Voice matching

Consent-based AI voice dubbing

Dubnivo does not use a stock narrator. It builds a voice from the speech already in your video and uses it to say the translated lines.

Dub a video free →No sign-up needed to try · 5 free videos a month with an account

How it works

  1. 01

    Add your video

    Upload a video file, or paste a public link from TikTok, Instagram, Facebook, X or Twitch.

  2. 02

    Pick the languages

    Choose the language spoken in the video and the language to dub into.

  3. 03

    Download the dub

    Dubnivo transcribes, translates and re-voices the video in natural AI voices, a matching one for each speaker, then gives you an MP4 to download.

No recording session

Each voice is taken from that person's own speech in the video you upload. There is nothing extra to record.

A voice for every speaker

When several people talk, Dubnivo tells them apart and gives each one a separate synthetic voice, so viewers can still follow who is speaking.

Choose the voice style

“Closest voice” keeps more of the speaker's accent, “Most natural” sounds most at home in the new language, and “Balanced” sits between the two.

Used for that video only

The voice model is made for the video you submit and is not kept to train anything or reused for other people. On Premium and Platinum you can also dub in a voice you upload, with that person's consent.

Consent matters

Only dub people who have given their consent: you confirm it before every dub. Dubnivo's terms forbid impersonating anyone or misleading people about who said something.

AI video dubbing and video translator features

Everything Dubnivo does today, named plainly.

  • Natural AI voices.

    Each person in the video gets their own matching AI voice. Copying a speaker's own voice is an option, with their consent.

  • Multi-speaker voices.

    When several people talk, each one is recognised and given a separate synthetic voice, so viewers can follow who is speaking.

  • 17 dubbing languages.

    Dub into English, Spanish, Hindi, Arabic, Chinese, Japanese, Korean, French, German, Portuguese and 7 more, from 29 spoken languages.

  • Review and edit the translation.

    Check the translated lines before anything is spoken, or correct a finished dub and dub it again.

  • Voice style.

    Choose the closest voice, a balanced delivery or the most natural sound in the new language.

  • Keep or drop the background.

    Get the new voice alone, or keep the video's music and background sound under it.

  • Subtitles in SRT and VTT.

    Download subtitles in the original and the translated language when the dub is done.

  • Short clips with captions.

    Cut a long video into up to 6 short captioned clips, vertical 9:16 for Shorts, Reels and TikTok.

  • Import from a link or Google Drive.

    Paste a public TikTok, Instagram, Facebook, X or Twitch link, upload a file, or pick one from Google Drive.

  • Publish to YouTube.

    On paid plans, send the finished dub straight to your YouTube channel.

  • Podcasts and audio files.

    Translate MP3, M4A or WAV recordings and get the translated audio back, on Premium, Platinum and pay as you go.

  • Free plan.

    Try 1 video without an account, then dub 5 videos a month free, up to 2 minutes each, with no card.

Questions

How does AI voice matching work in a dub?

By default each speaker gets a natural AI voice close to theirs, from voices people donated for this. With “Match the speaker's own voice” turned on, a speech model analyses a short sample of the speaker and generates new speech with a similar tone and timbre.

How much audio is needed to match the voice?

A few seconds of clear speech at the start of the video. Speech without music or other voices gives the closest match.

Is my voice stored?

No separate voice profile is kept. The uploaded video and its dub are deleted automatically after 3 to 30 days depending on your plan.