Stable Audio

What it does

Stable Audio 3.0 is a family of generative audio models from Stability AI. You describe the sound you want in a text prompt, such as genre, instruments, mood and intended use, and the models produce music tracks or sound effects. Full tracks can reach six minutes with a structured arrangement. The models were trained on fully licensed data.

People use it to generate audio from a written brief, then change it using editing options: modify a segment of a track, rework part of a song, or extend a composition. The family has Large (enterprise-grade production), Medium (full songs, open weights), and Small and Small SFX (built to run on a mobile device, open weights). It is distinct for combining downloadable weights, customization on your own library, and commercial licensing.

It suits musicians, game and video producers, developers building audio features, and companies wanting audio models on their own infrastructure. You can start in the Stable Audio web app, call the model through the Stability AI API, download the Small and Medium weights, or arrange an Enterprise license. Legal indemnification is offered only under the Enterprise license. Prices and usage limits are not confirmed here.

What it can do

• Generate music from text prompts describing genre, instruments and mood
• Create full tracks up to six minutes with dynamic structure
• Generate sound effects with a dedicated Small SFX model
• Edit audio: modify a segment, rework part of a song, or extend a composition
• Customize models on your own audio library
• Download open weights for Small and Medium models

Example workflows

Draft a game boss-fight track

Input: a text prompt describing a cinematic, emotional fantasy boss-fight piece with choir, organ and sparse percussion. Steps: generate a track, then rework or extend a section to fit the scene. Output: a composition of up to six minutes.

Fine-tune on a studio library

Input: a company's own audio library plus an Enterprise license arrangement. Steps: customize the model on the library with Stability AI's research team support, then deploy on your own infrastructure. Output: a customized model generating audio in the library's style.

Before you choose

• Open weights cover only Small and Medium; Large is described as enterprise-grade, with availability route not detailed
• Legal indemnification is stated only under the Enterprise license
• Pricing, usage limits and the terms for commercial use of the open weights are not confirmed here

Listed under

• Creative
• Music and audio
• Audio
• Audio models
• Creators
• Developers
• Enterprise
• Music generation

Get started

Product page links to a web app; API and downloads are alternative routes.

1. Open the Stable Audio web app from the product page to generate music immediately
2. For app integration, get access to the Stability AI API for managed hosting
3. For local use, download the Small or Medium open weights, or request an Enterprise license for self-deployment

Web app, API and Enterprise license prices were not confirmed; open-weight downloads are mentioned but their license terms and compute costs are not detailed.

We couldn’t find published prices for Stable Audio. See their pricing page ↗

Sources for this overview

Product pages, documentation, and pricing checked 29 Oct 2026. Each source supports part of the overview; availability and prices can change.

• Stable Audio 3.0 | Generative Audio Models — Stability AI ↗stability.ai

Company website

stability.ai: product, pricing and policy pages

YC directory mirror

Batch and company name

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

Last checked 29 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.