AI lip-sync for AI video clips
Create offers two ways to get a character speaking on screen with the mouth in sync.
Two ways
- Speaking clip: write the line in quotes in a /video prompt and choose Veo 3.1 (6 generations; 8 with a character sheet) or Wan 3.0 (5 generations) β the character says it with lips in sync.
- Lip-sync afterwards: add a line to any clip with π Voice (a character's voice or a narrator), then press π Lip-sync β the mouth is re-timed to the line (about 3 generations per 5 seconds). The original clip is kept.
- A lip-synced shot of an episode replaces the original shot in the episode; the old one can be restored.
Questions
Which clips can be lip-synced?
Clips that carry a spoken line (added with π Voice or your own audio) and show a face clearly; up to 60 seconds per clip.
Do failed lip-syncs cost generations?
No β a failed job is refunded in full.
More from DexunShort Create
- AI storyboard with 3D previs and multi-camera coverage
- Where to publish AI short dramas and earn a revenue share
- Keep AI characters consistent across every shot
- AI screenwriting with Final Draft (.fdx) and Fountain export
- Make an AI short drama from script to finished episode
- AI writing assistant for TV series scripts and series bibles
- DexunShort Create vs LibTV
Updated 2026-10-14 Β· DexunShort Create Β· DEXUN INC