Everything on tiny-voca
1 mention, newest first.
tiny-voca is an 18 MB open English TTS model that runs in the browser
A Reddit user released tiny-voca, a small English text-to-speech model with code, weights and dataset all open. It has 4,537,291 parameters, two voices (one female, one male) at 24 kHz, and runs about 70x faster than real time on 4 CPU threads. It was trained from scratch in 7.7 hours on one RTX 4090.
Usable today