tts ru-svetlana-neural sounds very bad. Can you add other female voices? Or can I create my own voice model from audiobook recordings?

It’s deployed on Vultr VPS. As long as their service is working, there won’t be any problem. There is another serverwatch script monitoring the API. So it should be stable/robust enough.

One more question. For example, how reliable is it to use your API? I mean, maybe you just close the API or it will be down for a long time?

It’s clear. There are no male voice in other languages. Okay, I just wanted to ask, if not planned to add.

try en-US-GuyNeural instead