Want an instrumental for karaoke, or a voice without the backing track? Attach a song or a video to ToolGPT and ask. An AI separation service splits the voice from the music, and you can switch between the two results without processing the file again.
An MP3, WAV, M4A or OGG works, and so does a video with sound. Open ToolGPT and add it with the + button.
Pick "keep the music" or "keep the voice". Before anything leaves your device, ToolGPT asks your permission to send the audio to the separation service.
Download the result. A toggle on the result card switches to the other side — vocals or instrumental — with no second upload.
The separation service always computes both parts of the track: the voice and everything else. ToolGPT shows you the one you asked for and keeps the other on the result card. If you meant the opposite — you wanted the singing, not the karaoke version — flip the toggle and it switches straight away, without sending the file again.
For a video, the chosen audio is put back under the original picture, so you get a video with the vocals (or the music) removed.
Unlike most ToolGPT jobs, this one uses a server, and ToolGPT is explicit about it. Voice and music separation runs on InverseAI's cloud service, so the file's audio is uploaded for processing. For a video, the audio is extracted in your browser first and only the sound is sent — the picture never leaves your device.
No upload starts until you agree. The first time you ask, a permission card explains what will be sent and waits for you to choose; the assistant cannot click it for you.
The same service can remove background noise from speech — traffic, hum, a noisy room — while keeping the voice. Ask ToolGPT to "remove the background noise" from a voice note, podcast clip or video, and the same permission step applies.
You do not need an account for any of this. Without one, the service allows a limited number of free minutes of audio per day; when you are signed in, jobs use your account's allowance instead.
The audio is, after you allow it: separation runs on a cloud service. For a video, only the extracted audio is sent; the picture stays on your device.
No. Guests get a limited amount of free processing each day; signing in lets jobs use your account's allowance.
Yes. Both are computed from one upload; use the toggle on the result card to switch between them and download each.
Try it with your own file
Open ToolGPT, attach a file, and type what you want done.