ToolGPT

Remove vocals from a song and keep the music

For karaoke, a practice track or a cover, you often want a song without the singer. Attach the song or a video to ToolGPT and ask it to remove the vocals. An AI separation service on a server splits the voice from the music, and the instrumental comes back to the chat.

  • Audio processed in the cloud, with your permission
  • No sign-in needed
Drop or choose your filesThe assistant opens with your files and does the rest.Choose Files
Things you can type
summer-song.mp3
Open ToolGPTAttach a file and describe what you need.

See a step-by-step example →

What it can do

  • The instrumental in one request

    Ask for the vocals to be removed, and the result is the music on its own.

  • Both parts from one upload

    The service always returns the voice and the music together. A switch on the result card moves between “Voice removed” and “Music removed”.

  • Songs and videos

    Use an MP3, WAV, M4A or OGG file, or a video with sound.

  • The picture stays with you

    For a video, only the audio is sent. The new soundtrack is put back under your picture in your browser.

  • Upload only with your permission

    A permission card explains what will be sent. The upload waits until you choose.

How it works

  1. 1

    Attach the song or video

    Open ToolGPT and add the file with the + button. You can do this without an account.

  2. 2

    Ask, then allow the upload

    Say “remove the vocals”. A permission card explains that the audio goes to a cloud service. Choose Allow once or Allow for this tab.

  3. 3

    Download the instrumental

    When processing finishes, the instrumental appears in the chat. Download it, or switch the card to the vocals.

How the vocal remover works

ToolGPT first makes an MP3 copy of the file's sound in your browser. That copy goes to ToolGPT's separation service, which splits it into two parts: the voice, and everything else.

You get the instrumental as the main result. The voice part is kept with it, so the switch on the result card changes sides on your device, without a second upload.

For an audio file, the result is an MP3. For a video, the instrumental replaces the original soundtrack, and you get the video back without the singing.

What is uploaded, and when

Unlike most ToolGPT tools, this one needs a server. Only the sound is sent. When the file is a video, the audio is taken out in your browser first, and the picture stays on your device.

Before the first upload, a permission card names the file and explains what will be sent. Allow once covers this job. Allow for this tab covers later cloud jobs in the same browser tab. The assistant cannot press either button for you.

Limits to know

Guests can use the vocal remover within a daily limit on how much audio they process. The vocal remover, the acapella maker and the noise reducer all count toward that same limit. When you are signed in, jobs use your account's allowance.

How clean the split sounds depends on the recording. A file with no sound track cannot be processed, and ToolGPT says so before it asks for permission.

Frequently asked questions

Is my song uploaded?

Its audio is, after you allow it, because the separation runs on a cloud service. For a video, only the extracted sound is sent.

Can I get the vocals as well?

Yes. Both parts come back from the same upload. Use the switch on the result card to change to the vocals, and download them too.

Do I need an account?

No. Guests can remove vocals up to a daily limit. Signed-in users use their own account's allowance.

Can I remove the vocals from a video?

Yes. The sound is separated, and the instrumental is put back under the original picture, so you get a video without the singing.

Try it with your own file

Open ToolGPT, attach a file, and type what you want done.

Open ToolGPT

Related tools

See all tools