Stability AI launches Stable Audio 2.0 with full track generation up to three minutes long

Stability AI has unveiled Stable Audio 2.0, a fresh standard in AI-generated audio. This new model is capable of producing high-quality, full tracks with a coherent musical structure, up to three minutes in length at 44.1kHz stereo.

Stable Audio 2.0 takes a leap beyond its predecessor, Stable Audio 1.0, by incorporating audio-to-audio capabilities in addition to the text-to-audio feature. This allows users to upload audio samples and transform them into a diverse range of sounds using natural language prompts. The update also broadens the scope of sound effect generation and style transfer, thereby offering musicians and artists greater control and flexibility in their creative process.

The inclusion of both text-to-audio and audio-to-audio prompting in Stable Audio 2.0 facilitates the production of melodies, backing tracks, stems, and sound effects. This enhancement is set to further enrich the creative process of users.

Stable Audio 2.0 is an advancement of Stable Audio 1.0, which was launched in September 2023. Stable Audio 1.0 was the first commercially viable AI music generation tool capable of producing high-quality 44.1kHz music, utilizing latent diffusion technology.

Stability AI has ensured that Stable Audio 2.0 was exclusively trained on a licensed dataset from the AudioSparx music library. This move respects opt-out requests and guarantees fair compensation for creators, thus maintaining the integrity of the platform.