Meet Audiobox from Meta AI. It’s the foundational AI model for generating audio, voices, sound effects, and voice restyling—all from a text description or voice input. Working with Meta and agency partner Author’s Projects, we developed a fun, immersive design for an academic and general audience to explore and understand the possibilities of these new models. Check out the page here.
Audiences can discover preset demos or play with an audio story maker to explore their own creativity.
In the demo section, there are six different models that allow audiences to generate audio in a variety of ways like text-to-speech, text-to-sound effect, and even a magic eraser that removes unwanted background noise.
Here is an example of one of the AI model demos featuring a way for the user to record their voice and then write a script for it to speak on its own.
Ready to have a little fun? With the audio story maker, you can create audio elements from scratch or use one of the preset stories and mess around!
