A new text-to-speech (TTS) model called Kokoro v1.0 has been released, featuring 82 million parameters and licensed under Apache 2.0. This multilingual TTS model supports English, Spanish, French, Italian, Japanese, and Mandarin. It operates fully offline and locally in web browsers, utilizing WebGPU for acceleration, allowing for real-time speech generation. Users can generate 10 seconds of speech in approximately one second at a cost of $0. The model is designed to be easily integrated, requiring only five lines of code to implement. Initial user feedback indicates a significant improvement in speed compared to previous versions without WebGPU support.