the missing part.

the last two years produced systems that write a finished song from a sentence. for a listener who wants a song, that’s magic. for the tens of millions of people who play an instrument — who have a guitar take recorded, a progression looping, an unfinished track open — it’s useless. it replaces exactly the thing they wanted to do themselves, and does nothing about the thing they can’t.

a guitarist can’t program drums. a beatmaker can’t write basslines. their sessions are full of their own playing and silent everywhere else. what they’re offered is loops and sample packs that don’t listen, virtual instruments that demand the production skill that’s missing, and now a text box that removes them from the music entirely.

a bandmate doesn’t write your song. they hear what you’re playing and answer it — on your grid, under your playing. that’s the seat stems takes: one part at a time, for the track you already have, in whole bars at your tempo.

turn up feel and the part starts from your track itself, so it sits on your groove instead of next to it. the more you’ve played, the more it has to follow.

the part, the style and the tempo come from your session — stems fills them in. what you type is the middle: how it plays and how it sounds. a kit, a feel, a room. a dozen words at most.

and the output isn’t the machine’s song. it’s your song, finished. the pride of authorship is the whole reason anyone opens a daw; a tool that erases it has misread the room.

it can’t sing. your key is a strong hint, not a lock — re-roll if a take clashes. a part takes about five seconds on an m3 pro and longer on older macs, and the model wants a few gigabytes while it writes. no fake progress, no spinners, no claim we haven’t measured. the tool that never once lied to a musician is the tool musicians keep.

stems ships as an audio unit. the model — stable audio 3, stability ai’s open model — runs on your mac, nothing you play leaves it, and takes arrive as plain audio in your project. just play.

get stems →