AI video translationVideo translation that sounds like the original.

Spelo translates the speech in a video and dubs it into 90+ languages. Each speaker keeps their own voice and the way they said it, and the music, effects and room sound stay exactly as they were.

One scene, up to two minutes, in one target language.

Three ways to translate the same documentary line.

Subtitles ask viewers to read instead of watch. A single voice-over flattens every speaker into one reader. A dub gives each speaker their own voice, delivered the way they said it.

Three ways to translate the same documentary line
Original · EnglishWe waited three days for the storm to pass.tired · quiet
SubtitlesEsperamos tres días a que pasara la tormenta.
Single voice-overOne narrator reads it over the original voice, for every speaker.
Spelo dub · SpanishEsperamos tres días a que pasara la tormenta.tired · quiet

What makes video translation hard

Translating the words is the easy part. Keeping the video watchable in the new language is what takes the work.

  • Every speaker needs a voice

    An interview, a panel or a drama has several people talking. One generic voice for all of them makes the video confusing.

  • Tone gets lost

    A tired line, a joke or a quiet threat reads very differently when it is spoken flat.

  • The soundtrack matters

    Music, effects and ambience are part of the video. Replacing the whole audio track throws them away.

  • Lines must fit the moment

    A translated line that runs long talks over the next speaker or the next cut.

How Spelo translates video

  • It listens before it translates

    Spelo hears who is speaking and how each line is delivered, then writes the translation for that scene.

  • A voice for every speaker

    Each person gets their own voice for the whole video, recognised by their voiceprint wherever they appear.

  • Only the speech changes

    Voices are separated from the mix, so the music, effects and ambience stay from the original.

  • Timed to the original

    Each new line is placed where the original was spoken, and your team can adjust timing and speed in the Studio.

How a project runs

Spelo makes the dub and the mix. Your team reviews every line and exports.

Send your film and notes

Share a video file or a link. Add who's-who notes if you have them, so names and relationships are right from the first line.

Choose the target language and your audience

Tell us who will watch it: the region, the variety of the language and how formal the film should sound.

Spelo creates the dub and mix

Every line is written for its scene, voiced with the original emotion and mixed over the film's own sound.

Review and fix lines in Studio

Your team signs off the language, changes any line and re-voices it in seconds.

Export your finished MP4

Download the dubbed video once your team is happy with every line.

Questions

Which languages can Spelo translate video into?

90+ languages, from Spanish, Hindi and Japanese to Swahili and Welsh. The source video can be in any language. See the full list.

Can one video go into several languages?

Yes. Choose as many target languages as you need. Each becomes its own dubbed version, reviewed in the same Studio.

What video files can we send?

MP4, MOV, MKV, WebM or AVI, up to 20 GB per file.

How long does it take?

A short scene is ready in minutes. A feature takes longer, and most of the calendar time is your own team's review.

Hear it on your own film.

Send us one scene, up to two minutes. We'll dub it into the language you choose and send it back next to the original, within a few working days.