Stable Audio

Overview

Stable Audio is a family of generative audio models trained on fully licensed data. Three of the models have open weights and are free to download, allowing developers and creators to build custom audio generation tools. The models can produce music, sound effects, and other audio from text prompts.

Best for

Best for
Audio creators and developers who want open, licensable generative audio models to integrate into their own projects

Use cases

• Generate background music and soundtracks for videos or games
• Create custom sound effects for apps and interactive media
• Build audio-generation features into creative tools and platforms

Notes

Stable Audio is a family of generative audio models trained on fully licensed data. Three of the models have open weights and are free to download, allowing developers and creators to build custom audio generation tools. The models can produce music, sound effects, and other audio from text prompts.

Use cases

• Generate background music and soundtracks for videos or games
• Create custom sound effects for apps and interactive media
• Build audio-generation features into creative tools and platforms

Pros

• Models are open weights and free to download
• Training data is fully licensed, reducing legal risk
• Multiple model sizes offer flexibility for different use cases

Cons

• Requires significant GPU resources for local inference
• Audio quality and prompt adherence vary across model sizes
• Limited official documentation for fine-tuning or deployment

Indexed from awesome-generative-ai and enriched against its public facts.