you can find them here: https://aistudio.google.com/app/live

from what I can hear, only 2 sound African American. The rest are all very corporate and generic.

Not even one british voice.

I'm looking to do a real-time voice <> voice POC .. but not sure if the STT or TTS will have the latency and pricing parity that 3.8 live does. What's been your experience with the latter? I do love the idea of cloning specific voice however cause these are just so crrrrrrrrringe.