Added IPADAPTER, improved Adetailer. Thanks to prodajie for feedback and suggested improvements!
Runtime profile
Remaining hardware or download-size estimates are not available yet.
AI-generated commentary
AI-generated explanation based on source and configuration details. Suggestions are clearly labeled.
This ComfyUI workflow is for inpainting and upscaling images, with FaceDetailer, ControlNet, and regional IPAdapter conditioning available; it takes an image and lets you select LoRAs, then preview or save an image.
Use shortcuts 1, 2, 3, and 4 to move around the workflow.
To use FaceDetailer, upscaling, or ControlNet, click Enable Group in Fast Groups Bypasser (rgthree) at the top left.
The listed features include CivitAI metadata and a token counter for SDXL, PONYV6, ILLUSTRIOUS, NOOBAI-XL, ANIMAGINEXL4, and SD1.5.
Before downloading, verify that the CivitAI page is for version v3 by Blinkdotleh.
Suggestion
The instructions place face and non-face SDXL IPAdapter Plus models in ComfyUI\models\ipadapter, CLIP Vision (ViT-H) in clip_vision, and Kim2091/UltraSharp in upscale_models.
The workflow lists these node packs: ComfyLiterals, ComfyUI_Comfyroll_CustomNodes, comfyui_controlnet_aux, comfyui_fearnworksnodes, comfyui_ipadapter_plus, ComfyUI_Mira, comfyui_smznodes, and comfyui_ultimatesdupscale.
It also lists comfyui-custom-scripts, comfyui-detail-daemon, comfyui-image-saver, comfyui-impact-pack, comfyui-impact-subpack, comfyui-inpaint-cropandstitch, and rgthree-comfy.
The workflow lists these model files: bbox/face_yolov9c.pt, clip_vision_g.safetensors, Comfyui_00001_.safetensors, controlnet-union-sdxl-1-promax.safetensors, djmilxlvae_dminVae.safetensors, noobIPAMARK1_mark1.safetensors, RealESRGAN_x4plus_anime_6B.pth, sam_vit_b_01ec64.pth, and
For anime, consider leaving FreeU_V2 bypassed; the author says it does not work well with anime.
Suggestion
For SDXL-based checkpoints, the instructions link a ControlNet model.
The linked 3D posing site requires a browser with hardware acceleration; the author says the workflow is best on a 1440p display.
No license is listed for this workflow; check the CivitAI page's terms before reusing or sharing it.
Suggestion
Does this need to be edited?
Sign in to send an edit request.
Sources
1 sourceSource excerpts
2 excerptsSource context: 1971 downloads · Type Workflows · Base model Other
Requirements
Object Detector · Hugging Face
Vision model · 3.44 GB · SAFETENSORS · fp16 · full · Unknown
Comfyui_00001_.safetensors
Not resolvedCheckpoint · Unknown
controlnet-union-sdxl-1-promax.safetensors
Not resolvedControlNet · Unknown
VAE · 319 MB · Civitai · 1257377
noobIPAMARK1_mark1.safetensors
Not resolvedCheckpoint · Unknown
Upscaler · 17.1 MB · PTH · Unknown
sam_vit_b_01ec64.pth
File unverifiedSegmentation · Unknown
Object Detector · Hugging Face
ComfyLiterals
FoundNode pack · Registry
ComfyUI_Comfyroll_CustomNodes
FoundNode pack · Registry
comfyui_controlnet_aux
PossibleNode pack · Registry
Source context: 1674 downloads · Type Workflows · Base model Other
Node pack · Registry
comfyui_ipadapter_plus
PossibleNode pack · Registry
ComfyUI_Mira
FoundNode pack · Registry
Node pack · Registry
comfyui_ultimatesdupscale
FoundNode pack · Registry
comfyui-custom-scripts
FoundNode pack · Registry
comfyui-detail-daemon
FoundNode pack · Registry
comfyui-image-saver
FoundNode pack · Registry
comfyui-impact-pack
FoundNode pack · Registry
comfyui-impact-subpack
FoundNode pack · Registry
comfyui-inpaint-cropandstitch
FoundNode pack · Registry
rgthree-comfy
FoundNode pack · Registry