Synchronize speech with the Lip Sync node

Combine a source video and new audio, choose a supported lip-sync route, and reconcile differences in clip and speech duration.

Where: Flowgen → Add → Audio → Lip Sync

Key ideas

Two required media inputs: Connect Source Video and New Audio; the output is a new lip-synced video.
Model-specific controls: The selected route decides which settings appear: duration reconciliation, guidance/loop/seed, or start/end time and speech options.
Duration reconciliation: Supported routes expose cut off, loop, bounce, silence, or remap behavior; other routes can match dynamic duration.
Lip Sync card: LipSyncNode uses the generic shared-card path. Synchronize connected replacement audio to a source video with model-specific duration and speech controls.
Properties sidebar: avatarAgenticNodes.tsx
Opened surfaces: Video preview/lightbox

Connect the source video and replacement audio.
Choose the lip-sync route and review the controls that become available.
Set duration behavior; when offered, choose captions, speech enhancement, music-track handling, trim times, guidance, loop mode, or seed.
Generate and review the whole clip for mouth timing, cuts, repeated motion, audio quality, and caption accuracy.

Add path: Flowgen → Add → Sound → Lip Sync.
Card: LipSyncNode.
Properties: avatarAgenticNodes.tsx.
Controls: Model; video/audio; duration reconciliation; captions; speech enhancement; loop/guidance/seed; trim; disable music.
Opened surfaces: Video preview/lightbox.

Use a clear frontal or three-quarter face with visible mouth motion.
Trim long silence before syncing unless the chosen duration mode is meant to preserve it.

Controls differ by selected route; changing the model can add or remove fields.
Review identity, consent, and disclosure requirements before publishing a synthetic speech edit.

Choose a different duration-reconciliation mode, trim the source/audio, or switch to a route whose controls match the clip.