Sharing a new way to work with Stable Audio

Summary

Stability AI has released two new interfaces for Stable Audio 3.0 that shift the focus from one-shot generation toward iterative editing inside familiar creative environments. The first is a DAW plugin that installs as an instrument in hosts such as Ableton Live and Logic Pro. Once inserted on an instrument track, the plugin can generate audio directly into the session, automatically locking to the project’s BPM, and supports lengths from short segments up to six-minute tracks. Users can store multiple takes in a playlist for later comparison, then treat the resulting clips like any other audio region for arrangement, mixing, and effects processing.

The second interface is an expanded web application at StableAudio.com. After an initial prompt generates a first take, the tool allows further text instructions to refine elements such as low-end balance, build timing, or instrumentation. An audio-to-audio mode produces variations from existing material, while per-track and master controls for level, pan, mute, solo, and effects support mixing of multiple stems. Finished sections can be bounced individually or exported as a complete mix for continued work elsewhere.

Both the plugin and the web application run on the same commercially-safe models, so generated outputs remain the user’s property and may be distributed without restriction. The releases are described as an early beta, with some features still experimental; the company intends to refine the tools based on ongoing user feedback.

Why it matters

This update is highly relevant for product teams and builders in the Dutch creative and media sectors. The emphasis on commercially-safe models aligns well with strict EU copyright and AI regulations, offering a compliant way to integrate generative audio into professional workflows.