Dubbing
Translate video or audio to another language.
AI Video Dubbing: Translate Any Video or Audio with ElevenLabs
Drop in a video or an audio file, pick a language, and ElevenLabs Dubbing rebuilds the speech in that language with the original voices preserved. Tone, pacing, and speaker identity carry over, so a dubbed clip still sounds like the person who recorded it.
It handles both media types from the same slot. Give it a video and you get the video back with a translated soundtrack; give it audio and you get a dubbed audio track.
What can you do with it?
- Localize video content. Ship the same product demo, course, or ad in Spanish, Japanese, and German without re-recording.
- Translate podcasts and voiceovers. Audio in, dubbed audio out, same voices.
- Reach a new audience with old content. Your back catalog already exists; dubbing it is one run per language.
- Dub multi-speaker recordings. Interviews and panels work too. Speaker detection is automatic, and you can pin the count if it guesses wrong.
- Watch foreign material in your language. Dub into English just as easily as out of it.
How to use it
- Upload your file. One slot takes either an audio or a video file.
- Pick the target language. 28 languages, from Spanish to Tamil.
- Run it. Source language and speaker count are detected automatically. You get back a dubbed MP4 for video input, or a dubbed audio file for audio input.
Settings
- Dub Into: the language the speech gets translated and re-voiced in.
- Source Language (hidden settings): set it explicitly if auto-detection picks wrong, e.g. for heavy accents or mixed-language clips.
- Number of Speakers (hidden settings): 1 to 50. Leave empty and ElevenLabs counts them itself.
- Highest Resolution (hidden settings): on by default; keeps the dubbed video at the best available resolution.
Pricing
$0.90 per minute of input, rounded up to the next whole minute. A 90-second clip bills as 2 minutes ($1.80); a 12-minute episode runs $10.80. Audio and video cost the same.
Tips
- Feed it clean speech. Clear dialogue with little background music translates best; heavy soundtracks can muddy the voice separation.
- Set the speaker count for crowded audio. Overlapping voices in a panel recording are the main place auto-detection slips.
- Mind the minute rounding. A 61-second clip costs the same as a 2-minute one, so trim dead air before dubbing.
- Dub the master, then edit. Dub your full-length source once and cut clips from it, instead of dubbing each clip separately.