peters.lab· project
live · since 2026

AI Video Translator

Any video, speaking 20+ languages — dubbed in the speaker's own voice.

AI Video Translator screenshot

overview

AI Video Translator is a full localization pipeline, not a subtitle button. From a single upload it recognizes the speech, translates for meaning rather than word-for-word, re-voices the result with natural AI dubbing that keeps the original speaker's voice, and lays down accurate subtitles and a transcript — across 20+ languages. Built for creators, educators, and marketers who need one video to reach everyone, in minutes instead of a dubbing studio's weeks.

what it does

  • Full pipeline: speech recognition → meaning-based translation → dubbing → subtitles
  • Natural dubbing that keeps the speaker's own voice, timing in sync
  • Subtitles and a transcript in the same pass, 20+ languages
  • AI dubbing
  • subtitles
  • 20+ languages

why i built it

I built it because "translating a video" almost always means burning in subtitles and calling it done — which leaves the audio, the voice, and the pacing stuck in the original language. That's subtitle translation, not video translation. I wanted the whole thing to travel: same speaker, same delivery, a different language. So the tool runs the full pipeline end to end instead of stopping at the text layer.

from the blog

Try AI Video Translator

ai-video-translator.com