consistent character - flux fill ace plus
5.0
0 reviews
Description
https://github.com/ali-vilab/ACE_plus/tree/main/workflow/ComfyUI-ACE_Plus
https://huggingface.co/ali-vilab/ACE_Plus/blob/main/portrait/comfyui_portrait_lora64.safetensors
Note: You still need to generate multiple times to get a decent result—that's something the ACE Plus repo actually explains. Also, the Flux Fill model is basically a painting model focused on patching things up, so the output quality isn't great. You'll need to use another workflow to improve the final image.
Discussion
(No comments yet)
Loading...
Reviews
No reviews yet
Versions (2)
- latest (5 months ago)
- v20250424-045128
Node Details
Primitive Nodes (13)
ACEPlusLoraConditioning (1)
ACEPlusLoraProcessor (1)
Anything Everywhere3 (2)
DualCLIPLoaderGGUF (1)
FluxGuidance (1)
Image Inset Crop (rgthree) (1)
LayerMask: LoadSegmentAnythingModels (1)
LayerMask: SegmentAnythingUltra V3 (1)
LayerUtility: NumberCalculatorV2 (1)
ReduxAdvanced (1)
TeaCache (1)
UnetLoaderGGUF (1)
Custom Nodes (15)
ComfyUI
- KSampler (1)
- VAEDecode (1)
- ConditioningZeroOut (1)
- VAELoader (1)
- LoraLoaderModelOnly (1)
- CLIPVisionLoader (1)
- StyleModelLoader (1)
- LoadImage (1)
- ConditioningConcat (1)
- DifferentialDiffusion (1)
- CLIPTextEncode (1)
- PreviewImage (1)
- LayerUtility: GetImageSize (1)
- ImageAndMaskPreview (2)
Model Details
Checkpoints (0)
LoRAs (1)
comfyui_portrait_lora64.safetensors