Kokoro TTS

Kokoro-82M is a compact text-to-speech model created by Hexgrad with 82 million parameters, released under the Apache-2.0 license. Despite its small parameter footprint and low hardware requirements, it generates natural speech and supports 54 distinct voices across American English, British English, Japanese, Mandarin Chinese, Spanish, French, Hindi, Italian, and Brazilian Portuguese. Its architecture allows fast inference on consumer CPUs, GPUs, and edge devices. Note: Third-party commercial domains incorporating 'kokoro' are unassociated; the official model is distributed via Hugging Face and GitHub.