#47 · Primary category: Image Generation

LanPaint

audio-processing comfyui comfyui-custom-node comfyui-custom-nodes comfyui-nodes diffusion-inpainting diffusion-models generative-ai generative-model image-editing inpainting inpainting-algorithm video-editing

High quality training free inpaint for every stable diffusion model. Supports ComfyUI

Project last updated:08/12/26

GitHub Stars

1.4K

Forks

50

Contributors

8

License

GPL-3.0

Why we included this project

LanPaint is a training-free sampler that improves inpainting results on pretty much any Stable Diffusion checkpoint you already have, without retraining or fine-tuning. It works by letting the model run several refinement passes before the final denoise, so you can spend a little extra compute and get cleaner fills and fewer artifacts around mask edges. It ships as a ComfyUI extension, which fits the node-based workflow most diffusion users are already in, and there's a Diffusers port if you'd rather call the same logic from your own Python scripts. If your image-editing work involves object removal, background replacement, or fixing damaged areas, this is a practical way to get better output from models that are already in use.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category