Official press release from Stability AI or the film production studio

Stability AI Launches Stable Audio Open for Generating Short Audio Samples, Sound Effects, and Songs

Jun 5, 2024, 06:55 PM

Stability AI has launched Stable Audio Open, an open-source model designed for generating short audio samples, sound effects, and songs from text prompts. The model can produce audio up to 47 seconds long and is trained on data from FreeSound and the Free Music Archive. Stability AI, known for its AI-powered art generator Stable Diffusion, aims to enhance audio production capabilities with this new tool. Stable Audio Open is particularly suited for video and film production, offering users the ability to create custom audio data. Researchers have noted that the model achieves the best FD results among competitors, a benchmark for realism in generated sounds. This release marks a key milestone for Stability AI as it further opens portions of its generative audio capabilities.