Stability AI releases a new audio model that can create 6-minute songs

Stability AI, the company renowned for Stable Diffusion, has introduced Stability Audio 3.0, a new family of audio models. These models are designed to generate professional-grade music, with the most advanced versions capable of producing compositions over six minutes long. This marks a substantial improvement, more than doubling the length achievable by its predecessor, Stable Audio 2.0.

The Stability Audio 3.0 family includes four models: small SFX, small, medium, and large. The small models are suited for on-device sound and music generation up to two minutes. The medium and large models can create full compositions of 6 minutes and 20 seconds, maintaining musical structure and melodic tone.

Stability AI is committed to open access for some of its models, with the small SFX, small, and medium models available with open weights for public use and modification. The large model, however, is accessible primarily through an API and paid self-hosting services, with enterprise licenses required for companies exceeding $1 million in revenue.

The company has strategically partnered with major music labels, including Warner Music Group and Universal Music Group, to ensure its models are built on fully licensed data. This approach addresses growing concerns in the industry regarding data licensing and intellectual property.

In a move to bolster its professional music offerings, Stability AI is developing a new suite of products for musicians. The company has also brought on Ethan Kaplan, former chief digital officer at Universal Audio and Fender, to lead these initiatives, further signaling its commitment to the professional music sector.