A new generative AI model by NVIDIA, which uses text and audio inputs, can create any combination or music, voices and sounds. A team of researchers in generative AI created a Swiss Army Knife for sound that allows users to control audio output using text. Some AI models can create a song or change a voice but none has the dexterity that the new offering offers. Fugatto is a model that can create or transform any mix of voices, music, and sounds described by prompts, using any combination text and audio files. It can, for example, create a music clip based on a prompt, remove or replace instruments in an existing song, alter the accent or emotion of a voice and even allow people to produce sounds they have never heard before. Fugatto, the first generative AI model to showcase emergent properties, can perform a wide range of audio generation and transformation tasks.