Skip to main content
voice-remover
Sign in
Checking your runs

AI Voice Remover

Separate the speaking voice from the music behind it. Upload a recording, listen to the isolated voice and the remaining audio, then save the track you need.

READY

You get the voice as one WAV, and the music as another.

Up to 25 minutes and 500 MB per recording. WAV, AIFF, FLAC, MP3, M4A, OGG, MP4, MOV and MKV.

Checking your runs

Your recording stays on your device until you confirm. Our processing partner deletes the audio within 24 hours. The separated tracks stay available for repeat runs for 24 hours.

Isolate the voice track

Start with the original mix of speech and background music. The voice remover gives you a Voice track and a Music track you can audition at the same point in the recording. Download the one you need as a WAV file.

Check the first words after a pause and the quiet end of a sentence. These spots make it easier to hear whether a word was clipped or whether the music is still audible under the voice.

If the soundtrack contains dialogue, ambience and sound effects, use the dialogue remover.

Take the voice out of a video

Keep the interview or narration from a clip with a soundtrack underneath. Open the video tool, choose Voice only, and compare a moment where the speaker talks over the music.

Save the isolated voice as WAV if you plan to edit it on a timeline. For a quick replacement, the video tool can save an MP4 with the new soundtrack in supported browsers. Keep a copy of the source clip so you can compare the edit against the original timing.

Remove background music from video

Voice for podcasts, interviews and dubbing

Podcast excerpts

Pull out a spoken segment that already has an intro bed underneath it. Listen where the bed fades into the first sentence, then add a new bed underneath the isolated speech in your edit.

Interviews and field recordings

Keep the interview track if the main distraction is music. For wind, traffic or a fan behind a voice, use the background noise remover.

Dubbing and narration

Use the voice track as a timing reference for a new read. Save the original and the separated background too, so you can line up the replacement and hear where a sound effect or room tone needs to be rebuilt.

Cost, privacy and time for voice separation

Which audio tracks can I save?

Preview the isolated Voice and the remaining Music track, then download the one you need as a 24-bit WAV. Keep the original if you want a reference for a later edit.

Can I use a long interview recording?

The upload accepts up to 25 minutes and 500 MB. If the recording is longer, trim a copy to the section you need before separating it. WAV, AIFF, FLAC, MP3, M4A and OGG are supported.

Can I keep the dialogue and put it back into an MP4?

The dedicated video tool can isolate the voice and export it as WAV. In a supported browser it can also put the replacement audio into an MP4.

Which tool should I use for hiss, wind or echo?

Use voice separation when the recording has music behind the speech. For hiss, fan noise or wind, choose a background noise tool. For a long room tail after each word, try reverb removal.

Can I separate two people who are talking at once?

The voice track can contain both speakers. If their voices overlap in the same recording, use separate microphone tracks when you have them, or edit the short overlaps by hand.

How many free runs are available without an account?

You get 3 free cloud runs each day without signing in. The upload panel shows the runs remaining before you start.

When is a recording uploaded?

The file stays on your device while you check it in the confirmation panel. Once you confirm, your audio goes to the cloud separation service. The privacy policy explains how long the files are kept.

How long will the separation take?

The time varies with the recording length and whether the service needs to start a worker. The progress panel shows each stage. Try a short excerpt first if you only need to check how the voice sounds.

Every audio tool, grouped

47 tools available now. Find all 49 listings in the tools directory.

Have another recording to separate?

It goes straight into the AI Voice Remover.