Sync-3 is our most capable lipsync model and it is available right now in the Lipsync tool. Feed it a video and an audio track and it returns accurate, natural-looking mouth movements at professional quality. No more stiff or slightly-off results. The sync is tight, the motion feels real, and it works across a wide range of content types.
What Is Sync-3?
Sync-3 is the latest generation of our lipsync engine. It was built to handle the cases that trip up older models: subtle jaw movement, realistic lip closure, and the way speech shapes the entire lower face - not just the mouth. The result is sync that holds up under close viewing, not just at a glance.
It also introduces Sync Mode, a new feature that gives you explicit control over what happens when the audio and video are different lengths. Previously you left that to chance. Now you choose.
What Makes It Stand Out
Accurate, natural-looking mouth movements. Sync-3 tracks the full phoneme shape of the audio and maps it to believable lip motion. Syllables land where they should. Hard consonants close. Vowels open. The face looks like it is actually speaking.
Works on any video type. Live-action footage, 3D rendered characters, and AI-generated avatars all respond well. You do not need a perfectly controlled input to get clean output.
Studio-grade results. The quality sits at a professional level out of the box. No manual cleanup, no frame-by-frame correction. Generate once and use it.
Sync Mode: you decide how audio fits video. When the audio and video clips are different lengths, pick from five alignment options: cut off, loop, bounce, silence, or remap. Each one gives you a different outcome and you can choose the one that fits your project instead of getting a random result.
How to Use It
It is already in your workspace.
- Open the Lipsync tool in Kolbo
- Upload your video clip and your audio track
- In the model selector, choose Sync-3
- If the audio and video lengths differ, choose your Sync Mode from the options
- Generate
No additional setup. Your credits work exactly the same way.
What It Is Best For
Sync-3 performs especially well for:
- AI avatar videos - characters generated in any image model, now fully lip-synced to voiceover or dialogue
- 3D animated content - rendered characters with clean mouth tracking applied in one step
- Live-action dubbing - swap the audio track on existing footage with accurate mouth sync
- Content in multiple languages - translate the audio, apply Sync-3, and the mouth movements follow
- Social media videos - tight, polished sync for short-form content where every frame matters
Sync-3 is live in your Kolbo workspace right now. Drop in a video and an audio track and see the difference.
Try Sync-3 →Best, Zohar Founder, Kolbo.AI
