openreader/README.md

📄🔊 OpenReader WebUI

OpenReader WebUI is an open source text to speech document reader web app built using Next.js, offering a TTS read along experience with narration for EPUB, PDF, TXT, MD, and DOCX documents. It supports multiple TTS providers including OpenAI, Deepinfra, and custom OpenAI-compatible endpoints like Kokoro-FastAPI and Orpheus-FastAPI

• 🧠 (New) Smart Sentence-Aware Narration: EPUB and PDF playback use shared NLP (compromise) and smart sentence continuation to merge sentences that span pages/chapters for smoother TTS trying to prevent hard cuts at page breaks
• 🎧 (New) Reliable Audiobook Export: Create and export audiobooks from PDF and EPUB files (in m4b or mp3 format using ffmpeg) with resumable, chapter/page-based export and per-chapter regeneration
• 🎯 (New) Multi-Provider TTS Support:
  • Deepinfra: Kokoro-82M, Orpheus-3B, Sesame-1B models with extensive voice libraries
  • OpenAI API (): tts-1, tts-1-hd, gpt-4o-mini-tts models
  • Kokoro-FastAPI: Self-hosted OpenAI-compatible TTS API server supporting Kokoro-82M and multi-voice combinations (like af_heart+bf_emma)
  • Orpheus-FastAPI: Self-hosted OpenAI-compatible TTS API server supporting Orpheus-3B
  • And other Custom OpenAI-compatible endpoints with a /v1/audio/voices endpoint
• 🚀 (New) Optimized TTS Pipeline: Next.js TTS backend with in-memory LRU audio cache, ETag-aware responses, and in-flight request de-duplication for faster repeat playback
• 💾 Local-First Architecture: IndexedDB browser storage for documents and settings (now using Dexie.js)
• 🛜 Optional Server-side documents: Manually upload documents to the Next.js backend (and Docker docstore) for all users to download
• 📖 Read Along Experience: Follow along with highlighted text as the TTS narrates PDF files, with per-sentence navigation and skip controls
• 📄 Document formats: EPUB, PDF, TXT, MD, DOCX (with libreoffice installed, plus hardened DOCX→PDF conversion for better reliability)
• 🎨 Customizable Experience:
  • 🔑 Select TTS provider (OpenAI, Deepinfra, or Custom OpenAI-compatible)
  • 🔐 Set TTS API base URL and optional API key
  • 🎨 Multiple app theme options
  • And more...

🆕 What's New in v1.0.0

• 🧠 Smart sentence continuation
  • EPUB and PDF playback now use smarter sentence splitting and continuation metadata so sentences that cross page/chapter boundaries are merged before hitting the TTS API.
  • This yields more natural narration and fewer awkward pauses when a sentence spans multiple pages or EPUB spine items
• 🎧 Chapter/page-based audiobook export with resume & regeneration
  • Per-chapter/per-page generation to disk with persistent bookId
  • Resumable generation (can cancel and continue later)
  • Per-chapter regeneration & deletion
  • Final combined M4B or MP3 download with embedded chapter metadata.
• 💾 Dexie-backed local storage & sync
  • All document types (PDF, EPUB, TXT/MD-as-HTML) and config are stored via a unified Dexie layer on top of IndexedDB.
  • Document lists use live Dexie queries (no manual refresh needed), and server sync now correctly includes text/markdown documents as part of the library backup.
• 🗣️ Kokoro multi-voice selection & utilities
  • Kokoro models now support multi-voice combination, with provider-aware limits and helpers (not supported on OpenAI or Deepinfra)
• ⚡ Faster, more efficient TTS backend proxy
  • In-memory LRU caching for audio responses with configurable size/TTL
  • ETag support (304 on cache hits) + X-Cache headers (HIT / MISS / INFLIGHT)
• 📄 More robust DOCX → PDF conversion
  • DOCX conversion now uses isolated per-job LibreOffice profiles and temp directories, polls for a stable output file size, and aggressively cleans up temp files.
  • This reduces cross-job interference and flakiness when converting multiple DOCX files in parallel.
• ♿ Accessibility & layout improvements
  • Dialogs and folder toggles expose proper roles and ARIA attributes.
  • PDF/EPUB/HTML readers use a full-height app shell with a sticky bottom TTS bar, improved scrollbars, and refined focus styles.
• ✅ End-to-end Playwright test suite with TTS mocks
  • Deterministic TTS responses in tests via a reusable Playwright route mock.
  • Coverage for accessibility, upload, navigation, folder management, deletion flows, and playback across all document types.

