Make real music.
No experience needed.
If you can describe it, you can make it. Type a vibe like “chill lo-fi for rainy nights” and AI turns your words into a complete song — vocals, instruments, everything. No instruments. No theory. No software to learn.
Three steps. That's the whole tutorial.
No timelines, no MIDI, no mixing desk. You talk, the AI makes music.
Say what you feel
Type an idea the way you'd text a friend: a happy acoustic song about summer road trips. Wrong words? There's no wrong words.
AI writes & performs it
The LM plans your song — tempo, key, structure, lyrics — then the diffusion model actually sings and plays it. All the "music stuff" is handled.
Listen, tweak, share
Don't like it? Hit generate again — every take is different. Love it? Download the audio and share it anywhere.
“An upbeat funk song about finally finishing my thesis, with a groovy bassline and hand claps”
One sentence in. A finished, singable track out — usually before your coffee cools down.
What you can do without knowing anything
Every one of these works from plain English (and 50+ other languages).
Full songs with vocals
Real singing — lead vocals, harmonies, lyrics in any language you type. Or say “instrumental” and skip vocals entirely.
Any genre you can name
Jazz, drill, K-pop, sea shanties, Gregorian trap — over a thousand styles and instruments. Mix them freely.
Lucky dip mode
No idea at all? Tap Surprise Me and the AI invents a whole song concept for you. Great for beating blank-page fear.
Grows with you
Start simple. When you're curious, optional knobs let you steer tempo, key, length and more — always optional, never required.
Remix & repaint
Upload a track and restyle it, extend it, or regenerate just a section — like editing text, but for music.
Private by design
Everything runs on this machine's own GPU. Your prompts and songs never leave it.
You write feelings. The AI handles theory.
This is what “no experience needed” actually means:
You say 🗣️
- “sad piano, rain outside”
- “make it faster”
- “like a campfire singalong”
- “verses about my cat, big chorus”
The AI figures out 🧠
- Key: A minor · Tempo: 68 BPM · Time: 4/4
- Structure: intro → verses → bridge → outro
- Instruments: felt piano, soft strings, vinyl crackle
- Melody, lyrics, phrasing and the actual performance
Listen live
AI-generated music, streaming now.
⚙️ Running your own engine? (for self-hosters)
This website talks to a local ACE-Step 1.5 API server. Start it once, then everything here just works:
pip install -U acestep python -m acestep.api_server --host 127.0.0.1 --port 8001
Needs an NVIDIA GPU (as little as 4 GB VRAM) or Apple Silicon. Full docs at the ACE-Step GitHub repo.
Your first song is
one sentence away.
Open the Playground, type whatever's in your head, and press generate.