SyncLipSyncNode - ComfyUI Built-in Node Documentation

This node re-syncs mouth movement in a video to new speech audio using the sync.so API. It handles close-ups, profiles and obstructions automatically while preserving the speaker’s expression. Cost scales with output duration.

Inputs

Parameter	Description	Data Type	Required	Range
`video`	Footage of the speaker to re-sync. Up to 4K (4096x2160); a constant frame rate of 24/25/30 fps works best.	VIDEO	Yes	-
`audio`	Speech audio to sync the mouth to.	AUDIO	Yes	-
`seed`	Seed controls whether the node should re-run; results are non-deterministic regardless of seed (default: 42).	INT	Yes	0 to 2147483647
`model`	sync.so generation model.	COMBO	Yes	See below

The model parameter is a dynamic combo that includes the following sub-parameters:

Sub-parameter	Description	Data Type	Required	Range
`sync_mode`	How to handle a duration mismatch between video and audio; this also sets the output length (default: “bounce”).	COMBO	Yes	`"bounce"` `"cut_off"` `"loop"` `"silence"` `"remap"`
`speaker_selection`	Which face to lipsync when several people are visible (default: “default”).	COMBO	Yes	`"default"` `"auto-detect"` `"coordinates"`
`speaker_frame`	Video frame used to locate the speaker. Only used when `speaker_selection` is “coordinates” (default: 0).	INT	No	0 to 1000000
`speaker_x`	X pixel coordinate of the speaker’s face. Only used when `speaker_selection` is “coordinates” (default: 0).	INT	No	0 to 4096
`speaker_y`	Y pixel coordinate of the speaker’s face. Only used when `speaker_selection` is “coordinates” (default: 0).	INT	No	0 to 4096

Sync mode options:

bounce: Video plays forward then backward until the audio ends (output = audio length)
loop: Video restarts until the audio ends (output = audio length)
remap: Video is time-stretched to match the audio (output = audio length)
cut_off: The longer track is trimmed (output = shorter length)
silence: Nothing is trimmed; the shorter track is padded (output = longer length)

Speaker selection options:

default: Let the model decide which face to lipsync
auto-detect: Detect and follow the active speaker
coordinates: Target the face at pixel (speaker_x, speaker_y) in the frame chosen by speaker_frame

Constraints:

Video resolution must not exceed 4K (4096x2160). Videos above this limit will raise an error.
Audio duration must not exceed 600 seconds (10 minutes).
The speaker_frame, speaker_x, and speaker_y parameters are only used when speaker_selection is set to “coordinates”.

Outputs

Output Name	Description	Data Type
`video`	The re-synced video with mouth movement matched to the provided audio.	VIDEO

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Source fingerprint (SHA-256): b41f8c9bf0d55059f081a66af20636ec96462c3fd9caeb685cab10278f84678a

​Inputs

​Outputs

Inputs

Outputs