Stable Audio Open: open weights for sound effects and audio samples
In one sentence Stability AI releases on Hugging Face the weights of Stable Audio Open, a roughly 1-billion-parameter text-to-audio model that generates up to 47 seconds of 44.1 kHz stereo audio, trained on Freesound and Free Music Archive recordings.
Stable Audio Open generates short audio clips from a text description: sound effects, foley, drum beats, instrument riffs, ambiences. It is not meant for full songs, melodies or vocals.
Unlike the commercial Stable Audio service, which produces tracks up to three minutes, here the weights are downloadable and the model can be fine-tuned on your own audio data.
The license is the Stability AI Community License, not a Creative Commons license.
Companies
Stability AI
Tools
Stable Audio Open 1.0
Tags
Sources