Upload a song and get two tracks back: the isolated lead vocal and the instrumental without it. Sing over the backing track, practice with the acapella, or use either one in your next mix.
Example: “Sing It Louder”, a synth-pop song with a female lead, split once. The same 30 seconds of the original, the isolated vocals and the instrumental.
Original
Sing It Louder
Synth-pop · female vocal
After vocal removal
Sing It Louder
Vocals only (acapella)
Sing It Louder
Instrumental (lead vocal removed)
The original song was made on ToSong. Loudness is matched for playback.
One upload gives you both halves of a song. Here is what each one is good for.
Nothing to install and nothing to adjust: the model decides where the voice ends and the music begins.
Add an MP3, WAV or M4A, up to 20 MB and 8 minutes. A studio mix separates most cleanly; a live recording with crowd noise leaves more bleed. Use music you made or have the rights to use.
Press Remove vocals. The AI finds the lead voice and lifts it out of the mix, usually within a few minutes. You can leave the page; the result waits in your library.
You get the isolated vocals and the instrumental as separate MP3 files. Play them on the page, then download the one you need, or both.
A vocal remover splits a finished song into two parts: the lead voice and everything behind it. You get an acapella with only the singing, and an instrumental, often called a karaoke track, with the lead vocal gone.
Older tools used a trick: lead vocals are usually mixed to the center, so cancelling the center channel removed them, and often the bass and kick drum too. An AI vocal remover instead uses a model trained to recognize a voice and separate it from the rest of the mix.
Nothing is re-recorded. Both tracks come straight from your upload, so the instrumental keeps the original key, tempo and arrangement, and the acapella keeps the singer's own phrasing.
Once the vocals are out, these tools pick up where vocal removal stops.
Bring a song you made or generated and keep both halves.