Make your first mix with Stable Audio
Stable Audio is built for the iterative creative process: you start with an idea, listen, make a change, and keep working until you get the right sound.
In our Stable Audio web app, you can start with a prompt, then refine a track with notes, try variations, work with multiple individual tracks, and extend a generation to the length you need, all in your browser.
To get started, create an account or sign in at stableaudio.com. This is an early beta, so features are experimental and we would love any feedback you have.
1. Get your first take down, as a full mix or multiple tracks.
Describe what you want to hear in a prompt. You have two options for what you want to generate:
FULL MIX: Your prompt goes to the model exactly as you wrote it, and comes back as one finished mixed track.
MULTI-TRACK: If you select Multi-Track, you’ll get separate tracks to mix, mute, solo and re-record individually. Note: Multi-Track is experimental and results may vary.
Read our latest Stable Audio 3.0 prompt guide for more details on prompting best practices.
2. Start mixing and editing in your session.
Once you generate your first track, you’ll land in the deck. Now we’re cooking with gas.
You’ll see four tracks available in your session, much like an analog tape deck. When all four are full and you want another, bounce two or more together first, which frees up slots without losing the sound.
There are two types of edit you can make: Tape, and splice. These do NOT cost any credits, since you are not generating new audio. Edit away.
| Quick edits | |
| Tape edits rearrange audio you already have rather than generating anything new. | |
| REVERSE | Plays a section backwards. |
| STUTTER | Repeats a short slice in place. |
| FREEZE | Holds a moment, sustaining it across the section. |
| Splice edits | |
| Underneath the tape edits, there is a full splice surface. Select a range, then cut, copy, and paste. | |
| CUT | Loads the range into the reel and removes it from the track. |
| COPY | Loads it and leaves the track alone. |
| PASTE | Click the mode label to choose. |
Hit PRINT to apply edits. When you make edits, apply it either to TRACK (one track only) or ALL, which hits every track at once so the whole arrangement turns together. Whole-mix edits are how you get transitions and drops.
3. Build on it by generating more tracks.
Now let’s keep building. Start layering in stems, regenerating the track to explore different directions, or varying the length.
Four ways to generate new audio
| ADD TRACK | Records one more part into the session, like a counter-melody, or a percussion layer. It listens to what is already there and plays along with it. |
| REGENERATE | Re-records a single track from scratch, leaving every other track untouched. Use it when one track is wrong and the rest are right. |
| REPLACE SECTION | Select a time range on a track and re-record just that stretch. Good for fixing a bad four bars without losing a take you like. |
| EXTEND | Continues a track and composes to the duration, with a real ending instead of a loop. |
Remember: Each of these will cost credits, since you’re generating new audio.
4. Keep iterating and give direction in your own words.
Listen for what's working, and what's getting in the way. Then tell the producer (the chatbox on the left) what you want to change.
For example:
Pull the low end back, there's dialogue over this.
The build starts too early. Hold it longer.
Take out the melody, leave the texture.
Too warm. Colder, more clinical.
Make it more ambiguous.
5. Finish your final mix and share it.
Bouncing
Bouncing commits several tracks down to one (the old tape trick of freeing up space by mixing finished parts together). You can bounce what is done, and keep working in the space that opens up.
Once you bounce, you get one hour to change your mind. UNDO BOUNCE puts the original tracks back.
Export
When the piece is finished you can export the mix or individual stems. You can also share a link to the session so someone else can listen. These are your export options:
MIXDOWN: Downloads the whole session as a single WAV.
STEMS: Downloads every track separately as a ZIP.
NORM: Toggles whether prints and exports are normalized or left at deck level.
You can also easily share your mix with others. Use SHARE to copy a link to the session, so someone can listen without needing to send a file.
Other pro tips and troubleshooting
Some other helpful tips:
Add in effects: Every track has its own effects chain, and the master bus has its own for the finished mix. Effects are applied live during playback, so you hear changes as you make them.
Use audio-to-audio: You can experiment with changes to things like genre, mood, instrumentation, or overall feel while starting from the audio you've already made.
Stick to instrumental prompts: Stable Audio 3.0 models are not designed for vocal output.
Get specific on length: Stable Audio 3.0 enables variable-length generation, which means you can output exactly the duration you want.
Common issues
| Generation failed | Failed generations are not charged. Try again; if one particular prompt keeps failing, rephrase it. |
| Out of credits | Anything that records new audio is unavailable until your balance tops up. Playback, mixing, editing and export keep working, so you can still finish what you have. |
| A track sounds wrong | Regenerate that one track rather than starting over; the rest of the session is preserved. |
| Nothing is playing | Check the track is not muted, and that no other track is soloed. A lit S anywhere on the deck silences everything else. |
| No free slots | Bounce two finished tracks together to open one. |
Want to continue working with Stable Audio in your DAW?
You can also download the Stable Audio plugin, now live in beta. The plugin puts generation inside your session, so you can arrange, edit, and process what you make like any other audio.

