OL Reader

Listen to your own EPUB library with on-device neural text-to-speech.

Getting started

1 · Add books

OL Reader scans the app's Documents folder for .epub files. The bundled sample Pride and Prejudice appears on first launch. To add more, use the Files app to copy .epub files into On My iPhone → OL Reader.

2 · Pick a voice

Tap the person icon in the controls bar to choose between Apple's built-in voices (instant) and the on-device Kokoro neural voices (~30 s warmup on first use, then real-time after).

3 · Tap to read

Tap any paragraph to start playing from there. The "Starting…" indicator shows briefly while the next chunk is generated. Lock-screen / Control Center controls work too.

Frequently asked questions

Does OL Reader send my books anywhere?

No. All text-to-speech happens entirely on your device using Apple's MLX framework. The app makes no network requests for narration or content.

Why does the first Kokoro chunk take a moment?

The neural model is ~330 MB and loads from disk on first use, after which the model stays warm and subsequent chunks generate in a few seconds. The "Starting…" indicator shows whenever audio isn't immediately available.

Audio stopped when my screen locked — what happened?

OL Reader is configured for background audio. If audio stopped, please make sure Settings → OL Reader → Background App Refresh is enabled, and that nothing else (a phone call, Siri, or another media app) interrupted playback.

How are books stored?

EPUBs you add live in the app's Documents folder. Reading position is saved per book in UserDefaults. Nothing is uploaded.

It can't read my book / a chunk sounds wrong

Please email the address below with the book title and a screenshot of the affected paragraph. Most issues are caused by EPUBs with unusual encoding or embedded scripts.

Contact

Questions, bug reports, feature requests:
darren.oakey@gmail.com