Stable Audio Open from Stability AI can create simple sound effects from text prompts
Not to be confused with the premium Stable Audio
2 min. read
Published on
Share this article
Improve this guide
Read our disclosure page to find out how can you help MSPoweruser sustain the editorial team Read more
Key notes
- Stability AI launched the open-source Stable Audio Open.
- Unlike its Stable Audio’s premium offering, this only best works for simple sound effects.
- It was trained using over 486,000 samples and is able to create up to 47 seconds.
Stability AI, the famed AI startup that created Stable Diffusion, is now expanding its reach in the audio market even further. The company has recently launched Stable Audio Open, its latest text-to-audio model, and according to its demo, it lets you generate high-quality audio samples and more.
This may remind you of the latest Suno v3 model, which claims that it can create “radio-quality music” and is available as a Copilot extension. But, it’s totally different. Stable Audio Open only lets you create up to 47 seconds and is ideal for drum beats, instrument riffs, ambient sounds, and foley recordings.
The best part of it? It’s open source, and you can try it out on HuggingFace today. It’s said to be trained using over 486,000 samples from libraries like FreeSound and Free Music Archive.
“While it can generate short musical clips, it is not optimised for full songs, melodies, or vocals. This open model provides a glimpse into generative AI for sound design while prioritising responsible development alongside creative communities,” the company says.
Still, that’s not much of a song, but if you want to create a 3-minute track, Stable AI has premium offerings for Stable Audio. The company just launched its v2.0 not too long ago and you can create up to 3 minutes in length at 44.1kHz stereo.
