Fully-upgraded AI Singing Model:
A brand new model based on pitch diffusion that significantly improves the effect of non-reference singing
All singers now support Chinese, Japanese, and English languages
Phoneme system upgraded for improved pronunciation
New AI singers:
Luo Tianyi's Japanese and English singing abilities are now available, Yan He and Yuezheng Ling are in public beta (also supporting Chinese, Japanese, and English)
Over ten new AI singers have been added, including KuroBousuku (native Japanese), Barber (Bel Canto Baritone), Chu Ming, Pop series, Rock series, Native English series, and more
Voice mixing function:
Timbre and singing style can be mixed as needed to create a unique new voice.
![](https://i.ytimg.com/vi/gbV274mow4U/maxresdefault.jpg)