⚠️ Disclaimer: Example/thumbnail images are AI-generated. No real persons were depicted or harmed. Use this workflow responsibly. Do not create images of real people without consent. I am not responsible for any damage or consequences. You are solely responsible for your actions.
Inpaint Example:
Image Edit Example:
Workflow Overview:
🔥 Advanced Image Edit & Inpaint Workflow for ComfyUI
A powerful all-in-one ComfyUI workflow designed for high-quality image editing and inpainting.
✨ Key Features:
• Smart image editing & precise inpainting
• Built-in Enhance Prompt system for better results
• Ready-to-use Presets for common editing styles
• Full LoRA support with strength control & loading guide
• Optimized for speed (Turbo mode) and quality
Simply load your image, choose a preset or write your own prompt, and generate. Perfect for creative editing, style changes, and detailed modifications.
Description
FAQ
Comments (12)
How can I ensure that the output image automatically retains the same dimensions as the original image? Thank you
Leave the resolution 0 and disable the upscale node beneath the primary image
屌
you sir, are a god among men. Thank you!
Amazing work
My first ever comment on here! Excellent work! I got it running ok on a rtx2060 6gb VRAM , 40gb Sys Ram. 3 mins an image, which is way faster than the other shite I've been using for Qwen. Most appreciated!
Unable to download Qwen-Image-2.1 NSFW Image EditV3.safetensors - file not found. Any help would be much appreciated :)
Use the link provided in the Notes node and place the model file in the Diffusion Model node. If you're using a different checkpoint, change it in the Load Diffusion Model node.
If you're using the ComfyUI desktop app, you can find the model files in the "ComfyUI/ComfyUI_shared" folder.
I hope this helps! 😊
@SkullReaper_22 Qwen-Image-2.1 NSFW Image EditV3.safetensors lora from the link in notes https://civitai.red/models/2994337/qwen-image-21-edit-lora-nude-edit?modelVersionId=3395802
It is unavailable
The pixels generated in the painting region has some position shift. Is there any idea to solve this?
I wouldn’t recommend using inpaint for heavy edits just yet, as I’m still figuring out how to fix the position-shift issue. For now, I’d suggest using it mainly for clothing swaps. Also, try to inpaint only the area you actually want to change instead of masking a large portion of the image. I’ve noticed that including areas that don’t need editing can sometimes cause unnecessary changes to details that were already correct. Keep the mask tight around the Cloths, with a little margin around the edges for blending. For heavier edits, I’d recommend using the normal Edit instead of inpaint.
Thank you very much. A fantastic, versatile, and user-friendly workflow that is a real pleasure to use. I’m very happy with it. I also couldn't find the LoRA file Qwen-Image-2.1 NSFW Image EditV3.safetensors. So, I downloaded Qwen-Image-2.1 NSFW Image EditV2.safetensors from Hugging Face and renamed it. It works perfectly for me.