Focus on prompting and generating
 
 
 
Go to file
Developer 8af0646df3 Add Z-Image dtype and turbo max-seq runtime overrides 2026-02-17 20:30:35 +02:00
.github ci: bump actions/checkout from 4 to 5 (#4085) 2025-09-02 22:28:40 +02:00
css fixes 2026-02-17 10:07:10 +02:00
extras Use weights_only for loading (#3427) 2024-08-03 12:33:01 +02:00
javascript fix 2026-02-17 11:06:58 +02:00
language initial 2025-10-07 17:14:17 +03:00
ldm_patched Use weights_only for loading (#3427) 2024-08-03 12:33:01 +02:00
models Add default text_encoders path and wire Z-Image custom lookup 2026-02-17 19:18:36 +02:00
modules Add Z-Image dtype and turbo max-seq runtime overrides 2026-02-17 20:30:35 +02:00
presets Sync branch 'mashb1t_main' with develop_upstream 2024-07-14 20:28:38 +02:00
sdxl_styles Sync branch 'mashb1t_main' with develop_upstream 2024-07-14 20:28:38 +02:00
tests change 2025-10-04 00:32:17 +03:00
wildcards feat: inline lora optimisations (#2967) 2024-05-20 17:31:51 +02:00
.dockerignore feat: build and push container image for ghcr.io, update docker.md, and other related fixes (#2805) 2024-05-23 00:19:54 +02:00
.gitattributes feat: configure line ending format LF for *.sh files (#2991) 2024-05-24 21:36:07 +02:00
.gitignore changes to negative promot location 2025-09-08 22:18:09 +03:00
Dockerfile fix: chown files directly at copy (#3066) 2024-05-31 22:41:36 +02:00
LICENSE Create LICENSE 2023-08-11 13:43:12 -07:00
args_manager.py feat: sort enhance images (mashb1t#62) 2024-07-25 15:21:56 +02:00
auth-example.json Added support for authentication in --share mode (via auth.json) 2023-10-24 18:29:37 +02:00
build_launcher.py Pass additional command line arguments when using run.bat scripts 2023-10-17 20:49:51 -04:00
development.md feat: add support for lora inline prompt references (#2323) 2024-05-18 17:19:46 +02:00
docker-compose.yml feat: build and push container image for ghcr.io, update docker.md, and other related fixes (#2805) 2024-05-23 00:19:54 +02:00
docker.md feat: build and push container image for ghcr.io, update docker.md, and other related fixes (#2805) 2024-05-23 00:19:54 +02:00
entry_with_update.py fixed typo in entry_with_update.py (#1306) 2023-12-14 09:22:49 -08:00
entrypoint.sh Revert "fix: use LF as line breaks for Docker entrypoint.sh (#2843)" (#2865) 2024-05-04 14:37:40 +02:00
environment.yaml i 2023-08-10 03:55:55 -07:00
experiments_expansion.py [2.1.831] Fooocus Describe 2023-12-12 19:34:18 -08:00
experiments_face.py [2.1.831] Fooocus Describe 2023-12-12 19:34:18 -08:00
experiments_interrogate.py [2.1.831] Fooocus Describe 2023-12-12 19:34:18 -08:00
experiments_mask_generation.py Sync branch 'mashb1t_main' with develop_upstream 2024-07-14 20:28:38 +02:00
fooocus_colab.ipynb chore: update packages #2 2024-07-16 16:31:15 +02:00
fooocus_version.py release: bump version to 2.5.5, update changelog 2024-08-12 08:10:20 +02:00
launch.py fix: move import to resolve colab issue (#3506) 2024-08-12 07:59:00 +02:00
notification-example.mp3 feat: change example audio file 2024-07-16 13:59:51 +02:00
readme.md Revert custom Z-Image text encoder and VAE overrides 2026-02-17 19:53:16 +02:00
requirements_docker.txt zimige 2026-02-17 15:40:22 +02:00
requirements_versions.txt zimige 2026-02-17 15:40:22 +02:00
shared.py feat: advanced params refactoring + prevent users from skipping/stopping other users tasks in queue (#981) 2024-02-24 19:01:06 +01:00
troubleshoot.md 2.1.844 2023-12-16 10:34:47 -08:00
update_log.md release: bump version to 2.5.5, update changelog 2024-08-12 08:10:20 +02:00
webui.py fix 2026-02-17 10:56:50 +02:00

