Turn Your Books into Audiobooks with AI

A new open-source tool allows you to convert PDF, EPUB, DOCX, and TXT files into high-quality audiobooks, leveraging Qwen3 TTS, an open-source voice synthesis model.

Main Features:

  • Converts documents in various formats (PDF, EPUB, DOCX, DOC, TXT).
  • Two voice modes: Pre-built speakers (Ryan, Serena, etc.) or voice cloning from a reference audio.
  • Uses the 1.7B model for optimal quality.
  • Smart chunking with sentence boundary detection.
  • Intelligent caching to avoid re-processing.
  • Automatic cleanup of temporary files.

Key Features:

  • Custom Voice Mode: Professional narrators optimized for audiobook reading.
  • Voice Clone Mode: Automatically transcribes reference audio and clones the voice.
  • Multi-format support: Compatible with PDFs, EPUBs, Word documents, and plain text.
  • Sequential processing: Ensures chunks are combined in the correct order.
  • Progress tracking: Real-time updates with time estimates.

Voice Cloning Example:

python audiobook_converter.py --voice-clone --voice-sample reference.wav

The tool automatically transcribes your reference audio, without the need for manual text input.

Performance:

  • Processing speed: approximately 4-5 minutes per chunk (1.7B model).
  • Quality: High-quality audio suitable for audiobooks.
  • Output: MP3 format, configurable bitrate.

GitHub Repository:

https://github.com/WhiskeyCoder/Qwen3-Audiobook-Converter