Description:
This workflow uses Nunchaku acceleration technology to speed up Zimage – Turbo image model generation, making it suitable for users with limited GPU memory. The workflow is designed to be operated within the DFCine canvas and requires the DigitFold ComfyUI integration package to be installed.
Large Model Downloads
- For 50-series GPUs: download
svdq-fp4_r128-z-image-turbo.safetensorsand place it in themodels\diffusion_models\folder. - For 30- or 40-series GPUs: download
svdq-int4_r256-z-image-turbo.safetensorsand place it in themodels\diffusion_models\folder.
Universal Model Downloads
ae.safetensors→ place inmodels\vae\qwen_3_4b.safetensors→ place inmodels\text_encoders\


