Voice & language

Your voice.Their language.

Voice-matched dubbing in 29 languages — cloned from thirty seconds of your audio, paced to your delivery, with phoneme-level lip-sync on the Pro tier.

29
dubbing languages
30 s
of audio to match your voice
2
lip-sync tiers: Standard · Pro
dubbing_agent
source_track.wavprocessed
Levels
−14 LUFS
Bed
auto-ducked
Noise floor
−58 dB

It sounds like you, because it is you

Generic TTS dubbing makes every creator sound like the same narrator. The dubbing agent clones from your actual audio — thirty seconds is enough — and carries your emphasis, pauses, and energy into every target language.

  • Voice cloned from 30 seconds of clean speech
  • Emotion, emphasis, and pacing preserved from the original
  • Multi-speaker diarization — every speaker keeps their own voice
  • Pronunciation overrides for names and product terms
dubbing_agent
Voice cloned from 30 seconds of clean speech
Emotion, emphasis, and pacing preserved from the original
Multi-speaker diarization — every speaker keeps their own voice

Two lip-sync tiers, one honest tradeoff

Standard times the dub to your existing cut — fast, and right for voiceover-heavy content. Pro goes further: the mouth region is re-rendered frame by frame to match the dubbed phonemes, for talking-head video where lips are the show.

  • Standard: wave-aligned dub, ready in minutes
  • Pro: phoneme-level mouth re-rendering on GPU
  • Pro renders take roughly 2× the video's runtime
  • Per-scene tier mixing — Pro for face shots, Standard elsewhere
dubbing_agent
Standard: wave-aligned dub, ready in minutes
Pro: phoneme-level mouth re-rendering on GPU
Pro renders take roughly 2× the video's runtime

Built for courses and product video

Dubbing earns its keep on content with a long shelf life. Re-record a single segment when a price changes, keep music and SFX stems untouched under the new voice, and export every language from one project.

  • Per-segment re-recording without redoing the whole dub
  • Music and SFX stems preserved under the dubbed voice
  • Shared glossary with AI Video Translator
  • One export per language, named and organized
dubbing_agent
Per-segment re-recording without redoing the whole dub
Music and SFX stems preserved under the dubbed voice
Shared glossary with AI Video Translator

From input to finished render

  1. 1

    Input

    Tell the agent what you're starting from — a recording, a script, or a prompt.

  2. 2

    Format

    Pick the output shape: aspect ratio, file type, or platform target.

  3. 3

    Length

    Set duration or let the agent find the natural cut points for you.

  4. 4

    Features

    Layer in captions, music, b-roll, or any other agent in the same pass.

  5. 5

    Model

    The model router picks the right model for the job — quality where it matters, speed everywhere else.

Frequently asked

Ready to try Dubbing?

No credit card. 10 free AI minutes every month. The dubbing agent is ready when you are.