI_TOKEN) environment variable or a local .env file. When configured, Fooocus automatically appends your token to any civitai.com or civitai.red download request.

How to configure:

1. Via `.env` File (Recommended for Local Windows / Linux)

Create a .env file in the Fooocus root directory (already included in .gitignore to protect your privacy):

CIVITAI_API_TOKEN=your_civitai_token_here

2. Via Google Colab Notebook

Set the environment variable in a notebook cell before launching Fooocus:

%env CIVITAI_API_TOKEN=your_civitai_token_here

(Or use Colab's built-in Secrets 🔑 feature in the left sidebar).

3. Via Windows Batch Script (`run.bat`)

Add this line before the python launch command:

set CIVITAI_API_TOKEN=your_civitai_token_here

4. Via Terminal (PowerShell / Bash)

PowerShell (Windows):

$env:CIVITAI_API_TOKEN = "your_civitai_token_here"

Bash (Linux/macOS):

export CIVITAI_API_TOKEN="your_civitai_token_here"

Where to get your Civitai API Token:

Log in to Civitai.com.

Go to Account Settings -> API Keys.

Click Add API Key, give it a name, and copy your token.

🔍 Custom Upscalers & Super-Resolution Models

Fooocus allows you to use custom super-resolution models (such as RealESRGAN_x4Plus_Anime_6B, 4x-UltraSharp, Remacri, DAT, etc.) alongside the default Fooocus upscaler.

How to Use:

Place your upscaler models into the models/upscale_models/ directory (supports .pth, .pt, .safetensors, and .bin).
Open the WebUI and navigate to Input Image -> Upscale or Variation (or the Enhance tab).
Select your desired model from the Upscaler dropdown:
- Default (Fooocus): Uses Fooocus's built-in upscaler model (fooocus_upscaler_s409985e5.bin).
- Custom Models: Any models detected in models/upscale_models/ appear directly in the list (e.g. RealESRGAN_x4Plus_Anime_6B.pt).
If you add new models while Fooocus is running, simply click 🔄 Refresh All Files in the Model tab to update the list without restarting.

Key Advantages:

Smart Download Prevention: When you select a custom upscaler, Fooocus will never download the default 32 MB model from Hugging Face. If Default (Fooocus) is chosen, it only downloads once if the file is missing locally.
Multi-Scale Compatibility: Even if a custom model has a fixed scale (e.g., 4x), Fooocus automatically upscales and precisely resamples the image to your requested scaling target (Upscale (Fast 2x), Upscale (1.5x), or Upscale (2x)), followed by the high-resolution diffusion refinement pass.
Magic Header Detection: If a PyTorch .pt or .pth model was saved with a .safetensors extension, Fooocus automatically inspects the first 8 magic bytes and safely routes it to torch.load to avoid SafetensorError: header too large crashes.

🎨 Anima ControlNet-LLLite (Structure & Pose Conditioning)

Unlike SDXL models that rely on heavy UNet ControlNets or IP-Adapters, the Anima DiT architecture uses ControlNet-LLLite (developed by Kohya-ss) for structural guidance:

Key Highlights:

Ultra-Lightweight: Models are only ~52 MB (compared to 1.4–2.5 GB for traditional ControlNets), loading in milliseconds with negligible VRAM overhead.
How to Use:
1. Check Input Image -> Image Prompt.
2. Load your reference pose, lineart, or sketch image.
3. Select Anima-LLLite (automatically selected when using Anima presets).
4. Fooocus will automatically download the default any-test-like-v2.safetensors model on first use.
Optimized Defaults: Default parameters are calibrated to Stop At: 0.4 and Weight: 0.2. This prevents over-saturation or burned colors when combined with Upscale (1.5x / 2.0x), where LLLite operates on both generation and upscale refinement passes.
Custom LLLite Models: You can drop additional Anima LLLite models into models/controlnet/ or models/prompt_expansion/controlnet/.

