sync-3
Sync's flagship lipsync model and the current default for all users.
- Up to
- 1800 seconds
- Resolution
- 3840x2160
- Weights
- Not released
- API
- Yes
- Commercial use
- Not established
Processes an entire shot holistically rather than in snippets, which lets it handle occlusions, extreme angles and close-ups automatically, and it outputs native 4K for the face region with built-in super resolution. Uniquely in the family it accepts a static image plus audio or text, turning stills into talking video; it removes the manual temperature, reasoning_enabled and occlusion_detection_enabled parameters because those behaviours are now automatic.
The research notes for this model record other terms but do not state a commercial-use position, and no licence page was retrieved, so it is not established here. Check the vendor's own terms.
- Depends on your plan tier. Watermarks are removed on paid plans
- Output is watermarked. Watermarks are removed on paid plans
- Requires consent from the person depicted. Consent requirements for syncing a real person's face were not stated on the pages fetched
Specification
- What it takes in
- video, image, audio, text
- What it makes
- video
- Maximum duration
- 1800 seconds
- Maximum resolution
- 3840x2160
- Aspect ratios
- not established
- Controls
- lipsync, voice-clone, translation
- Weights
- Not released
- API
- Yes
- Pricing
- Subscription plus usage. Plans: Hobbyist $5/mo ($0.05/sec, 1 min max, 1 concurrent job), Creator $19/mo ($0.05/sec, 5 min max, 3 concurrent, 3 seats), Growth $49/mo ($0.0475/sec, 10 min max, 6 concurrent, 5% usage discount), Scale $249/mo ($0.04/sec, 30 min max, 15 concurrent, 5 seats, 20% usage discount, Batch API), Enterprise custom (unlimited length). sync-3 itself is priced at $0.107-$0.133 per second at 25fps. Free accounts get one 15-second generation per month.
- Released or updated
- not established
Supported aspect ratios is not established here. That means no page we could retrieve stated it, not that the model lacks one.
Strengths
- Native 4K face-region output, rare in lipsync models
- Only model in the family that accepts still images as input
- Handles occlusion and profile/extreme angles without manual flags
- Active speaker detection for multi-person shots
Weaknesses
- Most expensive model in the family at $0.107-$0.133/sec, roughly 5x lipsync-1.9
- Generates generic rather than speaker-matched lip movement during silent segments
- Maximum clip length is a plan attribute (1 to 30 minutes) rather than a model capability
- No published aspect ratio constraints
Best for
- Close-up and profile dialogue where earlier lipsync models break
- Visual dubbing of premium footage
- Still-photo-to-talking-video
Where the sources disagree
Recorded rather than resolved. Picking the tidier number would hide the fact that the vendor's own pages do not agree.
- maxDurationSeconds of 1800 is the Scale-plan ceiling; Hobbyist is 60s, Creator 300s, Growth 600s and Enterprise unlimited. The docs index describes sync-3 as 'native 4K' while the model page specifies 4K for the face region, which is a narrower claim. sync.so/docs/models/lipsync-2 returns 404; lipsync-2 is documented on the shared /docs/models/lipsync page instead.
See it from the vendor
Shotcraft hosts no sample output. Judging a model means watching its own showreel with your own eyes, and a clip re-hosted here would be a clip we chose.
Nearby in the directory
lipsync-2 and lipsync-2-pro
lipsync-2 is Sync's general-purpose lipsync model, preserving each speaker's individual speaking style at 512x512 resolution. lipsync-2-pro adds diffu
Lip sync and avatars
Flawless DeepEditor and TrueSync
Neural performance editing tools for film and episodic post. DeepEditor applies new dialogue to existing shots and refines performances, covering in-v
Lip sync and avatars
Hedra character / avatar platform
Character and avatar video platform that has grown into a unified inference API putting multiple leading image, video and audio models behind one API
Lip sync and avatars