Stability AI Launches Sound Generating Model ‘Stable Audio Open’ With Short Audio Clips In Mind

Stability AI sound generating
  • Save

Stability AI sound generating
  • Save
Photo Credit: Stability AI

Stability AI announces a new AI model for generating sounds and music, Stable Audio Open, trained exclusively on royalty-free recordings from Free Sound and the Free Music Archive.

Stability AI, best known for its AI art generator, Stable Diffusion, has announced the launch of a new AI model for generating sounds and songs, trained exclusively on royalty-free recordings from Free Sound and the Free Music Archive. Named Stable Audio Open, the model can produce up to 47 seconds of audio based on text descriptions.

“We’re excited to announce Stable Audio Open, an open source model optimized for generating short audio samples, sound effects, and production elements using text prompts,” said Stability AI in a statement. “This release marks a key milestone as we further open portions of our generative audio capabilities to empower sound designers, musicians, and creative communities.”

With a training set of around 486,000 samples of royalty-free music and sound libraries, Stable Audio Open aims to provide a versatile tool for creating an array of audio elements based on a text input. Users can generate instrumentals and drum beats, ambient noise, and most audio production elements for use in videos, film, and television.

The tool is intended to be an open-source text-to-audio resource for sound designers, musicians, and creative professionals, enabling them to create high-quality audio from a simple text prompt. Stable Audio Open’s royalty-free training makes it particularly useful for creating sounds for use in music production and sound design.

Stability AI encourages sound designers, musicians, developers, and audio enthusiasts to download the current Stable Audio Open model, explore its capabilities, and provide the company with feedback.

“While an exciting step forward, this is still just the beginning for open and responsible audio generation capabilities,” said Stability AI. “We look forward to continuing research and prioritizing development hand-in-hand with creative communities. Let the open exploration of AI audio begin!”

Stability AI also offers a commercial model of Stable Audio, which enables users to produce high-quality full tracks with coherent musical structure of up to three minutes in length. Stable Audio Open, by contrast, is not optimized for full songs, melodies, or vocals. The company notes that, in its current state, the open model provides “a glimpse into generative AI for sound design” while prioritizing responsible development.