How to Add AI Lip Sync to a Music Video in BeatViz Workflow
Lip sync matches a visible character's mouth movement to the corresponding audio section. In BeatViz Workflow, lip sync is performed on individual generated scene clips, not on an Editor-style timeline.
Option 1: Automatic lip sync with Singing
Choose Singing (Beta) under Content Mode before starting the task. Singing uses a vocal-oriented analysis path and can identify lyric timings. After you confirm the storyline, inspect the first frames and generate the scene videos, Workflow runs the automatic lip-sync stage once the base clips are generated.
When you select Generate All Videos, the credit requirement includes both video generation and Singing lip sync. The scene card can display Syncing lips while that stage is processing. When completed, the scene preview uses the lip-synced video. If lip sync fails, a retry is available under the task's failure and credit rules.
Important: Singing is a Beta option, and selecting it does not promise flawless mouth movement for every angle or face.
Option 2: Manual lip sync with other Content Modes
For Storytelling, Dancing and Abstract, wait for the scene's base video to finish generating. The scene card then offers Lip Sync. Select it to process that scene using the matching portion of your selected song audio. This is a separate charged action, rather than part of the initial video-generation charge.
During processing, the card's lip-sync control is locked to prevent repeated submission. A scene already processed successfully is not charged again for the same completed action.
Current Workflow lip-sync costs
Lip sync is 20 credits per second, with fractional durations rounded up according to the product's charging rules. Examples:
| Clip duration | Lip-sync credits |
|---|---|
| 5 seconds | 100 |
| 8 seconds | 160 |
| 10 seconds | 200 |
| 15 seconds | 300 |
When the interface presents a duration tier, shorter partial durations may be billed according to the corresponding tier; check the displayed charge before submitting. A low balance opens the pricing interface instead of starting the job.
How to improve the likelihood of useful results
- Use a clear first frame with the intended person visible.
- Favor scenes where the face is clearly presented when visible singing is important.
- Check that the song section matches the intended vocal moment.
- Inspect the lip-synced clip before moving to final composition.
These are practical suggestions, not a guarantee of perfect output. Scene selection and model behavior still affect results.
Workflow Lip Sync vs Editor Lip Sync
Workflow's lip sync is attached to scene cards and uses the clip's corresponding audio section. It does not use the Editor timeline's 1× playback, minimum 0.5-second clip, or timeline-end constraints. Do not apply Editor-specific lip-sync instructions to Workflow.
FAQs
Can I lip-sync a Storytelling scene?
Yes. Use that scene's Lip Sync button after video generation completes.
Will Singing charge lip sync twice?
Singing includes lip-sync credits at the video-generation stage. Do not treat the automatic stage as an additional manual button charge.
Is lip sync available before the clip generates?
The manual Lip Sync action requires a finished video clip. Singing runs its automatic pass after generating the base video.
Continue learning: ← Previous: Keep Characters Consistent · Next: Credits & Costs →
Related guides
Try BeatViz Workflow
Upload a song, choose a Content Mode, and let Workflow plan scenes, first frames, and clips for you.