Text to speech in 80+ languages.

Hundreds of regional voices, each speaking every language. First audio in 80 ms.