scraed/LanPaint

Health v2 · maintenance only

Adoption not part of the score

1361 stars · 50 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

LanPaint is a training-free inpainting sampler for stable diffusion models, implemented as a ComfyUI custom node. It uses iterative 'think mode' denoising to achieve high-quality image and video inpainting without model retraining.

Use cases

• inpaint missing regions in images with stable diffusion
• edit photos by painting over areas with a mask
• inpaint video frames with ComfyUI
• use any diffusion model for inpainting without training an inpainting model
• improve inpainting quality by trading compute for better results

When to choose

• you want high-quality inpainting with any stable diffusion checkpoint without a dedicated inpainting model
• you already use ComfyUI and want a drop-in sampler node
• you need video or audio inpainting support

When to avoid

• you need fast real-time inpainting with minimal compute
• you don't use ComfyUI or diffusers-based workflows
• you need a simple GUI editor rather than a node-based pipeline

Facets

plugin · maturity active

image-processing machine-learning stable-diffusion image-processing machine-learning media python cross-platform inpainting comfyui-custom-node diffusion-models training-free generative-ai video-inpainting image-editing

Member repositories

For agents

markdown · JSON · MCP: product_card(name="scraed/LanPaint")