All tools Separate parts

Stem Splitter

Split a song into vocals, drums, bass and other, mute or solo each part, then download the stems or your own mix. The song never leaves this device; only the model downloads, once.

Drop an audio file here

MP3, WAV, M4A, FLAC, OGG or WebM

Your file stays on this device. Nothing is uploaded.

How to split a song into stems

  1. Choose a song

    Drop an MP3, WAV, M4A, FLAC, OGG or WebM file. It opens in this browser tab and is not uploaded.

  2. Press Split into 4 stems

    The first time, the browser downloads the separation model (90.4 MB) and saves it. The song is split into vocals, drums, bass and other, with the time left shown.

  3. Mix, solo, download

    Mute or solo each stem while it plays, or use a quick mix like No drums. Download single stems, all four in a ZIP, or the mix you hear.

What happens on your device

The tool runs Demucs (the htdemucs model), an open-source network that separates music into four sources, inside this tab. It uses the graphics processor through WebGPU when the browser offers it, otherwise the main processor. The song is cut into 7.8-second pieces that overlap, each piece is separated, and the pieces are joined back.

Nothing is uploaded: no server receives your audio. This browser remembers the settings you pick, never the song.

Limits on your device

Everything runs in your browser, so the longest song depends on your device’s memory. The tool checks the length before it starts and tells you if a file is too long.

DeviceLongest songWith little memory
Laptop or desktop15 min6 min (4 GB of memory or less)
Phone or tablet6 min4 min (4 GB of memory or less)

Four stems mean four full copies of the song in memory, plus the model while it runs. That is why the longest song here is shorter than for the effects. The model also needs about 2 GB of working memory while it runs, so a device with less than 4 GB of memory is told before anything downloads. Without WebGPU the model runs inside the tab and is much slower, so the longest song on that path is 8 minutes on a computer and 4 on a phone.

Opens
MP3, WAV, M4A, FLAC, OGG and WebM
Downloads
Each part as MP3 (128–320 kbps) or WAV (16-bit) at 44.1 kHz, or all parts in one ZIP
Speed we measured
Measured on a Windows PC with an RTX 3050 graphics card, in Chrome and Edge: a 4-minute song splits into 4 stems in about 42 seconds with WebGPU, plus a few seconds to load the model. Without WebGPU the same song takes about 12 minutes. Phones and computers without a graphics card are slower.

Separation speed, memory and quality by device: our full measurements

Questions

Can I split a song from a YouTube or Spotify link?

No. You open a file that is already on your device; the tool has no feature for taking audio from YouTube, Spotify or any other link. The song is decoded and split in this browser tab and never leaves the device. Only the separation model downloads from this site, once.

What is in “Other”?

Everything that is not vocals, drums or bass: guitars, keys, strings, synths and effects. The model has four outputs, so these share one stem.

Can I remove only the drums or only the bass?

Yes. After splitting, pick the quick mix No drums or No bass, or mute any stem yourself, then use Export this mix. The download is exactly what you hear.

How long does splitting take?

In our test with WebGPU (Chrome and Edge, RTX 3050 graphics card), a 4-minute song split in about 42 seconds, a third longer than Vocal Remover because the model returns four parts. Without WebGPU the processor does the work on one core and the same song takes about 12 minutes. The tool shows an estimate before you start and the time left while it runs.

Do the stems add up to the original?

Almost exactly: played together at full volume they sound like the song. Each stem may still carry a little of the others where instruments overlap in pitch and time.