# Stable Audio 3

> Generate up to 6 min of music, SFX, or ambient audio from text, plus audio-to-audio editing and inpainting.

Generate up to 6 min of music, SFX, or ambient audio from text, plus audio-to-audio editing and inpainting. Stability AI open-weights, 100 free credits.

## Facts
- Website: https://stableaudio3.com
- Pricing: freemium
- Categories: AI & Machine Learning
- LotsDirectory page: https://lots.directory/products/stable-audio-3
- Listing state: Unclaimed. We listed this product from its public website, and its owner has not claimed it yet.

## Detailed info

## What is Stable Audio 3?

Stable Audio 3 is an independent online AI audio generation tool powered by Seed Audio 1.0 that produces music, ambient soundscapes, sound effects, and short voiceovers up to 120 seconds.

## What Stable Audio 3 does

Stable Audio 3 generates custom audio clips up to 120 seconds in length using Text-to-Audio and Audio Reference modes. In Audio Reference mode, users provide up to 3 audio URLs or uploads to guide style and tone alongside a descriptive text prompt. It provides output controls for sample rates from 8 kHz to 48 kHz, playback speed, volume, and pitch adjustments, and exports audio in mp3 and wav formats. Generation costs 450 credits per minute with a 450-credit precharge settled against final duration.

## Who Stable Audio 3 is for

Creators exploring music sketches, game sound effects, audiobook excerpts, and film dubbing production.

## Key features

- Generates up to 120 seconds of audio per task using the Seed Audio 1.0 model.
- Supports Text-to-Audio prompting and Audio Reference guidance with up to 3 audio files.
- Outputs audio in formats including mp3, wav, and pcm.
- Provides output settings for sample rates from 8 kHz to 48 kHz, speed, volume, and pitch.
- Includes 100 signup credits and settles generation at 450 credits per minute.
- Allows commercial usage of generated audio subject to platform terms and input rights.

## Facts from stableaudio3.com, checked 8 November 2026

- Powered by Seed Audio 1.0 rather than Stability AI official model services.
- Produces audio clips up to 120 seconds in length per generation task.
- Charges 450 credits per minute with a required 450-credit initial precharge.
- Allows users to download audio in mp3, wav, and pcm formats.

Written from Stable Audio 3's own website ([Homepage](https://stableaudio3.com/), [Pricing](https://stableaudio3.com/pricing)). The maker can claim this listing to edit it.

## FAQ

### What AI model powers this service?
The service is an independent tool powered by Seed Audio 1.0, not Stability AI official model services.

### What is the maximum duration for a single audio generation?
Each generation task produces audio clips up to 120 seconds (2 minutes) in duration.

### What audio formats can be downloaded?
Audio clips can be exported in formats including mp3, wav, and pcm with sample rates up to 48 kHz.

### How does the credit pricing work?
Audio generation costs 450 credits per minute. Tasks require a 450-credit precharge and settle based on actual output duration.

### What creation modes are available?
Users can generate audio using Text-to-Audio with a prompt or Audio Reference with up to 3 reference audio files.

### Can generated audio be used commercially?
Commercial use is supported subject to the Terms of Service and rights to user inputs.

---
Source: https://lots.directory/products/stable-audio-3 · Machine-readable context by LotsDirectory.