🖼️ 3-Stage Gallery Zoom & Lightbox Viewer

Fooocus includes a fluid 3-stage image viewing workflow designed for evaluating generated outputs at any zoom level:

Stage 1 — Grid Thumbnails: View all generated batch outputs side-by-side in a responsive grid. Clicking any thumbnail seamlessly switches the gallery to that image in the canvas view.
Stage 2 — Canvas Detailed View: Inspect the image fitted comfortably within the WebUI workspace while retaining immediate access to generation settings, seeds, and metadata.
Stage 3 — Fullscreen Lightbox Modal: Click the canvas image to open the centered fullscreen lightbox overlay.
- Keyboard Navigation: Press ArrowLeft / ArrowRight to cycle through your generation history without closing the viewer.
- Quick Close: Press Esc, click the dark background, or click (X) to exit back to the canvas view.

UI Access and Authentication

In addition to running on localhost, Fooocus can also expose its UI in two ways:

Local UI listener: use --listen (specify port e.g. with --port 8888).
API access: use --share (registers an endpoint at .gradio.live).

In both ways the access is unauthenticated by default. You can add basic authentication by creating a file called auth.json in the main directory, which contains a list of JSON objects with the keys user and pass (see example in auth-example.json).

List of Architecture Optimizations

Click to view built-in sampling and pipeline optimizations.

Prompt Expansion: Offline GPT-2 dynamic prompt enhancement engine ("Fooocus V2").
Native Refiner Swap: Seamless momentum-preserving model swap inside a single k-sampler for SDXL.
Negative ADM Guidance: Compensates for lack of cross-attention contrast in SDXL highest resolution level.
Self-Attention Guidance (SAG): Anisotropic kernel variation of SAG for artifact prevention and structural preservation.
Style Normalization: Balanced multi-style blending and automatic A1111-compatible prompt emphasizing.
Rectified Flow Matching: Shifted Euler Ancestral scheduling tuned for DiT (Anima) models.
Dual-Path VAE Decode: Specialized handling for both 4-channel 2D KL-Autoencoder and 16-channel 3D Causal WanVAE.
Sequential CFG Batching: Sequential positive/unconditioned passes preventing high-resolution activation spikes.
Zero-VRAM Step Preview: Instant linear RGB latent projection preview for DiT latents.
DiT ControlNet-LLLite Injection: Low-rank linear adapters and ASPP conditioning extractor for DiT models.
3-Stage Image Inspection: Gradio gallery pass-through to canvas view with centered fullscreen lightbox modal.

Customization

After the first time you run Fooocus, a config file will be generated at Fooocus\config.txt. This file can be edited to change the model path or default parameters.

For example, an edited Fooocus\config.txt (this file will be generated after the first launch) may look like this:

{
    "path_checkpoints": "D:\\Fooocus\\models\\checkpoints",
    "path_loras": "D:\\Fooocus\\models\\loras",
    "path_embeddings": "D:\\Fooocus\\models\\embeddings",
    "path_vae_approx": "D:\\Fooocus\\models\\vae_approx",
    "path_upscale_models": "D:\\Fooocus\\models\\upscale_models",
    "path_inpaint": "D:\\Fooocus\\models\\inpaint",
    "path_controlnet": "D:\\Fooocus\\models\\controlnet",
    "path_clip_vision": "D:\\Fooocus\\models\\clip_vision",
    "path_fooocus_expansion": "D:\\Fooocus\\models\\prompt_expansion\\fooocus_expansion",
    "path_outputs": "D:\\Fooocus\\outputs",
    "default_model": "realisticStockPhoto_v10.safetensors",
    "default_refiner": "",
    "default_upscale_model": "Default (Fooocus)",
    "default_loras": [["lora_filename_1.safetensors", 0.5], ["lora_filename_2.safetensors", 0.5]],
    "default_cfg_scale": 3.0,
    "default_sampler": "dpmpp_2m",
    "default_scheduler": "karras",
    "default_negative_prompt": "low quality",
    "default_positive_prompt": "",
    "default_styles": [
        "Fooocus V2",
        "Fooocus Photograph",
        "Fooocus Negative"
    ]
}

Many other keys, formats, and examples are in Fooocus\config_modification_tutorial.txt (this file will be generated after the first launch).

