A Qt frontend for image generation that uses ComfyUI
Supports Windows, Linux and Mac(?)
Features:
- Supports SD 1.x, SD 2.x, SDXL, SDXL Refiner
- txt2img, img2img, hires fix
- Inpainting
- Outpainting
- Loras
- Upscaling
- Prompt styles
- Controlnet
Installation:
First make sure Python (3.10+ recommended) and GIT are installed. Then:
git clone https://github.com/shinomakoi/SnugQt
cd SnugQt
Optionally create a virtual environment (recommended)
python -m venv .venv
source ./.venv/bin/activate ### For Linux
.\.venv\Scripts\activate ## For Windows
pip install -r requirements.txt
You can update to latest version with: git pull
Usage:
Make sure you have launched ComfyUI first, then to launch SnugQt run:
python main.py
You can change settings in File > Settings. Make sure to set the path to the 'ComfyUI/models' folder (click the ... button to select). Images are saved in the ComfyUI/ouput/SnugQt folder.
