Show HN: Kokoro TTS in the browser, 41 voices in 6 languages, runs locally (sandbook.app)

Kokoro TTS has launched an innovative text-to-speech tool that operates entirely within your web browser, offering 41 voices across six languages including English, Spanish, French, Italian, Brazilian Portuguese, and Hindi. The tool leverages the Kokoro-82M open-weight text-to-speech model, enabling users to type text and generate speech locally without any uploads, ensuring complete privacy. Users can enjoy this functionality for free, with the option to save generated audio as WAV files directly from the browser. The application supports efficient audio rendering on both WebGPU-enabled devices for faster performance or slower CPUs via WebAssembly, making it accessible to a broad range of users. This development is significant for the AI/ML community as it highlights the increasing capabilities of local models to provide high-quality speech synthesis without reliance on external servers, thus enhancing user privacy and reducing latency. With commercial rights granted under the Apache-2.0 license, creators can use the generated audio in various applications without concerns over copyright claims. This tool not only enriches the landscape of text-to-speech technology but also serves as a valuable resource for developers and content creators seeking customizable and high-performance speech generation solutions.