​

Models, speed, and limits

Each lip sync tier has a different tradeoff between speed, quality, and project size.

​

Model differences at a glance

​

What each model is better at

​

Lip Sync

Lip Sync is the most practical default for many projects. It is best when you want:

• a fast turnaround
• good quality on normal talking-head footage
• a strong quality-to-cost balance

For straightforward videos, this is often enough.

​

Lip Sync Pro

Lip Sync Pro focuses more on facial detail than standard lip sync. Its main advantages are:

• stronger beard rendering
• better teeth generation
• cleaner mouth-region detail
• better handling of larger face regions in frame

This is why Pro is often a better fit for:

• tighter close-ups
• higher-end creator content
• premium marketing videos
• projects where facial detail matters more than speed

The tradeoff is that it is slower and more expensive than standard Lip Sync.

​

Lip Sync Studio

Lip Sync Studio is the strongest model when the shot itself is difficult or when the output standard is especially high. Its biggest differences are:

• it builds a fuller understanding of the whole shot instead of relying on smaller independent chunks
• it handles obstructions natively instead of depending on extra settings
• it is much stronger on profile views, over-the-shoulder framing, and non-frontal lip positions
• it preserves more of the speaker’s cadence, emotion, and performance
• it is built for higher-end output, including stronger 4K positioning

This is why Lip Sync Studio is better than Lip Sync Pro on:

• hard camera angles
• shots with partial face coverage
• scenes where consistency matters across a longer uninterrupted take
• premium short-form work where visual artifacts are less acceptable

​

Why some videos take longer than others

Two videos with the same duration can still process very differently. Common reasons:

• one video has many scene cuts while another is mostly one stable shot
• one video has multiple faces on screen
• one video has more profile views or extreme angles
• one video has more partial obstructions over the mouth or face
• one video has fewer frames where the speaker is clearly visible and actively talking
• one video is higher resolution
• queue load is different at the time of processing

For Lip Sync and Lip Sync Pro, videos with many scene changes or many frames without clear talking faces can become much harder to process efficiently. That can increase runtime and make failures more likely on difficult long-form content. Lip Sync Studio is usually more resilient on these harder shots, but it is still the slowest tier because the model is doing more work per shot.

​

Why some clips fail or look weak

Quality can drop when the source contains:

• still or nearly still faces that do not already look like they are speaking
• many fast cuts
• faces that are too small in frame
• strong profile shots on lower tiers
• heavy obstructions
• non-human or non-humanoid faces

As a rule:

• Lip Sync works well for the majority of standard videos
• Lip Sync Pro improves premium facial detail
• Lip Sync Studio is the best choice when the footage is difficult, high-value, or visually demanding

​

Typical performance

These timings are estimates, not guarantees. Actual runtime depends on queue load, video duration, resolution, and source complexity.

​

Current duration limits

​

What affects speed

• video duration
• resolution
• current queue depth
• face angle complexity
• number of scene changes
• number of visible faces
• how often the speaking face is obstructed
• whether the speaker is visibly talking throughout the shot
• audio clarity

As a rule of thumb, 1080p is usually the best balance between speed and quality. 4K can work, but it increases processing time.

​

Current billing impact

These are the current extra billed minutes on top of the translated video minutes:

• Lip Sync: +4
• Lip Sync Pro: +9
• Lip Sync Studio: +14

​

Related pages

• Lip sync overview
• Best practices