One singer's mouth kit (13 mouth shapes from their own face): the "mouth LoRA" every dub needs. Bring a kit folder, or build one from a shot of the source with…

Lip-sync for music videos in ComfyUI: dub the singers' mouths to the song, shot by shot, with QA. Every input is a slot (bring a clip and song, or generate them with LTX-2.5, Qwen-Image-Edit and ACE-Step); vocal split, lyric alignment, lip kits, sync QA.

LSDub

project

identity

kit_dir

source

shots

shot_id

keep

seeds

One singer's mouth kit (13 mouth shapes from their own face): the "mouth LoRA" every dub needs. Bring a kit folder, or build one from a shot of the source with Qwen-Image-Edit-2511 (open weights; local GPU or Comfy Cloud per the project, ~39 edits).

Name	Type	Default	Description
project	LSDUB_PROJECT		The LSDub project.
identity	STRING		Who this kit is for: the singer's cast name.
kit_dir	STRING		A finished kit folder, or a .zip of one (a zip travels with `comfy deploy run`; a folder does not). Leave empty to build one from the source.
sourceopt	LSDUB_SOURCE		Build from this video: the frame with the most closed mouth of the chosen shot is the base face.
shotsopt	LSDUB_SHOTS		The shots, to pick shot_id from.
shot_idopt	INT	-1-1–100000	The shot to build from (a close-up of this singer); -1 = the whole source.
keepopt	STRING	person, same face shape, skin tone, hair, eyewear and makeup	Who to keep: describes the singer, after "the same ...", so the edits keep the face.
seedsopt	INT	31–6	Takes per mouth shape; the best-measured take is kept.

LSDub Lip Kit

Inputs (8)

Outputs (1)