Translate video into 40+ languages
AI video translation that goes beyond subtitles: TH-Labs translates what is said and speaks it back in the original speaker's voice, with lip sync and captions included.
More than subtitles
Subtitles ask the viewer to read instead of watch. A translated voice track lets them simply listen. TH-Labs produces both from one upload: a dubbed audio track in the speaker's cloned voice, and subtitle files in every target language.
Live and recorded
Recorded video is translated faster than real time. Live streams — lectures, conferences, broadcasts — are translated and voiced continuously at about two seconds of delay, so a multilingual audience stays in step with the room.
Developers can translate video programmatically through the TH-Labs API.
How video translation works
- 01Speech recognition transcribes the original audio.
- 02Machine translation carries the meaning into each target language.
- 03A clone of the speaker's voice reads the translation, timed to the picture.
- 04Lip sync and subtitles are generated from the same run.
Frequently asked questions
- Which languages can I translate a video into?
- Over forty, in both directions. If you need a specific pair, ask and we'll tell you exactly where it stands.
- Can I translate a live stream?
- Yes. Live translation runs continuously at roughly two seconds of end-to-end delay.
- Do I also get subtitles?
- Yes. Subtitles and captions are exported for every target language.