Free · No signup · In your browser
Suno hands you one flat file with no mix behind it. This pulls it apart into vocals, drums, bass and music, mixes each part like a real session, then masters the result. Drop your song, flip Stem mix: ON in the studio, done.
No vague AI magic — this is what actually happens to your stems, with the same moves a mix engineer would make.
Every stem is measured and pulled toward a role-weighted balance — vocal on top, bass under the kick, music supporting. Faders set before anything else, like a real session.
gated RMS · role-weighted targets
Instruments are ducked in the 1.8–4 kHz zone by up to ~2.5 dB, driven by the vocal's own energy — so they yield to the singer only while the singer is singing.
sidechain duck · 1.8–4 kHz · ≤2.5 dB
Slow-attack opto-style compression levels the vocal, punch-preserving timing keeps the drums alive, then a gentle glue compressor sums it back into one record.
2–4 dB GR per stem · bus glue
Bass is summed to mono below 120 Hz so it translates on any speaker, and a dynamic de-harsh tames the 2.5–5 kHz AI fizz only when it actually spikes — never dulling the whole track.
mono <120 Hz · dynamic 2.5–5 kHz
Mastering polishes the whole file at once — it can make a Suno song louder and cleaner, but it can't un-bury a vocal or tighten a bass that's fighting the kick, because everything is baked together. Stem mixing gets underneath: each part is treated on its own, then glued back into one record. Mix first, master second — that's the real pipeline, and here it happens in one pass.
The separation runs in your browser
Your song is split into stems on your own device before anything else happens. Files are processed, delivered, and deleted — no account, no library of your unreleased music anywhere.
Upload your Suno song here, turn on Stem mix, and hit Master. The engine separates the track into vocals, drums, bass and music in your browser, then runs a real mix pass: it rebalances the levels so the vocal sits on top and the bass sits under the kick, carves a pocket so instruments yield to the vocal, sums the bass to mono below 120 Hz, and glues everything back together before mastering. You don't need a DAW or any mixing knowledge.
Suno prints your song as one flat stereo file — every instrument baked together at whatever balance the AI chose. Sometimes the vocal is buried, the bass is loose, or everything crowds the same space. Mastering alone can only polish that whole file at once. Stem mixing gets underneath it: each part is treated on its own — vocal lifted, bass tightened, instruments carved around the singer — the way a real mix session works.
It's the trick that makes vocals cut through without turning everything else down. The engine ducks the instruments by up to about 2.5 dB in the 1.8–4 kHz range — the zone where voices live — and only while the vocal is actually singing. When the vocal rests, the instruments come back. That's sidechained frequency ducking, the same move professional mix engineers use, done automatically.
The stem separation runs in your browser, on your own device — your song is split locally before anything else happens. The rebalanced result is then processed for mastering and delivered back to you, and files are deleted right after. No account, no library of your unreleased music sitting anywhere.
Yes. The separation and mix pass work on any stereo song — Suno, Udio, Mureka, or a track you recorded yourself. AI songs benefit most, because they ship as a single flat bounce with no mix decisions behind them.
Mixing balances the parts of a song against each other — vocals, drums, bass, instruments — and needs those parts separated. Mastering is the final polish on the finished stereo file: loudness, tone, a safe peak ceiling. This tool does both in one pass: it separates, mixes, then masters, so what you download is ready for streaming.
Separate, balance, pocket, glue, master — one upload, one download. Free.
Mix my Suno song freeChoose a file or drag it here
Supports WAV · FLAC · MP3 · M4A · AIFF