OpenAI is engaged on a brand new software that will generate music primarily based on textual content and audio prompts, in response to a report in The Information.

Such a software could possibly be used so as to add music to present movies, or so as to add guitar accompaniment to an present vocal observe, sources stated. It’s not clear when OpenAI plans to launch such a software or whether or not it could be accessible as a standalone product (versus integrating with OpenAI’s ChatGPT and video app Sora).

One supply informed The Data that OpenAI is working with some college students from the Juilliard College to annotate scores as a means to supply coaching knowledge.

Whereas OpenAI has launched generative music fashions previously, they predate the launch of ChatGPT; extra lately, the corporate has been developing audio models centered on text-to-speech and speech-to-text. Different firms with generative music fashions embody Google and Suno.

TechCrunch has reached out to OpenAI for remark.