Lightweight local text-to-speech, running entirely on kittentts.

0 / 2000
1.00×

First request can take up to a minute (the free server wakes up and loads the model). After that it's faster.