Extract the Voice from a Video (Isolate Dialogue)

Get only the voice out of a video, without the music and effects behind it: as an MP3, or as the video itself. AI does it in your browser; nothing is uploaded.

Your file never leaves your device. How?

Your media is processed in your browser and is not uploaded to our servers for conversion. Page assets (including the converter engine) still load over the network.

Select or drop a video

MP4, MOV, WebM or a sound file. The music is removed on your device: nothing is uploaded.

Click, tap, or drop a file

How to extract the voice from a video

Choose the video

Drop an MP4, MOV, WebM or MKV video from your device.

Let the model download once

The first time, about 43 MB downloads and is saved in your browser for next time.

Separate the voice

Tap Remove the music. The AI splits the sound into the voice and everything else.

Save the voice

Download "Voice only (MP3)" for an editor, a dub or a transcript, or the video with only the voice in it.

What a clean voice track is for

Dubbing and voice-overs start from the original speech without the music. Transcripts and subtitles are more accurate on a clean voice. Editors want dialogue on its own track, to re-mix it with new music or level it separately; podcasters pull a guest's answer out of a video with music behind it.

Speech comes out cleanest. Singing counts as voice too, so for a song you get the vocals (for songs, the Vocal Remover is made for that).

How it works

An MDX-Net model (from the open-source Ultimate Vocal Remover project) runs in your browser and splits the video's sound into the voice and everything else. The voice is put back in the video with FFmpeg: the picture is copied as it is, so it loses no quality.

Speech over music separates cleanly. Singing counts as voice, and sound effects go with the music. A voice buried in very loud music can keep a little of it.

Features

Voice as MP3

A separate voice track for any editor, dub or transcript.

Background too

The music and effects as their own MP3, if you need them.

Private

Interviews and unreleased footage never leave your device.

Music Mixer (Pro)

Turn the music down instead of off, let it dip whenever someone speaks, or put your own track under the voice, heard live as the video plays. Your first mix is free.

Frequently asked questions

Is my video uploaded?

No. Only the AI model is downloaded, once, from this site. The voice is separated from the music in your browser and the new video is made on your device.

Is the voice clean enough for transcription?

Usually, yes: with the music gone, speech is easier to transcribe. Use Auto Subtitles on the result for subtitles or a transcript.

Does it remove background noise too?

It removes most steady background sound along with the music, but hiss, wind or hum close to the voice can remain. Run the voice through the Noise Remover afterwards for that.

How is this different from the Vocal Remover?

Same AI model. The Vocal Remover is made for songs (karaoke and acapellas, as audio); this is made for videos: it puts the separated voice back into the video, and the Music Mixer works with the picture.

Which videos work best?

Speech over music: vlogs, interviews, talks, tutorials. Videos up to 15 minutes on a computer or 6 minutes on a phone.