Early access

Generate a sound. Drag it into your session.

Sona is a small audio plugin that puts a curated set of AI sound models behind one screen of real knobs and switches. Describe what you want, turn a couple of dials, and drop the result straight onto a track — without leaving your DAW.

Three steps, then you’re back to writing

  1. 01

    Pick a model

    A small, curated set. Each one is good at something specific, and Sona tells you which.

  2. 02

    Turn the knobs

    The provider’s real parameters, on real controls. No hidden prompt engineering behind your back.

  3. 03

    Drag it in

    The take lands on your timeline. That’s the whole loop, and it takes seconds.

Runs where you work

In your DAW

VST3 and AU — Logic, Ableton, Cubase, Reaper, Pro Tools

Desktop

A standalone app for macOS, Windows and Linux

Browser

The whole thing, no install, at app.sona.audio

iOS

In the works — the app already runs anywhere a webview does

One account, all of them. Sign in with a code sent to your email — no password to forget.

Straight answers

What does it cost?
Nothing right now. Sona is in early access and every account comes with a generous allowance of generations. When that changes, it will be obvious and fair — no surprise meter.
Which models are in there?
A handful of the best text-to-audio, loop and sound-effect models available, plus audio-to-audio for remixing your own material. We add them slowly and only when they earn their place.
Do you keep my audio?
Your generations are stored privately so you can come back to them, and only you can access them. We don’t train on your material.
Who is making this?
A very small team who got tired of alt-tabbing to a browser, downloading a file, finding it in Finder, and dragging it into a session. That’s it. That’s the whole origin story.

Try it, then tell us it’s wrong

It’s early and we’d rather hear what’s broken than guess. If a model misbehaves, a control feels off, or you run out of generations — write to us.