source: techcrunch ai: stability ai releases a new audio model that can create six-minute songs
level: business
stability ai released stability audio 3.0, a new set of audio generation models. the lineup includes four sizes: small sfx and small at 459 million parameters, medium at 1.4 billion, and large at 2.7 billion. the small models can create up to two minutes of sound or music and are designed for on-device use. the medium and large models can produce full compositions lasting six minutes and twenty seconds, maintaining musical structure and melody. this is more than double the length possible with the previous stable audio 2.0.
the small sfx, small, and medium models are available with open weights for anyone to use and modify. the large model is only accessible through an api or self-hosted paid service. companies with over one million dollars in revenue need an enterprise license. stability ai says the models are trained on fully licensed data, following deals with warner music group and universal music group last year. the company is also building new tools for professional musicians, though details are not yet public.
ethan kaplan, former chief digital officer at universal audio and fender, is joining stability ai to lead its professional music efforts. this move mirrors a trend of ai music companies hiring industry veterans, such as suno and elevenlabs bringing on executives from merlin and kobalt. the release comes amid ongoing legal disputes over training data in ai music, making licensed data a key factor for long-term viability.
why it matters: open-weight models for longer music generation let developers and researchers experiment with ai audio while addressing licensing concerns.
source: techcrunch ai: stability ai releases a new audio model that can create six-minute songs