Consider twice before you really change the config. If you find yourself breaking things, just delete Fooocus\config.txt. Fooocus will go back to default.

A safer way is just to try "run_anime.bat" or "run_realistic.bat" - they should already be good enough for different tasks.

All CMD Flags

entry_with_update.py  [-h] [--listen [IP]] [--port PORT]
                      [--disable-header-check [ORIGIN]]
                      [--web-upload-size WEB_UPLOAD_SIZE]
                      [--hf-mirror HF_MIRROR]
                      [--external-working-path PATH [PATH ...]]
                      [--output-path OUTPUT_PATH]
                      [--temp-path TEMP_PATH] [--cache-path CACHE_PATH]
                      [--in-browser] [--disable-in-browser]
                      [--gpu-device-id DEVICE_ID]
                      [--async-cuda-allocation | --disable-async-cuda-allocation]
                      [--disable-attention-upcast]
                      [--all-in-fp32 | --all-in-fp16]
                      [--unet-in-bf16 | --unet-in-fp16 | --unet-in-fp8-e4m3fn | --unet-in-fp8-e5m2]
                      [--vae-in-fp16 | --vae-in-fp32 | --vae-in-bf16]
                      [--vae-in-cpu]
                      [--clip-in-fp8-e4m3fn | --clip-in-fp8-e5m2 | --clip-in-fp16 | --clip-in-fp32]
                      [--directml [DIRECTML_DEVICE]]
                      [--disable-ipex-hijack]
                      [--preview-option [none,auto,fast,taesd]]
                      [--attention-split | --attention-quad | --attention-pytorch]
                      [--disable-xformers]
                      [--always-gpu | --always-high-vram | --always-normal-vram | --always-low-vram | --always-no-vram | --always-cpu [CPU_NUM_THREADS]]
                      [--always-offload-from-vram]
                      [--pytorch-deterministic] [--disable-server-log]
                      [--debug-mode] [--is-windows-embedded-python]
                      [--disable-server-info] [--multi-user] [--share]
                      [--preset PRESET] [--disable-preset-selection]
                      [--language LANGUAGE]
                      [--disable-offload-from-vram] [--theme THEME]
                      [--disable-image-log] [--disable-analytics]
                      [--disable-metadata] [--disable-preset-download]
                      [--disable-enhance-output-sorting]
                      [--enable-auto-describe-image]
                      [--always-download-new-model]
                      [--skip-pip]
                      [--rebuild-hash-cache [CPU_NUM_THREADS]]

Inline Prompt Features

Wildcards

Example prompt: __color__ flower

Processed for positive and negative prompt.

Selects a random wildcard from a predefined list of options, in this case the wildcards/color.txt file. The wildcard will be replaced with a random color (randomness based on seed). You can also disable randomness and process a wildcard file from top to bottom by enabling the checkbox Read wildcards in order in Developer Debug Mode.

Wildcards can be nested and combined, and multiple wildcards can be used in the same prompt (example see wildcards/color_flower.txt).

Array Processing

Example prompt: [[red, green, blue]] flower

Processed only for positive prompt.

Processes the array from left to right, generating a separate image for each element in the array. In this case 3 images would be generated, one for each color. Increase the image number to 3 to generate all 3 variants.

Arrays can not be nested, but multiple arrays can be used in the same prompt. Does support inline LoRAs as array elements!

Inline LoRAs

Example prompt: flower <lora:sunflowers:1.2>

Processed only for positive prompt.

Applies a LoRA to the prompt. The LoRA file must be located in the models/loras directory.

Advanced Features

Click here to browse the advanced features.

Forks

Below are some Forks to Fooocus:

Fooocus' forks
fenneishi/Fooocus-Control runew0lf/RuinedFooocus MoonRide303/Fooocus-MRE mashb1t/Fooocus and so on ...

Thanks

Many thanks to twri and 3Diva and Marc K3nt3L for creating additional SDXL styles available in Fooocus.

The project starts from a mixture of Stable Diffusion WebUI and ComfyUI codebases.

Also, thanks daswer123 for contributing the Canvas Zoom!

