Autiobooks addresses the need to convert EPUB e-books into audiobook format (.m4b), targeting users who wish to listen to their digital books. It leverages the Kokoro text-to-speech model to generate natural-sounding audio, making e-books accessible for auditory consumption.

How It Works

Autiobooks processes EPUB files, extracting content chapter by chapter. It then utilizes the Kokoro speech synthesis model, an 82-million-parameter open-weight model, to convert the text into speech. The generated audio for each chapter is saved as a .wav file, which are subsequently combined into a single .m4b audiobook file. This approach allows for high-quality, natural-sounding speech output that can run on consumer hardware.

Quick Start & Requirements

• Installation: Due to current issues with pip installations, download the repository and run directly. The README notes that pip installs are not working and are being investigated.
• Prerequisites: Requires ffmpeg and python3-tkinter. For Linux, install via sudo apt install ffmpeg python3-tkinter. For macOS, use brew install ffmpeg python3-tk. espeak-ng is recommended for improved processing of unknown words.
• Running: Execute using python3 -m autiobooks.
• Documentation: No specific links provided in the README for quick-start or demos.

Highlighted Details

• Utilizes Kokoro, an open-weight TTS model with 82 million parameters for natural speech.
• Supports multiple languages (though currently only English is supported by the project).
• Outputs .m4b files compatible with audiobook players.

Maintenance & Community

• The project is maintained by David Nesbitt.
• Recent changelog entries indicate fixes for race conditions, output file issues, and chapter indexing.
• No community links (Discord, Slack, etc.) are provided in the README.

Licensing & Compatibility

• Distributed under the MIT license.
• The MIT license generally permits commercial use and linking with closed-source projects.

Limitations & Caveats

• Pip installations are currently not functional, requiring direct repository execution.
• Python 3.13 is not supported.
• While Kokoro supports multiple languages, the project currently only supports English.