Most ”free audiobook makers” are trials: a few thousand characters, then a subscription. This one is free the way open source is free — you run the pipeline on your own free-tier accounts, and there's simply nobody to bill you.
Audiobook-maker apps pay for TTS compute on your behalf, so they must charge you. This project inverts the model: the pipeline is AGPL open source and runs on your accounts — Cloudflare R2's free 10GB for storage, free-for-personal-use neural voices for synthesis, optional pennies of LLM for summaries. There is no company in the loop with a burn rate.
npm run setup once (two Cloudflare values in .env, one command, ~2 minutes).Chapter-sized episodes with AI one-line summaries; an intro episode where two AI hosts discuss the book; automatic language detection and voice switching; whole-book translation if you want to listen in another language; resume positions everywhere via standard podcast RSS.
No. AGPL-3.0 source, your own storage and keys. The only genuinely optional cost is LLM usage for summaries (cents).
You need to run three commands in a terminal once. After that, everything happens in the web player.
The pipeline is AGPL and fine to use; the default free TTS voices are personal-use only — swap in a commercial TTS provider (interface provided) for commercial audio.