Particle.news

Download on the App Store

Nvidia Unveils Fugatto AI Model Capable of Generating Unique Sounds

The generative AI tool promises unprecedented audio flexibility, but its release timeline remains unclear.

  • Nvidia's Fugatto model can create and transform music, voices, and sounds using text and audio prompts, including sounds never heard before.
  • The AI tool offers granular control, allowing users to modify accents, emotions, and even combine disparate audio elements into new compositions.
  • Fugatto uses a 2.5 billion-parameter architecture trained on millions of audio samples, leveraging Nvidia's H100 GPUs for development.
  • Potential applications include music prototyping, video game sound design, localized advertising, and language learning tools.
  • Nvidia has not announced public availability, citing concerns about misuse and the need for careful consideration before release.
Hero image