Separate voice and music
AI separates the voice and the music in your video or audio.
The first run takes longer.
How it works
- 1
Add a file
A video or audio up to 5 minutes: a Reel, a voiceover, a song.
- 2
Let the AI split it
The model separates the voice from everything else. With a GPU it usually takes less time than the clip itself.
- 3
Download
Voice and music as MP3 or WAV, or the same video with only the voice or only the music.
Frequently Asked Questions
Is my file uploaded?
No. The separation runs in your browser. Only the AI model is downloaded, once.
How do I remove background music from a video?
Add the video, press Separate, then download Video (MP4) next to Voice. You get the same video with only the speech.
How good is the separation?
It uses HTDemucs, one of the most accurate open models for this. Speech over music comes out clean. A song where the music itself has vocals is harder, some echoes may remain.
Can I use it on any song?
Use it on your own content or tracks you have the rights to. Separating a song doesn't give you the rights to it.