Clone your speaker’s voice into every language

Dub a video ad in the voice of the person who recorded it, with their consent. Or audition stock voices and design a new one where no voice fits.

English · originalI sleep cooler in this linen than in anything I’ve owned.

German · dubbedIn diesem Leinen schlafe ich kühler als in allem, was ich je hatte.

English · originalAnd it gets softer with every wash.

German · dubbedUnd es wird mit jeder Wäsche weicher.

The script, in their voiceThe translated lines are read in the voice you chose for each speaker.
  • Speaker 1Original voiceConsent confirmed
  • Speaker 2Stock voiceAuditioned in German
  • NarratorDesigned voiceWarm, mid-paced, close to the mic
One voice per speakerCast each speaker separately: their own voice, a stock voice or a designed one.
Original speaker voiceKeep the speaker’s own voice, with their consent.
Clone from a recordingBuild a voice from a recording of someone who agreed to it.
Stock and designed voicesAudition stock voices, or describe a new one and create it.
27 languagesEvery official EU language, plus Serbian.

How voice dubbing works

Dub a video ad
  1. 1. Upload your video ad

    Add an MP4, MOV or WebM file of up to 3 minutes. Speakers are told apart for you.

  2. 2. Cast each speaker

    Keep a speaker’s own voice with consent, clone one, or pick a stock or designed voice.

  3. 3. Listen and fix

    Play the dub, add a pronunciation for a name, and generate that language again.

Voice cloning

A voice your audience already knows

Consent comes first

Use a person’s voice only with their permission. The app asks you to confirm consent before it keeps an original speaker’s voice.

Names that sound right

Brand and product names trip up every voice. Add how a name is pronounced and generate the language again.

What is included

The voice options, and what they do not cover.

FeatureDetailsStatus
Original speaker voiceKeep the speaker’s own voice, with their consent.Included
Designed voicesDescribe the voice you want and create it when no stock voice fits.Included
Stock voicesAudition and pick a voice for each language.Included
Translated voiceoverThe spoken message is translated and voiced to fit the original edit.Included
Script reviewEdit phrases, timing and speakers before you generate, and fix words after.Included
Lip-syncOptional: the speaker’s mouth is matched to the new voiceover. Shots with two faces keep the original mouth.Included
Publishing to ad platformsYou upload the files through your own campaign tools.Not included

Start with a supported file

What you can upload and the voices you can use.

More ways to localize an ad

Image ad translation

Translate headlines, on-image copy and calls to action into 47 languages.

Translate image ads

Video ad dubbing

A translated voiceover that fits the original edit, in 27 languages.

Dub video ads

Ad resizer

One image ad in 11 aspect ratios, with or without translation.

Resize an ad

Video subtitles

Burned-in translated subtitles, over the original voice or a dub.

Translate video subtitles

Supported languages

Every language for image and video ads, with a note where one needs a closer look.

See all languages

Questions about this workflow

Do I need permission to use someone’s voice?

Yes. Only use the voice of a person who has agreed to it, and keep a record of that consent. The app asks you to confirm it.

Can the speaker sound like themselves in a language they do not speak?

That is what keeping the original voice does: the translated script is read in a voice built from theirs.

What if the cloned voice does not sound right?

Pick a stock voice for that language after hearing it, or design a new one, and generate that language again.

Can the mouth match the new voice too?

Yes, with optional lip-sync at an extra cost per second. A shot with two faces keeps the mouth as filmed.

Hear your speaker in another language