Update Log

The log is here.

Localization/Translation/I18N

You can put json files in the language folder to translate the user interface.

For example, below is the content of Fooocus/language/example.json:

{
  "Generate": "生成",
  "Input Image": "入力画像",
  "Advanced": "고급",
  "SAI 3D Model": "SAI 3D Modèle"
}

If you add --language example arg, Fooocus will read Fooocus/language/example.json to translate the UI.

For example, you can edit the ending line of Windows run.bat as

.\python_embeded\python.exe -s Fooocus\entry_with_update.py --language example

Or run_anime.bat as

.\python_embeded\python.exe -s Fooocus\entry_with_update.py --language example --preset anime

Or run_realistic.bat as

.\python_embeded\python.exe -s Fooocus\entry_with_update.py --language example --preset realistic

For practical translation, you may create your own file like Fooocus/language/jp.json or Fooocus/language/cn.json and then use flag --language jp or --language cn. Apparently, these files do not exist now. We need your help to create these files!

Note that if no --language is given and at the same time Fooocus/language/default.json exists, Fooocus will always load Fooocus/language/default.json for translation. By default, the file Fooocus/language/default.json does not exist.

Name		Name	Last commit message	Last commit date
.github		.github
css		css
extras		extras
javascript		javascript
language		language
ldm_patched		ldm_patched
models		models
modules		modules
presets		presets
sdxl_styles		sdxl_styles
tests		tests
wildcards		wildcards
.dockerignore		.dockerignore
.gitattributes		.gitattributes
.gitignore		.gitignore
Dockerfile		Dockerfile
LICENSE		LICENSE
args_manager.py		args_manager.py
auth-example.json		auth-example.json
build_launcher.py		build_launcher.py
development.md		development.md
docker-compose.yml		docker-compose.yml
docker.md		docker.md
entry_with_update.py		entry_with_update.py
entrypoint.sh		entrypoint.sh
environment.yaml		environment.yaml
experiments_expansion.py		experiments_expansion.py
experiments_face.py		experiments_face.py
experiments_interrogate.py		experiments_interrogate.py
experiments_mask_generation.py		experiments_mask_generation.py
fooocus_colab.ipynb		fooocus_colab.ipynb
fooocus_version.py		fooocus_version.py
launch.py		launch.py
notification-example.mp3		notification-example.mp3
readme.md		readme.md
requirements_docker.txt		requirements_docker.txt
requirements_versions.txt		requirements_versions.txt
shared.py		shared.py
troubleshoot.md		troubleshoot.md
update_log.md		update_log.md
webui.py		webui.py

Latest commit

History

Folders and files

Repository files navigation

Fooocus (SayMaven Fork - SDXL & Anima DiT Dual-Architecture)

🚀 SayMaven Fork Highlights: SDXL & Anima DiT Support

Features

Download

Windows

Colab (Modern Runtime Support - SayMaven Fork)

Linux (Using Anaconda)

Linux (Using Python Venv)

Linux (Using native system Python)

Linux (AMD GPUs)

Windows (AMD GPUs)

Mac

Docker

Download Previous Version

Minimal Requirement

Troubleshoot

Default Models

🔑 Civitai API Token Authentication

How to configure:

1. Via .env File (Recommended for Local Windows / Linux)

2. Via Google Colab Notebook

3. Via Windows Batch Script (run.bat)

4. Via Terminal (PowerShell / Bash)

🔍 Custom Upscalers & Super-Resolution Models

How to Use:

Key Advantages:

🎨 Anima ControlNet-LLLite (Structure & Pose Conditioning)

Key Highlights:

🖼️ 3-Stage Gallery Zoom & Lightbox Viewer

UI Access and Authentication

List of Architecture Optimizations

Customization

All CMD Flags

Inline Prompt Features

Wildcards

Array Processing

Inline LoRAs

Advanced Features

Forks

Thanks

Update Log

Localization/Translation/I18N

About

Topics

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages

1. Via `.env` File (Recommended for Local Windows / Linux)

3. Via Windows Batch Script (`run.bat`)