Inpaint Anyone or Anything Anywhere Doing Whatever Seamlessly With Krita, Kontext, and Nunchaku

Use this workflow with Krita and ComfyUI to insert an object into a background image: provide the background image, the image to insert, a selected target area, and a prompt, with an edited image as the output.

Load the background image in Krita and the image to insert in ComfyUI.

Select the area for the insertion, then adjust the prompt to guide the edit.

Inputs are a background image, an image to insert, a selected area, and a prompt; the output is an edited image with the object inserted.

Before editing, open the background image in Krita and the image to insert in ComfyUI.

Node packs listed include comfyui_essentials, comfyui-easy-use, comfyui-inpaint-cropandstitch, comfyui-kjnodes, comfyui-logic, comfyui-logicutils, comfyui-nunchaku, and comfyui-rmbg.

Other named node packs are comfyui-tooling-nodes and pr-was-node-suite-comfyui-47064894.

Model files listed include flux.1 turbo alpha.safetensors, flux1DevVAE_safetensors.safetensors, Long-ViT-L-14-BEST-GmP-smooth-ft.safetensors, svdq-int4_r32-flux.1-kontext-dev.safetensors, and t5xxl_fp8_e4m3fn.safetensors.

The listed base model is Flux.1 Kontext.

This workflow can be adapted for in-browser ComfyUI inpainting.

Tip: After selecting the insertion area, try different prompts to guide the edit.

Sugestão · não verificado

Before use, verify these filenames in your setup: flux.1 turbo alpha.safetensors, flux1DevVAE_safetensors.safetensors, Long-ViT-L-14-BEST-GmP-smooth-ft.safetensors, and pr-was-node-suite-comfyui-47064894.

Sugestão · não verificado

No license information is listed for this workflow.