OpenAI’s voice cloning AI model only needs a 15-second sample to work.

Voice cloning seemed, for years, to be a technology available only to those in possession of a certain economic liquidity, to those belonging to a certain elite class, to those in the 'powerful' category. The latest update of OpenAI, on the contrary, has made this innovation available to anyone in possession of a computer and a small amount of money.

More precisely, this feature, called Voice Engine, is currently available to a limited number of users. The platform has been under development since late 2022 and is what enables the 'read aloud' function within ChatGPT.

The text-to-voice software can create a synthetic voice based on a 15-second sample of any subject's voice, both in English and in other languages.