logoalt Hacker News

simonwtoday at 2:58 PM0 repliesview on HN

That was about the text-to-speech model, the speech-to-text one was release in January.