Drop a track you own or have rights to. The Studio remembers it as this project's reference.
MP3 · WAV · FLAC · M4A · stays on your device, only the name & details are saved
Tempo & loudness are measured free on your device the moment a reference lands. The full AI read — genre, mood, vocals, instrumentation, lyrics — runs on demand.
Saved with the project. Write, Generate and Iterate all inherit these fields.
Use [Verse], [Chorus], [Drop] section tags. The pre-flight check runs as you type.
Remixed versions appear here — keep your structure, tags and flow, fresh words. The remix instructions live on the server, never in this page.
Your persona + lyrics, compiled into the target generator's syntax. Copy, generate, then bring the takes back to Iterate.
Drop your generated takes back in. Each is measured on your device — tempo, loudness, clipping — and scored against the persona's targets. Free, no credits.
multiple files welcome · audio stays on your device
Next build: AI vocal/style scoring and the splice bench (combine best sections across takes).
A genre-tuned EQ → glue compression → loudness chain runs right here in the browser on your winning take. Free — it never leaves your machine.
The linter checks what playlists and algorithms care about, then exports your pack.