LanPaint

LanPaint_AVDecode

Decode the nested AV latent, merge the inpainted video and audio with the original (mask-blended), and write the result with the original fps and bit depth.

Video · 6 inputs · 2 parameters · 2 outputs

LanPaint_AVEncode

Encode a video's frames and audio into a nested AV latent with per-stream video and audio masks (1 = regenerate, 0 = keep).

Video · 5 inputs · 0 parameters · 1 output

LanPaint_ImageDecode

Decode an inpainted latent, resize to the original image's exact dimensions, and merge with the original inside the mask (replaces VAEDecode + MaskBlend).

Image · 4 inputs · 1 parameter · 1 output

LanPaint_ImageEncode

Encode an image and attach an inpainting mask to the latent (replaces VAEEncode + SetLatentNoiseMask).

Image · 3 inputs · 0 parameters · 1 output

LanPaint_KSampler

Uses the provided model, positive and negative conditioning to denoise the latent image.

Sampling · 1 output

LanPaint_KSamplerAdvanced

Full parameter control: Key Parameters | Parameter | Range | Description | |-----------|-------|-------------| | Steps | 0-100 | Total steps of diffusion sampling. Higher means better inpainting. Recommend 20-50. | | LanPaintNumSteps | 0-20 | Reasoning iterations per denoising step ("thinking depth"). Easy task: 2-5. Hard task: 5-10 | | LanPaintLambda | 0.1-50 | Content alignment strength (higher = stricter). Recommend 4.0 - 10.0 | | LanPaintStepSize | 0.1-1.0 | The StepSize of each thinking step. Recommend 0.1-0.5. | | LanPaintBeta | 0.1-2.0 | The StepSize ratio between masked / unmasked region. Small value can compensate high lambda values. Recommend 1.0 | | LanPaintFriction | 0.0-100.0 | The friction of Langevin dynamics. Higher means more slow but stable, lower means fast but unstable. Recommend 10.0 - 20.0| | LanPaintEarlyStop | 0-10 | Stop LanPaint iteration before the final sampling step. Helps to remove artifacts in some cases. Recommend 1-5| | LanPaintPromptMode | Image First / Prompt First | Image First mode focuses on the image context, maybe ignore prompt. Prompt First mode focuses more on the prompt. | For detailed descriptions of each parameter, simply hover your mouse over the corresponding input field to view tooltips with additional information.

Sampling · 1 output

LanPaint_MaskBlend

This node blends the original image with the inpainted image based on the mask. It is useful if you want the unmasked region to match the original image pixel perfectly.

Image/Postprocessing · 3 inputs · 1 parameter · 1 output

LanPaint_MiniMaxAudioDecode

Decode a MiniMax H3 audio latent to a waveform.

Audio · 2 inputs · 0 parameters · 1 output

LanPaint_MiniMaxAudioEncode

Encode audio with the MiniMax H3 audio VAE (correct layout). Audio inpainting masks come from the video mask editor via SetLatentNoiseMask.

Audio · 2 inputs · 0 parameters · 1 output

LanPaint_SamplerCustom

Runs custom sampling with the supplied model, sampler, sigmas, conditioning, and latent. The add_noise input controls noise generation; LanPaint parameters configure the sampling model. Returns sampled and denoised LATENT outputs.

Sampling/Custom Sampling · 6 inputs · 6 parameters · 2 outputs

LanPaint_SamplerCustomAdvanced

Runs the supplied guider with noise, sampler, sigmas, and latent, applying LanPaint step count, lambda, step size, and prompt mode. Uses the latent noise mask when present and returns sampled and denoised LATENT outputs.

Sampling/Custom Sampling · 5 inputs · 5 parameters · 2 outputs

LanPaint_UpSale_LatentNoiseMask

Copies the LATENT and attaches a noise_mask initialized to ones with zeros at every scale-th spatial coordinate. The scale input is an integer from 2 through 8.

Latent/Inpaint · 1 input · 1 parameter · 1 output

LanPaint_VideoMaskEditor

Loads a video (like LoadVideo) and also outputs a per-frame video inpainting mask and a per-frame audio mask painted in the mask editor (1 = regenerate, 0 = keep).

Video · 3 outputs