How do I make a karaoke version of a song?

Drop the song file into the box above. SplitVocals separates it into stems and automatically mixes down an instrumental (everything except the vocal), ready to download as WAV or MP3.

Will the bass and drums stay in the instrumental?

Yes. Unlike old center-channel-cancellation tricks, the AI model removes only the identified vocal source, so bass, drums and other center-panned instruments stay in the mix at full volume.

Can I get the acapella (vocals only) from the same song?

Yes. The same separation produces an isolated vocal stem, which you can download alongside or instead of the instrumental — no need to process the file twice.

Is there a limit on how many karaoke tracks I can make?

No. Because the separation runs on your own device, there's no per-song fee or daily cap — make an instrumental for as many songs as you need.

Why does the vocal still faintly show up on some songs?

That usually happens on live recordings or mixes with heavy reverb and vocal doubling, where the voice physically bleeds into other instrument tracks during recording. It's a limit of the source material, not a setting to adjust.

What file formats can I turn into karaoke tracks?

MP3, WAV, FLAC, M4A, OGG, AAC and WebM, up to 50 MB or 10 minutes per file.

Karaoke maker: remove vocals from any song, free

Drop in a song and SplitVocals removes the lead vocal using AI, leaving a clean instrumental you can sing over — processed entirely on your device, with no upload and no per-song limit.

Make a karaoke track

or drop it here · MP3, WAV, FLAC, M4A, OGG · up to 50 MB / 10 min

Runs locally · Nothing uploaded

How the vocal removal works

Older "vocal remover" tricks work by cancelling out whatever is identical in the left and right stereo channels, which also mutes any instrument mixed dead-center — usually the bass and kick drum go with it. SplitVocals instead runs a Demucs v4 AI model that identifies the vocal as its own source and separates it from drums, bass and the rest of the instrumentation, so the backing track keeps its low end and center-panned instruments intact.

The model, and your audio, stay on your device the whole time — the karaoke track is mixed down locally after separation, with nothing sent to a server.

Who this is for

Karaoke hosts and home singers get an instrumental of a song that doesn't have an official karaoke release. Cover artists use the same instrumental as a backing track for a video or live set. Wedding and event DJs build custom instrumentals for first-dance songs on request. Because there's no per-track fee, it's practical to make an instrumental for every song on a set list, not just the ones that already have one.

Tips for a cleaner instrumental

Vocal removal works best on studio recordings where the lead vocal is a distinct, well-recorded track. Live recordings, vocals with heavy reverb or doubling, and songs where backing vocals are mixed very loud can leave faint traces of the voice in the instrumental — this is a property of the source recording, not something a setting can fix. If a track doesn't clean up well, try the studio version instead of a live or remastered edit.

Download the instrumental as a WAV for the best quality or a 320k MP3 for a smaller file, or grab the isolated vocal from the same pass if you want an acapella too.

No queue, no per-minute fee

Cloud vocal removers often charge by the minute or limit you to a few free tracks a day. Because SplitVocals processes on your own CPU or GPU, there's nothing to meter — make as many instrumentals as your karaoke night needs. A song is usually ready in one to three minutes on a GPU-capable browser like Chrome or Edge; the CPU fallback in Safari and Firefox takes a bit longer.

How it works

1
Drop a song
MP3, WAV, FLAC, M4A, OGG or AAC, up to 50 MB or 10 minutes.
2
AI splits it on your device
A Demucs v4 model runs locally via WebGPU or WebAssembly — nothing is uploaded.
3
Play, solo, download
Preview each stem, then export WAV, 320k MP3, or a ZIP of all four.

Why running locally matters

Private by construction

Your audio is decoded and processed inside the browser tab. There's no upload step, so there's nothing to verify in a privacy policy — check the Network tab yourself.

No queue, no waiting

Processing starts the moment you drop a file. There's no server-side render queue, so a slow day for other users never slows you down.

Works offline after first use

The AI model downloads once (~170 MB) and is cached by your browser. Come back tomorrow, even without a connection, and separation still works.

Free and unlimited

No per-song cost, because there's no per-song server compute to pay for. Split as many tracks as you want, whenever you want.

Frequently asked questions

How do I make a karaoke version of a song?: Drop the song file into the box above. SplitVocals separates it into stems and automatically mixes down an instrumental (everything except the vocal), ready to download as WAV or MP3.
Will the bass and drums stay in the instrumental?: Yes. Unlike old center-channel-cancellation tricks, the AI model removes only the identified vocal source, so bass, drums and other center-panned instruments stay in the mix at full volume.
Can I get the acapella (vocals only) from the same song?: Yes. The same separation produces an isolated vocal stem, which you can download alongside or instead of the instrumental — no need to process the file twice.
Is there a limit on how many karaoke tracks I can make?: No. Because the separation runs on your own device, there's no per-song fee or daily cap — make an instrumental for as many songs as you need.
Why does the vocal still faintly show up on some songs?: That usually happens on live recordings or mixes with heavy reverb and vocal doubling, where the voice physically bleeds into other instrument tracks during recording. It's a limit of the source material, not a setting to adjust.
What file formats can I turn into karaoke tracks?: MP3, WAV, FLAC, M4A, OGG, AAC and WebM, up to 50 MB or 10 minutes per file.

Stem splitter: split any song into 4 stems, free

Upload a track and get four separate stems back — vocals, drums, bass and other — using a Demucs v4 model that runs entirely in your browser, with no upload and no limit on how many songs you split.

Drum extractor: isolate drums from any song, free

SplitVocals pulls the drum track out of a full mix using AI, so you can hear or download just the kit — processed on your own device, with no upload, no signup, and no limit on songs.

Split any song into vocals, drums, bass and more — free, in your browser

SplitVocals separates a song into four stems (vocals, drums, bass, other) using an AI model that runs entirely on your device. No upload, no account, no limits.