Kokoro

Kokoro is a text-to-speech model and inference library for developers adding speech generation to personal projects or production applications. It runs on your own hardware and uses the compact Kokoro-82M model.

• Speech generation: The Python library turns text into speech with Kokoro-82M.
• Compact model: Its lightweight architecture offers a smaller model for local speech generation, with reported speech quality comparable to larger models.
• Apple Silicon acceleration: Speech inference can use the GPU on Apple Silicon Macs.

Kokoro supports Windows and macOS. The library uses the Apache 2.0 license, and the model weights are Apache-licensed. You can run inference locally; the Google Colab examples run in a cloud notebook.