AI video dubbing

Upload a video, check the translation line by line, and download it dubbed into another language, with the original music and background sound kept.

También en español

Cloud processing

MP4, MOV, WebM and other video up to 100 MB and 5 minutes.

Voice sample

0:00--:--

Uploaded and processed audio is stored on SoundHalo. Audio access expires after 30 days; expiry does not confirm storage deletion. Delete the stored files here before leaving this page. Speech providers have their own retention policies. Privacy policy.

Example: English to Spanish

A 13-second clip with background music, dubbed into Latin American Spanish with the Dora voice. The output kept the original music and background sound: the English voice was removed and the Spanish dub was laid over what was left. The first Spanish line was longer than the English one and was sped up to fit its time. Nothing was edited.

TimeOriginalDub lineLine check
0:01.0 → 0:03.7Welcome to a short tour of our community garden.Bienvenidos a un breve recorrido por nuestro jardín comunitario.Sped up 1.25x
0:03.7 → 0:07.4Every spring, neighbors plant tomatoes, beans, and herbs here.Cada primavera, los vecinos plantan tomates, frijoles y hierbas aquí.Sped up 1.05x
0:07.4 → 0:13.0The garden is open to everyone on Saturday mornings, so come and say hello.El jardín está abierto a todos los sábados por la mañana, así que ven a saludar.Fits

Original video

Dubbed video, original music kept

What you check before anything is spoken

  • Every line of the translation sits next to the original speech and its time, so you can read them side by side.
  • Lines that look too long for their time, even sped up, are marked before you use any speech characters.
  • You can edit a line, translate one line again, or clear it to skip it. The voice is made only when you choose Create dubbed video.
  • After the dub is made, the line check shows which lines fit, which were sped up and which still run over. Watch the dubbed video on the page before you download it.

How the video sounds

  • Dub over the original music and sound: the original voice is removed and the dub is laid over the music and sound effects that are left, which dip a few decibels while the dub speaks. This takes about half a minute per minute of video.
  • Dub over the original sound, lowered: the whole original sound, including the original voice, drops by about 15 dB while the dub speaks and comes back between lines. Quicker, but the original voice stays faintly audible.
  • Dub only: the video keeps only the new voice. Music and sound effects are removed too.

Speech removal uses the UVR-MDX-NET 9482 vocal model from Ultimate Vocal Remover (UVR), released under the MIT license. Thanks to the UVR developers for training and sharing it. It works best on clear speech over music; crowd noise, singing and speech that overlaps loud effects can leave traces of the original voice.

Dub in the original speakers' voices

With a free account, choose "Original speakers' voices (cloned)" as the voice. Transcription then adds speaker labels (up to four people), and each speaker's clearest stretches of speech, up to about 15 seconds, become a reference for their voice. Before anything is generated, the review shows which voices will be copied and the exact cost, and you confirm you have the right to dub those people in their own voices. Music or noise in those reference seconds would be copied into the voice, so when a speaker's seconds have background sound, the speech separation model removes it first and the voice is copied from their speech alone. In our tests this left the pauses in the cloned speech 10 to 50 dB quieter, with a small loss in likeness. Traces can remain where the music is loud, so videos with clear speech still work best. Cloned speech is processed by Fish Audio and charged at its real cost; the full rules are on the translate audio page.

Match lip movements

After the dub is made, signed-in users can choose Match lip movements. The speaker's mouth in the original video is edited to follow the dub voice, then the finished dubbed sound is put back on the video. Paid plans include Standard lip-sync minutes each month (Pro 5, Creator 15, Business 35); beyond them, and for the sharper Pro and Studio qualities, it is paid from your prepaid balance. The page shows the minutes and balance it uses before you confirm. It is processed by fal; read the lip-sync page for qualities, prices and the consent rules.

How the timing works

The original speech is transcribed with the time of every word. The transcript is cut into lines at the end of each sentence, at pauses of a second or more, and before a line passes about 15 seconds. Each translated line starts at the time the original line started, and may use up to 2 seconds of the pause after it. A line that is longer than the original is sped up without changing the pitch, up to 1.25 times. Beyond that, speech sounds rushed, so the line is flagged instead and you can shorten it.

Limits

  • No account: videos up to 5 minutes and 100 MB, and a dub preview of up to 300 speech characters (about 20 seconds of speech) and 20 lines. Longer scripts dub the first lines that fit. Speech characters (1,200 a day), translations (5 a day), transcription and video processing (60 minutes a day) are shared by everyone on your network and capped by a site-wide free budget, so they can run out.
  • Free account: videos up to 10 minutes and 100 MB, up to 3,000 speech characters (about 3 minutes of speech) and 200 lines per dub, 20,000 speech characters and 90 minutes of video processing a day.
  • Pro: videos up to 10 minutes and 100 MB, up to 20,000 speech characters and 500 lines per dub, 300,000 speech characters and 1,800 minutes of video processing a day.
  • Keeping the music and background sound counts the video length three times toward video processing (removing the speech, then two mixes). The lowered original and dub only choices count it once. Speech characters are shared with the text to speech, voice-over and translate audio tools. Limits reset at midnight UTC.

