Kokoro is an open-weight Text-to-Speech (TTS) inference library designed
for the Kokoro-82M model.

It provides fast, small, and free text-to-speech capabilities using PyTorch
and the Hugging Face ecosystem. The library supports multiple languages
and voices through the Misaki grapheme-to-phoneme engine.

For full English language support, install textproc/py-phonemizer separately.

Kokoro is the fastest production quality English Text-To-Speech convertor
available as of early 2026.