On October 26, according to The InformationOpenAI A new tool is being developed that can generate music based on text and audio hints。
The blogger says:This tool can be used to add music to existing videos or guitars to existing human voice tracksI don't know. It is unclear when OpenAI plans to launch the tool, whether it will be published as a stand-alone product or whether it will be integrated into its ChatGPT and video application Sora。
One source reported that:OpenAI is working with a number of students at the Juilliard School of Music to provide training data by placarding the music。
According to 1AI, although OpenAI had in the past introduced generating music models, they were all earlier than ChatGPT; in the recent past, the company had focused on the development of audio models that focused on text-to-text and voice-to-text. At present, companies like Google and Suno also have generating music models。