AI Stem Splitter

1. 1

   Upload the mix. Anything the platform reads works; a 320 kbit/s MP3 splits as readily as a master WAV.

2. 2

   Choose which stems you want. Vocals, drums, bass and other are the standard set, with guitar and piano available as separate parts.

3. 3

   Download the stems individually, or take the zip. Each one is a full-length file aligned with the original, so they sum back to the mix.

One audio file up to 500 MB, and the list of stems you want out of it. No prompt and no tuning.

One WAV per stem at the rate and depth you choose, plus a zip of the set. The stems are time-aligned with the input, so dropping all of them on a timeline reconstructs the original mix.

WAV, FLAC and MP3, at 44.1 or 48 kHz and 16 or 24-bit. You pick the container and the rate on the form before you run it.

Editors who need the music under a dialogue scene turned down without the drums going with it. Producers studying an arrangement. Game teams who were handed a finished track and need the percussion on its own layer so it can duck.

Five credits per split, flat. Not per stem and not per minute — a four-stem split of a six-minute track and a two-stem split of a thirty-second clip cost the same five credits.

Every run is a queued job. It holds its credit estimate while it works and settles to the real cost when it finishes; if it fails or you cancel it, the hold is released in full. You see the estimate before you spend anything.

See pricing

Everything you generate is royalty-free for commercial use — client work, monetised video, games you sell, ads. No per-use fee and no attribution requirement. Generations are yours; uploads stay private by default, and the inputs to the isolator and splitter are deleted automatically.

Which stems can I get?

Vocals, drums, bass and other as the standard four, with guitar and piano available as separate stems. You pick the set on the form, and the cost is the same whichever you choose.

Do the stems add back up to the original?

They are full-length and time-aligned with the input, so laying all of them on a timeline reconstructs the mix. That is also the honest test of a split: if summing them sounds wrong, the separation is what you are hearing, not the alignment.

Why is it a flat five credits?

Because the work does not scale the way generation does, and a per-stem price would charge you for asking a clearer question. Five credits covers the split however many stems you asked for and however long the track is.