Upload a song (or a video with music) and an AI model strips out the singing, leaving the instrumental. Play it right here and sing along, or download it as MP3 or WAV. The separation runs on our server, not in your browser, and your files are deleted within 30 minutes.
Drop the song you want to sing
or click to browse
Audio: MP3, WAV, FLAC, OGG, M4A, AAC · Video: MP4, MOV, WEBM, MKV · Up to 100 MB
Direct audio or video file links (ending in .mp3, .wav, .mp4, etc.) and Dropbox share links work. YouTube, Instagram and other social links aren't supported — save the song or video to your device first, then upload it above.
The song is processed on our server by an AI model (Demucs), not in your browser. Files are deleted automatically within 30 minutes.
Upload it. The vocals are removed on our server and the instrumental plays here.
Upload the video. The tool uses its audio track.
Paste the share link. The file must be shared with "Anyone with the link".
We do not download from streaming sites. Upload a copy of the song you have instead.
This makes the backing track only, not a lyrics video.
AI vocal removal (Demucs) that keeps the music intact
Play the karaoke track on this page as soon as it is ready
Download the instrumental as MP3 (192 kbps) or WAV
The isolated vocals come too, for practice or guide tracks
Takes MP3, WAV, FLAC, M4A and video files like MP4 and MOV
Songs up to 20 minutes and 100 MB, no signup, no watermark
Make backing tracks for the songs your group actually sings, including regional and film songs that karaoke apps do not carry.
Rehearse over the original arrangement, then record your cover on top of the instrumental.
Prepare a performance track for a function or competition when only the full song is available.
Your file is uploaded to our server, where Demucs, an open-source AI model trained to separate music, splits it into two tracks: the instrumental and the vocals. Most songs take one to three minutes. Both tracks come back to this page so you can listen before downloading.
Files are deleted from our server automatically within 30 minutes. Nothing is published or shared.
They use the same AI model. This page is set up for singing along: it puts the instrumental first and plays it in the page. The Vocal Remover gives equal weight to the instrumental and the acapella, for remixing and editing.
Vocal Remover
AI-powered vocal removal to create instrumentals and karaoke
Pitch Changer
Change the key of a song without changing its speed
Stem Separator
AI-separate audio into vocals, drums, bass, and instruments
Audio Cutter
Trim and cut audio files with precision controls
Audio Converter
Convert audio between MP3, WAV, FLAC, AAC, OGG, M4A formats
M4A to MP3
Convert M4A audio and iPhone voice memos to MP3 in your browser
WAV to MP3
Convert WAV to MP3 and make files about 10x smaller
Audio Compressor
Reduce audio file size by adjusting bitrate and quality
Strip the vocals out of any track and keep a clean instrumental. How AI separation works, and which songs it struggles with.
Remove the vocals, trim the intro, match the levels and join a whole set into one file.
Noise does not just cause mishearing - it makes an AI agent interrupt at the wrong moment. Which audio problems are fixable, and how to tell them apart.
Built by Edesy
Edesy builds AI products for Indian businesses. Start self-serve, no demo call needed.
AI agents that answer and make phone calls 24/7. Handle appointments, inquiries and support in Indian and global languages.
₹50 free credit. No demo call needed.