Kokoro TTS

Kokoro TTS converts text into high-quality speech using AI models. It addresses the need to generate audio from written material such as audiobooks, articles, and other long-form content.

The service provides a selection of voices including American female options Sky, Bella, Nicole, and Sarah, American male options Adam and Michael, British female options Emma and Isabella, and British male options George and Lewis. Users choose between two processing modes: Local VPS, which runs on CPU and is slower, or Hugging Face, which requires zero GPU and processes faster. Text-to-speech computation occurs server-side for improved reliability. The system uses local models; the initial request may require initialization time while later requests complete more quickly.

Input occurs either by uploading TXT or MD files up to 10 MB or by entering text directly with a limit of 500 characters. For larger uploads the tool automatically splits content into chunks and produces separate audio files to preserve quality and keep outputs manageable. The interface includes options to view all voices and to clear input.

It is delivered as a web-based application built for Algoran and described as ready to use. The page indicates the TTS system is running with local models.