Hacker News Reader

What's the state of the art in run-it-yourself TTS? I haven't seen anything that even comes close to what ElevenLabs is doing

I converted a book to audiobook a couple of months ago using the tts code here: https://github.com/coqui-ai/TTS

Wasn't as good as eleven labs, but it had enough options that I found a voice that I liked to listen to.