Stable Audio

Stable Audio is Stability AI’s audio model: high-fidelity short-form music and textures for intros, stings and loops.

Stable Audio specialises in production-quality short pieces: 10–45 second cues, stings, risers and ambient textures with notably clean, hi-fi output. It is the pick for polished intros, outros and transition moments.

Where CassetteAI covers full-length underscores, Stable Audio covers the punctuation: the five-second riser into a reveal, the loopable bed under a title.

At a glance

  • Type: Music & sound textures
  • Duration: 10–45 seconds
  • Fidelity: High-fidelity stereo output
  • Best for: Intros, stings, risers, loopable beds

What it is best at

  • Clean, hi-fi output suited to polished edits
  • Purpose-built for short cues and stings
  • Prompt-controllable genre, mood and instrumentation

What it costs

From 3 tokens ($0.03) per track; 8 tokens ($0.08) at the default settings.

SettingsTokensPrice
10 s3$0.03
20 s6$0.06
30 s8$0.08
45 s12$0.12

Prices come from the live catalog at $0.0100 per token, the pay-as-you-go rate; subscriptions include tokens at a lower rate. The exact price is shown before you generate, and a failed generation is refunded.

Questions

How does Stable Audio compare with CassetteAI?
Stable Audio for short, polished cues (10–45 s); CassetteAI for long underscores (up to 3 min). Many edits use both: a Stable Audio intro sting over a CassetteAI bed.
Can Stable Audio make sound effects?
It handles textures and ambience well. For discrete effects (impacts, whooshes, UI sounds) the dedicated ElevenLabs SFX model is usually the better pick.
How is Stable Audio billed?
Per generation, scaling with duration, with the token price in the picker.

Related models