Stability AI launches Stability Audio 3.0 with advanced music generation capabilities
Here's what it means for you.
The launch of Stability Audio 3.0 by Stability AI represents a significant leap in the capabilities of AI-generated music. This innovation not only caters to professional musicians but also opens doors for casual users to explore music creation. As the demand for creative tools in the music industry grows, this technology could redefine how music is produced and consumed. With the ability to generate tracks up to six minutes long, Stability Audio 3.0 enhances the creative possibilities for artists and content creators alike. The introduction of on-device functionality further democratizes access to music generation tools, making it easier for anyone to experiment with sound.
What happened
Stability AI has officially launched Stability Audio 3.0, a new suite of audio models designed to generate music tracks of varying lengths. The top model in this suite can create songs lasting up to six minutes, showcasing the advanced capabilities of the technology. Additionally, some models are optimized to run on-device, allowing for the generation of shorter two-minute tracks.
This release, which occurred on May 20, 2026, includes three models with open weights, enhancing accessibility for users. The models were trained entirely on licensed data, ensuring compliance and quality in the generated music.
The Context
The introduction of Stability Audio 3.0 comes at a time when the music industry is increasingly embracing AI technology. As artists and producers seek innovative tools to enhance their creative processes, the demand for advanced audio generation models has surged. This launch positions Stability AI as a key player in the evolving landscape of music creation.
By offering both long and short track generation capabilities, Stability Audio 3.0 caters to a diverse range of users, from hobbyists to seasoned professionals. The ability to produce music on-device further aligns with current trends in mobile and accessible technology, making it easier for users to create and share their work.
Takeaway
The release of Stability Audio 3.0 is poised to enhance the creative landscape for musicians and content creators. As AI technology continues to evolve, we can expect further advancements in audio generation capabilities, potentially leading to innovative applications in music and sound design.
Future developments may include collaborations with artists utilizing Stability Audio 3.0, as well as updates that expand the functionality of these models. The implications of this technology could reshape how music is produced, offering new avenues for creativity and expression.
Curated tech headlines including AI stories.
"Influential aggregator surfacing the day’s top tech/AI links."
— A47 Editor
Stability AI releases a new family of audio models called Stability Audio 3.0 that is trained on licensed data; the top model can generate six-minute songs (Ivan Mehta/TechCrunch)
Stability AI has launched a new family of audio models named Stability Audio 3.0, which is trained on licensed data and capable of generating music tracks up to six minutes long. This release marks a significant advancement in the company's audio gen...
Startup news with frequent AI coverage.
"Covers launches, funding, and product updates in AI."
— A47 Editor
Stability AI releases a new audio model that can create six-minute songs
Stability AI has unveiled its latest audio model, Stability Audio 3.0, which is capable of generating music tracks up to six minutes long, marking a significant advancement in AI-generated music technology. The model can also run on-device and produc...
Daily AI news: models, tools, and policy.
"Independent outlet tracking the fast pace of AI."
— A47 Editor
Stability AI launches Stable Audio 3.0 with up to six-minute tracks and open weights
Stability AI has launched Stable Audio 3.0, a new generation of audio models capable of generating music tracks up to six minutes long, with three models shipping with open weights and trained entirely on licensed data.