How to make a VST plugin with AI

Can AI make a VST plugin?

Studio 56 builds the sound rather than a new plugin. You type what you want to hear, like “a warm Rhodes with slow tape wobble” or “a gritty 808 that glides between notes,” and the AI builds it in your browser, here on the site. Every sound you make plays in your DAW inside the Studio 56 plugin, a VST3 and AU, so there’s nothing new to install for each one.

The AI doesn’t write new code for each sound. It picks blocks from a library of more than a hundred verified DSP blocks and then sets and wires them together, and you can change any of them by hand.

What you need

You need a Mac or Windows computer and a DAW. You make sounds in your browser, and they play inside Studio 56 in Ableton Live, Logic Pro, GarageBand, FL Studio and most other DAWs except Pro Tools. On a Mac it runs as VST3 or AU, and on Windows as VST3.

Describe the sound

Open Studio 56 in your browser and type the sound where it says Describe a sound. Say what it is first, like a bass, a pad or a pluck, and then how it should feel and move. Words like warm, glassy, wide and punchy all help.

The AI builds it right there out of blocks you can see, and you can play it in your browser the moment it’s made.

Play it and keep changing it

Ask again for a change, like a darker filter or a slower attack, or turn any knob yourself. You can also build a sound by hand from the start, free with no limit. Saving a sound to come back to later is free with an account.

Add Studio 56 to your DAW

When a sound is ready for a track, download Studio 56 and add it to an instrument track like any other plugin, then sign in with the same account. Press Send to my DAW in your browser and the sound shows up inside Studio 56 ready to play. Your first 3 sends to your DAW are free, and Pro sends as many as you want. If you use FL Studio, the FL Studio guide has the exact clicks.

What it’s good at

Studio 56 builds instrument sounds such as basses, leads, pads, plucks, keys and bells. Effects like reverb, delay and distortion go inside a sound, so it doesn’t make an effect to put on an audio track. It can’t clone voices or split a song into stems either. The AI builds the sound and you make the music.

What it costs

Building a sound by hand is free with no limit, and your first 3 asks to the AI are free. Pro is $10 a month, and the AI builds about 70 sounds a month on it. You can use what you make in songs, videos, live shows and commercial releases. The one thing you can’t do is sell the sounds themselves.