Finrandojin

A multi-voice AI audiobook generator built on Qwen3-TTS — annotate scripts with an LLM, assign unique voices to each character, per-line style instructions for delivery, clone voices from reference audio, design new voices from text descriptions, train custom voices with LoRA fine-tuning, and export to MP3 or Audacity multi-track projects

336
35
100% credibility
Found Feb 07, 2026 at 13 stars 26x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Alexandria is a browser-based application that converts text novels into complete audiobooks by using AI to annotate scripts, assign and generate character voices, and compile the final audio.

How It Works

1
🔍 Find the audiobook maker

You hear about Alexandria, a simple tool that turns any book into a lively spoken story with different voices for each character.

2
📱 Install with one click

Using a helper app called Pinokio, you paste a link and click install – everything sets up automatically on your computer.

3
🌐 Open the web playground

Click start to launch a colorful browser window where you control everything with easy buttons and tabs.

4
🔧 Link your storytellers

Tell it where your AI friends live (like a local storyteller or online helper) so it can think and speak.

5
📖 Upload your book

Drop in your text file, hit generate, and watch the AI break it into scenes with speakers, words, and feelings – pure magic!

6
Pick voices for characters
👤
Use built-in voices

Select fun voices like calm narrator or excited hero, add a style like whispery or bold.

🎤
Clone a voice

Upload a 10-second clip of someone talking and type what they said to copy their unique sound.

7
✏️ Tweak and create audio

Review each line in the editor, preview sounds, regenerate bits, then batch-make the whole audio with natural pauses.

🎉 Enjoy your audiobook

Download the full MP3 story or pro tracks for editing – now anyone can listen to your book come alive!

Sign up to see the full architecture

6 more

Sign Up Free

Star Growth

See how this repo grew from 13 to 336 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is alexandria-audiobook?

Alexandria-audiobook is a Python-based AI tool that converts plain text books into full multi-voice audiobooks, using LLMs to annotate scripts with per-line directions for emotion, tone, and delivery. Upload a novel, let it parse dialogue and narration into JSON chunks with speaker tags, then generate audio via built-in Qwen3-TTS—no external servers needed. Output a polished MP3 or Audacity-ready tracks with natural pauses, perfect for creating alexandria quartet audiobook-style productions or harry potter multi voice audiobook experiences.

Why is it gaining traction?

It stands out with granular per-line control over voice delivery in multi-voice audiobooks, plus voice cloning from 5-second clips, on-the-fly voice design from descriptions, and LoRA training for custom identities—all in a browser UI with batch processing at 3-6x real-time speeds. Pinokio one-click install skips dependency hell, and Audacity export zips per-speaker WAVs with labels for pro editing. Developers dig the REST API for scripting alexandria bellefleur audiobook or alexandria house audiobooks workflows.

Who should use this?

Indie authors turning novels into audiobooks without hiring narrators, podcasters needing quick multi-character scripts, or hobbyists experimenting with best multi voice audiobooks like alexandria of africa audiobook. Voice actors or AI tinkerers cloning refs for alexandria audiobook projects will love the editor for tweaking LLM-annotated lines before batch export.

Verdict

Grab it if you have a beefy GPU (16GB+ VRAM) and want local control over audiobook generation—docs and wiki are solid for a 30-star repo. At 1.0% credibility, it's early and rough-edged, but Pinokio makes it dead simple to spin up; test on short texts first.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.