Repolode
Submit
denizsafak

denizsafak/abogen

Desktop and web app that converts EPUB, PDF, text and subtitle files into audio with synced subtitles, using Kokoro-82M TTS.

Open-source alternative to:Audible

View on GitHubWebsite
Stars
6,131
Forks
500
Open issues
48
Last commit
3 days ago

What is it?

Abogen turns ePub, PDF, plain text, markdown and subtitle files into speech audio, generating matching subtitles alongside. It is built on the Kokoro-82M text-to-speech model and is aimed at audiobooks as well as voiceovers for short-form video.

It ships as two interfaces: a PyQt6 desktop app launched with `abogen`, and a Flask web UI launched with `abogen-web` that runs jobs in a background worker. The web UI also includes an optional LLM-based text normalization step and an Audiobookshelf upload integration. Output can be WAV, FLAC, MP3, Opus or M4B with chapters, and subtitles can be SRT or ASS.

The tool supports chapter selection, batch queue processing, a voice mixer for blending voice models, and text files with chapter or timestamp markers. A Docker image and Compose file are included for GPU or CPU deployment.

Key features

  • 01Converts ePub, PDF, text, markdown and subtitle files to audio
  • 02Generates synchronized subtitles in SRT or ASS format
  • 03Exports WAV, FLAC, MP3, Opus or chaptered M4B files
  • 04Includes a voice mixer for combining voice models into custom voices
  • 05Processes multiple files in a batch queue

Use cases

  • Turning an EPUB or PDF book into an audiobook with chapters
  • Producing voiceovers with subtitles for videos
  • Running a self-hosted conversion service via Docker and sending results to Audiobookshelf

Quick start

  1. 1

    Install espeak-ng (Ubuntu/Debian)

    sudo apt install espeak-ng
  2. 2

    Install abogen with uv (Linux, no GPU or NVIDIA)

    uv tool install --python 3.12 abogen
  3. 3

    Start the desktop GUI

    abogen

Things to know

  • Requires installing espeak-ng separately on every platform.
  • Some GPU setups need platform-specific extra steps, and AMD GPU acceleration requires Linux.
  • The Web UI has newer features than the desktop app, which is still being updated.

Topics

#text-to-speech#audiobook#epub#subtitles#kokoro#tts#pdf#python#audiobooks#content-creation

FAQ

What is abogen?
Desktop and web app that converts EPUB, PDF, text and subtitle files into audio with synced subtitles, using Kokoro-82M TTS. Abogen turns ePub, PDF, plain text, markdown and subtitle files into speech audio, generating matching subtitles alongside. It is built on the Kokoro-82M text-to-speech model and is aimed at audiobooks as well as voiceovers for short-form video.
Is abogen free to use? What license does it have?
abogen is open source under the MIT license. Check the license terms before using it in commercial products.
Is abogen still maintained?
Actively maintained. The last commit to abogen was on Oct 17, 2026.
What can I use abogen for?
Turning an EPUB or PDF book into an audiobook with chapters Producing voiceovers with subtitles for videos Running a self-hosted conversion service via Docker and sending results to Audiobookshelf
How do I install abogen?
Install espeak-ng (Ubuntu/Debian): sudo apt install espeak-ng Install abogen with uv (Linux, no GPU or NVIDIA): uv tool install --python 3.12 abogen Start the desktop GUI: abogen
Where can I get abogen?
The source code and installation instructions are on GitHub: https://github.com/denizsafak/abogen

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects