Getting started
Models
Getting started
Models
Explore and compare the capabilities of our different models
At sync, we’re building foundational models to understand and manipulate humans in video. Our suite of lipsyncing models allows you to edit the lip movements of any speaker in any video to match a target audio. Explore and compare the capabilities of the different models below.
| Feature | lipsync-2 | lipsync-1.9.0-beta | lipsync-1.8.0 | lipsync-1.7.1 |
|---|---|---|---|---|
| Description | Our most natural lipsyncing model yet. The first model that can preserve the unique speaking style of every speaker. Best across all kinds of video content. | Our fastest lipsyncing model. Standard, general-purpose, accurate lipsync. | Slow, legacy model, suited for budget-constrained tasks. Use lipsync-1.9 & later for best results. | Fast, legacy model, best suited for simple low-res avatar videos. |
| Price / min @ 25 fps | $2.4 — $3 | $1.2 — $1.5 | $1 — $0.8 | $1 — $0.8 |
| Accuracy | ||||
| Speed | ||||
| Style | Lip movements in the unique style of the speaker | Standard generic lip movements | Standard generic lip movements | Standard generic lip movements |
| Identity Preservation | ||||
| Teeth | ||||
| Face Detection | ||||
| Face Blending | ||||
| Pose Robustness | ||||
| Beard | ||||
| Face Resolution | 512×512 | 512×512 | 512×512 | 256×256 |
| Best for | Best across all kinds of videos, outperforms every other lipsync model across all key attributes. | Simpler avatar-style use cases | Legacy model, use 1.9.0 and above | Legacy model, might work for low-quality videos |
All models are available in both Playground and API.