readme.md

FooocusPocus

A Quality-of-Life Enhanced Fork of Fooocus

>>> Click Here to Install FooocusPocus <<<

About This Fork

FooocusPocus is a fork of Fooocus by lllyasviel.

Goal: keep core Fooocus generation behavior while improving usability, workflow, and configuration quality of life.

For full upstream capability docs, see the original Fooocus repository.


Highlights

Image Library (Major Feature)

The Image Library is an in-app browser for generated images with metadata-aware management.

  • Browse generated images in a dedicated modal gallery
  • Search images by prompt text
  • Filter by tags
  • Single-image preview with full metadata panel
  • Multi-select with checkboxes for bulk operations
  • Unified delete action:
    • Deletes one image when single-selected
    • Deletes multiple images when checkbox selection is active
  • Edit tags for one or many images
  • Optional auto-reload when opening the library

Configuration Tab

A dedicated tab for managing runtime settings without manual config editing.

  • Add/remove checkpoint folders and LoRA folders
  • Auto-reload models after folder changes
  • Edit key paths (output/temp/embeddings/VAE/ControlNet/upscale)
  • Set generation defaults (steps/CFG/sampler/scheduler/model/style/aspect ratio/output format)
  • Per-setting reset buttons and full restore-to-defaults
  • Auto-save configuration changes

UI and Workflow Improvements

  • Prompts are organized for faster day-to-day use
  • Better operation feedback and console logging
  • Random LoRA option for exploration

Prompt Syntax Reference

FooocusPocus supports both wildcard placeholders and dynamic prompt groups.

1) Wildcard files (__name__)

Use wildcard placeholders in prompts:

  • a portrait of __artist__
  • __color__ sports car in __city__

Wildcard files are loaded from your configured wildcards folder (path_wildcards), with one option per line.

2) Dynamic choices ({...})

Use inline dynamic groups:

  • Single choice: {red|green|blue}
  • Single choice (spaces are fine): {red | green | blue}
  • Multi-select count: {2$$red|green|blue|yellow}
  • Multi-select range: {1-3$$red|green|blue|yellow}

Both positive and negative prompts support this syntax.

3) Read wildcards in order

When enabled, wildcard file entries are consumed deterministically by index (useful for reproducible batches).
When disabled, wildcard entries are chosen randomly.


Download

Windows

You can download FooocusPocus from the Releases page.

After downloading, extract and run run.bat.

System Requirements

  • Minimum: 4GB Nvidia GPU VRAM and 8GB system RAM
  • Recommended: 6GB+ VRAM and 16GB+ RAM

Changes from Upstream (Summary)

Feature Description
Image Library In-app generated image browser with metadata, tag filter/search, multiselect, bulk delete, and tag editing
Configuration Tab UI-based settings management with auto-save and reset controls
Dynamic Model Folders Add/remove checkpoint and LoRA folders without restart
Prompt Utilities Wildcard placeholders and dynamic prompt groups
Random LoRA Optional random LoRA selection for experiments
UX/Logging Improved feedback and operational visibility

Original Fooocus Features

FooocusPocus includes upstream Fooocus capabilities, including:

  • High-quality text-to-image generation
  • Inpainting and outpainting
  • Image prompt workflows
  • Style systems and model switching
  • Upscale and variation workflows

For full documentation, see Fooocus upstream.


Contributing

Contributions are welcome via pull requests and issues.

License

This project inherits Fooocus licensing. See LICENSE.

Credits

  • Original Fooocus by lllyasviel
  • FooocusPocus enhancements by contributors

Note: This is an unofficial fork. For the official project, visit github.com/lllyasviel/Fooocus.