Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Audiobook Creator is an app that converts books (EPUB, PDF, TXT etc.) into fully voiced audiobooks with intelligent character voice attribution. It uses LLMs and Kokoro/Orpheus TTS to generate engaging, multi-voice audiobooks. Features include emotion tag addition, character identification, and customizable narration. Licensed under GPL-3.0
| Date | Stars |
|---|---|
| 2026-07-31 | 518 |
| 2026-08-06 | 519 |
Today
+1 stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# Audiobook Creator
## Overview
Audiobook Creator is an open-source project designed to convert books in various text formats (e.g., EPUB, PDF, etc.) into fully voiced audiobooks with intelligent character voice attribution. It leverages modern Large Language Models (LLMs), and Text-to-Speech (TTS) technologies to create an engaging and dynamic audiobook experience. The project features professional-grade TTS engines including [Kokoro TTS](https://huggingface.co/hexgrad/Kokoro-82M) and [Orpheus-TTS](https://github.com/canopyai/Orpheus-TTS) with advanced async parallel processing, offering exceptional audio quality with memory-efficient processing. The project is licensed under the GNU General Public License v3.0 (GPL-3.0), ensuring that it remains free and open for everyone to use, modify, and distribute.
Sample multi voice audio for a short story generated using *Orpheus TTS* (with added emotions and better sounding audio) : https://audio.com/prakhar-sharma/audio/sample-orpheus-multi-voice-audiobook-orpheus
Sample multi voice audio for a short story generated using *Kokoro TTS* : https://audio.com/prakhar-sharma/audio/generated-sample-multi-voice-audiobook-kokoro
Watch the demo video:
[](https://www.youtube.com/watch?v=E5lUQoBjquo)
<details>
<summary>The project consists of four main components:</summary>
1. **Text Cleaning and Formatting (`book_to_txt.py`)**:
- Extracts and cleans text from a book file (e.g., `book.epub`).
- Normalizes special characters, fixes line breaks, and corrects formatting issues such as unterminated quotes or incomplete lines.
- Outputs the cleaned text to `converted_book.txt`.
2. **Character Identification and Metadata Generation (`identify_characters_and_output_book_to_jsonl.py`)**:
- Step 1: Identifies unique characters and their age, gender in the text using an LLM via an OpenAI-compatible API.
- Step 2: Attributes the speakers identified in each line of text using the characters identified in step 1 using an LLM via an OpenAI-compatible API.
- Outputs two files:
- `speaker_attributed_book.jsonl`: Each line of text annotated with the identified speaker.
- `character_gender_map.json`: Metadata about characters, including name, age, gender, and gender score.
3. **Emotion Tags Enhancement (`add_emotion_tags.py`)**:
- Adds emotion tags (e.g., `<laugh>`, `<sigh>`, `<gasp>` etc.) to enhance narration expressiveness.
- Processes `converted_book.txt` and outputs enhanced text to `tag_added_lines_chunks.txt`.
- Requires Orpheus TTS engine for emotion tag support.
4. **Audiobook Generation (`generate_audiobook.py`)**:
- Converts the cleaned text (`converted_book.txt`) or speaker-attributed text (`speaker_attributed_book.jsonl`) into an audiobook using advanced TTS models.
- **Multi-Engine Support**: Compatible with both Kokoro TTS ([Hexgrad/Kokoro-82M](https://huggingface.co/hexgrad/Kokoro-82M)) and Orpheus TTS ([Orpheus-TTS](https://github.com/canopyai/Orpheus-TTS)) with engine-specific voice mapping.
- Offers two narration modes:
- **Single-Voice**: Uses a single voice for narration and another voice for dialogues for the entire book.
- **Multi-Voice**: Assigns different voices to characters based on their gender scores.
- Saves the audiobook in the selected output format to `generated_audiobooks/audiobook.{output_format}`.
</details>
## Key Features
- **Advanced TTS Engine Support**: Seamlessly switch between Kokoro and Orpheus TTS engines via environment configuration
- **Async Parallel Processing**: Optimized for concurrent request handling with significant performance improvements and faster audiobook generation.
- **Gradio UI App**: Create audiobooks easily with an easy to use, intuitive UI made with Gradio.
- **M4B Audiobook Creation**: Creates compatible audiobooks with covers, metadata, chapter timestamps etc. in M4B format.
- **Multi-Format InpExcerpt of 19,233 characters
Read on GitHub
113
3
1
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:e579a8f4f98e830e, llm:Repository description: 'converts books ... into fully voiced audiobooks with intelligent character voice attribution. It uses LLMs and Kokoro/Orpheus TTS to generate engaging, multi-voice audiobooks. Features include emotion tag addition, character identification, and customizable narration.'
matched fp:e579a8f4f98e830e, llm:Repository description: 'converts books ... into fully voiced audiobooks with intelligent character voice attribution. It uses LLMs and Kokoro/Orpheus TTS to generate engaging, multi-voice audiobooks. Features include emotion tag addition, character identification, and customizable narration.'
matched fp:e579a8f4f98e830e, llm:Repository description: 'converts books ... into fully voiced audiobooks with intelligent character voice attribution. It uses LLMs and Kokoro/Orpheus TTS to generate engaging, multi-voice audiobooks. Features include emotion tag addition, character identification, and customizable narration.'