Skip to content

LanPaint

Image inpainting for Stable Diffusion without extra training

Visit official site

What LanPaint is

The plugin is described as an image inpainting plugin for Stable Diffusion models that reaches high quality repairs through multi-round iterative inference without extra training. It is stated to work with any Stable Diffusion model, including user custom models.

What you can do with it

  • Run zero-training inpainting on any Stable Diffusion model
  • Use multi-round iterative inference for seamless repairs
  • Adjust advanced parameters to fine-tune the result
  • Install the plugin quickly through ComfyUI-Manager

Getting started

  • Set up the ComfyUI environment following the official guide
  • Install ComfyUI-Manager to manage extensions
  • Search for and install LanPaint through ComfyUI-Manager, or clone the repository into the target folder
  • Restart ComfyUI so the LanPaint nodes load

Example uses

  • Repair a basket into a basketball
  • Replace a white shirt with a blue shirt
  • Change a smiling expression into a sad one

Pros & cons

✓ What we like

  • Advanced parameters let users fine-tune repair results
  • Custom Stable Diffusion models are supported

! What to watch out for

  • A ComfyUI setup and ComfyUI-Manager are part of the install path
  • Advanced parameters are exposed for manual tuning

FAQ

Does it need extra training?

The feature list states it works with any Stable Diffusion model with no extra training.

How is it installed?

The steps list installing ComfyUI, adding ComfyUI-Manager, installing LanPaint through it, and restarting ComfyUI.

Last reviewed: 2026-11-15

More AI image generation tools

View all →

How we review