duplaja/epub-to-audiobook-hf: Epub to MB4 Audiobook, with StyleTTS2 via HuggingFace Spaces API
Epub to MB4 Audiobook, with StyleTTS2 via HuggingFace Spaces API - duplaja/epub-to-audiobook-hf
Python CLI Tool
Summary
epub_to_audiobook is an open-source Python tool that converts EPUB ebooks into audiobooks using multiple TTS providers (Azure, OpenAI, EdgeTTS, Piper, Kokoro), generating chapter-by-chapter MP3 files optimized for Audiobookshelf. It offers a CLI, a Gradio WebUI for visual interaction, and Docker deployment options.
Highlights
- Supports five TTS backends: Azure TTS, OpenAI TTS, EdgeTTS (no API key needed), Piper TTS, and Kokoro TTS via local OpenAI endpoint
- Generates separate MP3 files per chapter with extracted titles as metadata, specifically optimized for Audiobookshelf navigation
- Includes a Gradio WebUI (added May 2025) for file upload, provider selection, voice configuration, and conversion monitoring
- Chapter title extraction uses HTML title tag parsing with a fallback to first few words of chapter text
- Deployment options include direct Python (3.6+), CLI, and Docker
auto-generated
Context
Audience
Self-hosted audiobook enthusiasts, Audiobookshelf users, and developers interested in TTS pipelines or EPUB processing
DomainText-to-Speech / Audiobook Production
Formatopen-source software project
Accessopen source (GitHub)
Topics
Related
AudiobookshelfAzure Cognitive Services SpeechOpenAI TTS APIEdgeTTSPiper TTSKokoro TTS (local inference server)
