Any new text to speech lately? Local still desperately needs a good local Text 2 speech that can emote joy, anger, etc. A fast image 2 video model can probably replace Live2D or VRM in the future.