Synthetic voice and acceptable use

The dub is read by a text to speech voice, or by a cloned copy of each speaker's voice when you choose that option. Either way it is synthetic speech, not a recording of what anyone said. Dub videos you made or have the rights to translate and publish, such as your own tutorials, courses or product videos. Do not use it to make someone appear to say something they did not, to impersonate a person, or for spam and fraud. Tell your viewers the dub is AI generated; some platforms require that label.

Related tools

FAQ

Does it keep the original music and sound effects? +

Yes, if you choose "Dub over the original music and sound". A speech separation model removes the original voice and keeps the music and background sound, and the dub is laid over them. Separation is not perfect: faint traces of the original voice can remain, mostly where speech and music overlap. You can also keep the whole original sound lowered under the dub, or replace it with the dub alone.

Will the dub sound like the original speaker? +

By default the dub uses a stock synthetic voice you pick, such as Dora or Alex for Spanish, and each voice has a sample to play first. Signed in, you can instead choose original speakers' voices: each person in the video is cloned from a few seconds of their own speech and speaks the translation. That option is charged at the voice provider's real cost and needs your confirmation that you may dub those people.

Does it match lip movements? +

Not by default: the picture is not changed, and each dub line starts when the original line started, so speech lines up with the scene but mouths do not match the new words. Signed in, you can choose Match lip movements under the finished dub. It edits the speaker's mouth to match the dub, using the Standard lip-sync minutes your plan includes or your prepaid balance; see the lip-sync page for how it works and what it costs.

Which languages can I dub into? +

25 languages: English (US), English (UK), Spanish, French, German, Italian, Portuguese (Brazil), Portuguese (Portugal), Japanese, Chinese (Mandarin), Hindi, Russian, Turkish, Polish, Dutch, Swedish, Danish, Norwegian, Finnish, Romanian, Hungarian, Czech, Indonesian, Vietnamese, Arabic. The original video can be in any language the speech recognition supports; it is detected automatically.

Can I correct the translation before the voice is made? +

Yes. After transcription and translation you see every line next to the original speech, with its time. Edit any line, ask for a new translation of one line, or clear a line to skip it. Speech characters are only used when you choose Create dubbed video.

What happens when the translation is longer than the original? +

Each dub line starts at the original line's start time. If the speech is longer than the original line, it is sped up, up to 1.25 times, without changing the pitch. If it still does not fit, it plays in full and the line check shows how far it runs over, so you can shorten it and dub again. Words are never cut.

What files do I get? +

The dubbed video as MP4, the dub as a voice-only track (MP3 or WAV) and the translated script as an SRT subtitle file with the original timings. When you keep the music and background sound, you also get the finished dubbed audio track on its own. The picture is copied as it is when MP4 can hold it, so its quality does not change; other video formats, such as VP8 from WebM, are converted to H.264.

Which video files work? +

MP4, MOV, WebM, MKV and most other video files with an audio track, up to 100 MB. The page checks the length before uploading.

Where is my video processed? +

The video is uploaded to SoundHalo. Its speech is transcribed by Lemonfox. The transcript is translated by the DeepSeek model through OpenRouter, using only hosts that do not store or train on prompts. Each translated line is read by a text to speech provider: Lemonfox for its languages and Giggy for the others, or Fish Audio when you choose the original speakers' voices. Removing the original speech and mixing the video run on SoundHalo's own processing servers on Cloudflare, and no other company receives the video for those steps. The video and the generated files are stored on SoundHalo so you can download them. Access expires after 30 days, and you can delete the files from the page. When you are signed in, the transcript is also saved to your account library, where you can delete it; the translation is not kept.

What are the limits? +

No account: videos up to 5 minutes and 100 MB, and a dub preview of up to 300 speech characters (about 20 seconds of speech) and 20 lines. Longer scripts dub the first lines that fit. Speech characters (1,200 a day), translations (5 a day), transcription and video processing (60 minutes a day) are shared by everyone on your network and capped by a site-wide free budget, so they can run out. Free account: videos up to 10 minutes and 100 MB, up to 3,000 speech characters (about 3 minutes of speech) and 200 lines per dub, 20,000 speech characters and 90 minutes of video processing a day. Pro: videos up to 10 minutes and 100 MB, up to 20,000 speech characters and 500 lines per dub, 300,000 speech characters and 1,800 minutes of video processing a day. Keeping the music and background sound counts the video length three times toward video processing (removing the speech, then two mixes). The lowered original and dub only choices count it once. Speech characters are shared with the text to speech, voice-over and translate audio tools. Limits reset at midnight UTC.