How accurate is acapella extraction?

It varies by song — the sparser the backing, the cleaner the result. When quality matters more than time, try High Quality.

What can I use the extracted acapella for?

Remixes, practicing covers, checking your ear-transcription, and more.

Why does this page open on High Quality?

Because an acapella gets layered over other things. Rough edges hide behind a full arrangement when you are only listening to a backing track; a voice on its own, or over a different instrumental, puts every bit of bleed in the open. Quality is the first thing that shows here. Switch to Standard if you are in a hurry.

Can I separate the lead from the harmonies?

No. The model treats human voice as one part, so lead, harmonies and backing vocals all land in the same track. Splitting them apart is beyond what this tool reaches.

How do I get rid of the backing that is still audible?

Not with anything on this page. EQ is a bad answer — it cuts the vocal band before it cuts the bleed. What actually helps is re-running on High Quality, and choosing the sparser sections of the song. The denser the backing at that moment, the more of it survives.

Is it really free? Do I need to sign up?

It’s completely free with no account needed. Just open the page and pick a file.

Is the song I choose sent anywhere?

No. All separation happens entirely inside your browser.

Which file formats are supported?

MP3, WAV, M4A, AAC, FLAC and OGG, plus the audio track of video files.

Audio · processed on this device

Pull out just the vocals.

Extract the vocal part from a song to create an acapella track.

Free, no account
Stays on your device

About this process

Free in your browser · instant

One mix, separated into its individual parts

Acapella extraction keeps the voice and discards everything else — the mirror image of vocal removal. It is how people source material for mashups and remixes, or study a performance phrase by phrase.

When the voice is the part you keep, whatever backing survives becomes noticeable. Cymbals and reverb tails overlap the vocal range most, so a little can bleed through. If the acapella is going into a production, export at High Quality and clean up the unwanted bands afterwards.

There is a reason cymbals in particular get through. A cymbal has no definite pitch and spreads energy across a wide band — much like the consonants at the front of sung words, the breath in an s or a t. The model is deciding between those on probability, so whatever could be either lands on the vocal side. The trade runs the other way too: consonants sent to the backing leave the extracted voice sounding slightly muffled.

Whatever reverb was on the voice in the original comes with it. That is not a failure of the separation — the reverb was recorded as part of the vocal. If you plan to place the acapella over a different backing, that inherited space can fight the new one. There is no de-reverb here, so the practical answers are to use the drier sections or to find an acapella that was released as one.

Three simple steps

Open a file
Choose an MP3, WAV, M4A, FLAC or video with “Open file”, or drag it into the workbench.
Check settings and separate
This page opens on High Quality — an acapella is going to be layered into something else, which is where the difference shows first. Standard pulls it out too if you are in a hurry. On a phone, Fast is the one that runs. Check the settings selected for this task, then press “Separate”.
Listen and export
Use Solo and Mute to listen to the parts you need, then choose “Export” to save WAV or MP3.

Acapella extractor — frequently asked questions

How accurate is acapella extraction?: It varies by song — the sparser the backing, the cleaner the result. When quality matters more than time, try High Quality.
What can I use the extracted acapella for?: Remixes, practicing covers, checking your ear-transcription, and more.
Why does this page open on High Quality?: Because an acapella gets layered over other things. Rough edges hide behind a full arrangement when you are only listening to a backing track; a voice on its own, or over a different instrumental, puts every bit of bleed in the open. Quality is the first thing that shows here. Switch to Standard if you are in a hurry.
Can I separate the lead from the harmonies?: No. The model treats human voice as one part, so lead, harmonies and backing vocals all land in the same track. Splitting them apart is beyond what this tool reaches.
How do I get rid of the backing that is still audible?: Not with anything on this page. EQ is a bad answer — it cuts the vocal band before it cuts the bleed. What actually helps is re-running on High Quality, and choosing the sparser sections of the song. The denser the backing at that moment, the more of it survives.
Is it really free? Do I need to sign up?: It’s completely free with no account needed. Just open the page and pick a file.
Is the song I choose sent anywhere?: No. All separation happens entirely inside your browser.
Which file formats are supported?: MP3, WAV, M4A, AAC, FLAC and OGG, plus the audio track of video files.

Guides for this task

Using stems for sampling and remixing
Lift a drum track to build on, or take a vocal phrase somewhere new. Three ways to put stem separation to work in a DAW, and three tricks for making the separation artefacts sit down.
Making your own backing track for a cover
No official off-vocal for the song you want to sing? Stem separation will make you one. The three-step process, plus two tricks for cleaning up whatever vocal is left behind.