Free online vocal and instrumental splitter
Try splitting a song into voice-focused and instrumental-focused WAV files. Listen to both results; vocal or music traces can remain.
Separate vocals and instrumental audio from a song
Use SnipyCut to attempt to separate a mixed recording into a voice-focused WAV and an accompaniment-focused WAV. Select the source in your browser without installing a vocal-removal program. The recording is uploaded so the separation model can run on the server.
This does not recover original studio tracks. The model estimates components from a mixed file. Music may remain in the vocal output, while vocals or reverb may remain in the instrumental. Compare both results with the source before using them.
How to use the free online vocal splitter
- Select one audio or video file, up to 5 minutes long and no larger than 490 MB.
- Check the source preview. There is no separate separation-strength control to configure.
- Select “Start Processing” and follow upload and processing status. Your job may wait in a queue.
- Listen to both results, then download a WAV file or the ZIP containing both outputs.
Understand voice.wav and instrumental.wav
voice.wav is the voice-focused component. It can help you inspect words, timing or a vocal passage. instrumental.wav is the accompaniment-focused component. You can try it as a rehearsal backing track, but it is not guaranteed to be completely free of vocals.
Exporting to WAV does not improve the quality of the original recording. Distortion and detail lost through earlier compression cannot simply be restored by separation. Keep the source file and use it for further attempts rather than repeatedly processing an already separated copy.
Review the separation result
Backing vocals, strong reverb, dense instrumentation and sounds similar to the singer can make separation difficult. Test a short section first. Listen to word beginnings and endings, high-frequency sounds and gaps between vocal phrases. Remaining sound in either stem does not by itself mean processing failed; separation is approximate.
Choose the right tool for video and speech
Video input here produces audio stems without a new video file. For a video with voice-focused audio, use remove music from video. For steady background hiss, try noise reduction. For spoken words as text, use subtitles and transcription.
Server processing and downloads
Processing can take longer than the recording. Refreshing may resume a saved job in the same browser. “Close job tracking” stops tracking but does not cancel the server job. Download the outputs when they are ready and check they suit your intended use.
Frequently asked questions
Is the vocal splitter free?
Yes. Start separating vocals and instrumental audio in your browser for free. Choose one file, up to 5 minutes long, with a total upload limit of 490 MB.
Which files contain the vocals and instrumental?
voice.wav contains the voice-focused component and instrumental.wav contains the accompaniment-focused component. Download them individually or together in results.zip.
Can I make a backing track for karaoke?
You can try the instrumental output for this purpose, but complete vocal removal is not guaranteed. Backing vocals, reverb or vocal traces can remain. Listen to the result first.
Does uploading a video create a new video?
This tool separates the audio from video input into two WAV files; it does not produce a new MP4. Use the remove-music-from-video tool if you need a video with voice-focused audio.
Where does processing happen and how long does it take?
The file is uploaded and the model runs on the server. Processing time depends on the recording and queue load and can exceed the recording length. Closing job tracking does not cancel the server job.