HushBook: Read while you listen

HushBook provides a word-by-word read-along experience for audiobooks, highlighting each spoken word while the narration plays. It is designed for readers who want to follow along with the audio and keep the focus on the text, with the transcription generated on the device. The project emphasizes privacy, as there is no cloud or account required and everything stays on the phone. The app supports importing user-owned files and free public-domain titles from LibriVox and the Internet Archive, offering offline use after setup.

The HushBook Engine performs transcription locally on the device, turning spoken words into synchronized text. It operates without uploading data, with audio, transcripts, notes, and scores stored locally. The system does not require sign-in and works offline after a one-time engine download, with a design that accounts for heat and long-running transcriptions by processing in chunks when needed.

Users interact with a built-in read-along reader that lights up each word in sync with narration and allows tapping a word to jump to that position. The app emphasizes private by design transcription and provides a library of public-domain titles accessible offline, alongside user-imported files.

HushBook focuses on accessibility features, including reading profiles that adapt the UI for comprehension, dyslexia, and low-vision reading. It includes fonts, overlays, contrast, and pacing controls, along with a private analytics suite that stays on-device. The project promotes a habit-friendly approach with streaks and goals to encourage daily reading and listening together.

Use cases

• Read along with owned audiobooks on iOS and Android, offline
• Import and transcribe personal audio files