OpenReaderFormats
Native EPUB and PDF, kept intact
Your file renders in a built-in EPUB and PDF reader, never flattened to plain text. Layout-aware parsing (PP-DocLayoutV3, ONNX) follows the true reading order, even in dense multi-column PDFs.
Open-source document reader
Every word, lit as it’s spoken.
OpenReader turns EPUB, PDF, TXT, Markdown, and DOCX into synchronized read-along audio, highlighting every word on your original page as it’s spoken. Multilingual, open source, self-hostable.
How it works
From file to voice in three steps.
Upload a document
Drop an EPUB, PDF, TXT, Markdown, or DOCX into your library, or import one straight from the server. It stays organized for every session after.
Pick a voice
Hosted OpenAI, Replicate, or DeepInfra, or your own Kokoro, KittenTTS, or Orpheus server. Set the language, choose a compatible voice, and tune the speed.
Read, listen, export
Follow word-level highlighting on the original page as it plays, resume on any device, and export an audiobook for the road.
Under the hood
Made for deep reading.
Alignment
Word-by-word timing
ONNX Whisper alignment through a JetStream-backed compute worker maps each spoken word back to the page, so the cursor tracks speech precisely.
Languages
Multilingual narration
Choose a document language for language-aware narration, highlighting, and compatible voice selection.
Export
Audiobook output
Render m4b and mp3 files with resumable processing, ready for any offline player you already use.
Sync
Progress that follows you
Reading position, queue state, and per-document settings sync across sessions and devices through your account.
Backend
A stack you actually control
Embedded SeaweedFS or any S3-compatible bucket, SQLite or Postgres, Docker on amd64 or arm64, with built-in auth and automatic startup migrations.
Formats & voices
Five formats in. Any voice out.
Read EPUB, PDF, TXT, MD, DOCX natively. Narrate with hosted providers or a speech server you run yourself: Kokoro, KittenTTS, Orpheus, OpenAI, Replicate, DeepInfra.
macOS · Windows · Linux
The whole reader, no server needed.
OpenReader Desktop bundles everything this site runs, from local Kokoro voices to word alignment and audiobook export, in one app that keeps your library on your computer and updates itself.
iPhone · iPad · Mac · Beta
OpenReader for Apple devices, entirely on-device.
Several voice models and voice cloning, powered by Apple’s Core AI. No account, no analytics, no data collected.
Open source · MIT
Host it yourself.
Deploy for yourself or your team in minutes. The docs cover Docker, provider integration, object storage, and the external compute worker.
# run the maintained Compose example
$ git clone https://github.com/richardr1126/openreader.git
$ cd openreader
# add first-admin credentials to .env
$ docker compose -f examples/docker/compose.yml up -d
# open the reading room
$ open localhost:3003