🐳 Docker Quick Start

Prerequisites

• Recent version of Docker installed on your machine
• A TTS API server (Kokoro-FastAPI, Orpheus-FastAPI, Deepinfra, OpenAI, etc.) running and accessible

Note: If you have good hardware, you can run Kokoro-FastAPI with Docker locally (see below).

1. 🐳 Start the Docker container:

docker run --name openreader-webui \
  --restart unless-stopped \
  -p 3003:3003 \
  -v openreader_docstore:/app/docstore \
  ghcr.io/richardr1126/openreader-webui:latest

(Optionally): Set the TTS API_BASE URL and/or API_KEY to be default for all devices

docker run --name openreader-webui \
  --restart unless-stopped \
  -e API_KEY=none \
  -e API_BASE=http://host.docker.internal:8880/v1 \
  -p 3003:3003 \
  -v openreader_docstore:/app/docstore \
  ghcr.io/richardr1126/openreader-webui:latest

Note: Requesting audio from the TTS API happens on the Next.js server not the client. So the base URL for the TTS API should be accessible and relative to the Next.js server. If it is in a Docker you may need to use host.docker.internal to access the host machine, instead of localhost.

Visit http://localhost:3003 to run the app and set your settings.

Note: The openreader_docstore volume is used to store server-side documents. You can mount a local directory instead. Or remove it if you don't need server-side documents.

2. ⚙️ Configure the app settings in the UI:

• Set the TTS Provider and Model in the Settings modal
• Set the TTS API Base URL and API Key if needed (more secure to set in env vars)
• Select your model's voice from the dropdown (voices try to be fetched from TTS Provider API)

3. ⬆️ Updating Docker Image

docker stop openreader-webui && \
docker rm openreader-webui && \
docker pull ghcr.io/richardr1126/openreader-webui:latest

🗣️ Local Kokoro-FastAPI Quick-start (CPU or GPU)

You can run the Kokoro TTS API server directly with Docker. We are not responsible for issues with Kokoro-FastAPI. For best performance, use an NVIDIA GPU (for GPU version) or Apple Silicon (for CPU version).

Note: When using these, set the API_BASE env var to http://host.docker.internal:8880/v1 or http://kokoro-tts:8880/v1. You can also use the example docker-compose.yml in examples/docker-compose.yml if you prefer Docker Compose.

Docker CPU

docker run -d \
  --name kokoro-tts \
  --restart unless-stopped \
  -p 8880:8880 \
  -e ONNX_NUM_THREADS=8 \
  -e ONNX_INTER_OP_THREADS=4 \
  -e ONNX_EXECUTION_MODE=parallel \
  -e ONNX_OPTIMIZATION_LEVEL=all \
  -e ONNX_MEMORY_PATTERN=true \
  -e ONNX_ARENA_EXTEND_STRATEGY=kNextPowerOfTwo \
  -e API_LOG_LEVEL=DEBUG \
  ghcr.io/remsky/kokoro-fastapi-cpu:v0.2.4

Docker GPU

docker run -d \
  --name kokoro-tts \
  --gpus all \
  --user 1001:1001 \
  --restart unless-stopped \
  -p 8880:8880 \
  -e USE_GPU=true \
  -e PYTHONUNBUFFERED=1 \
  -e API_LOG_LEVEL=DEBUG \
  ghcr.io/remsky/kokoro-fastapi-gpu:v0.2.4

Note:

• These commands are for running the Kokoro TTS API server only. For issues or support, see the Kokoro-FastAPI repository.
• The GPU version requires NVIDIA Docker support and works best with NVIDIA GPUs. The CPU version works best on Apple Silicon or modern x86 CPUs.
• Adjust environment variables as needed for your hardware and use case.

Local Development Installation

Prerequisites

• Node.js (recommended: use nvm)
• pnpm (recommended) or npm

  npm install -g pnpm
  

• A TTS API server (Kokoro-FastAPI, Orpheus-FastAPI, Deepinfra, OpenAI, etc.) running and accessible Optionally required for different features:
• FFmpeg (required for audiobook m4b creation only)

  brew install ffmpeg
  

• libreoffice (required for DOCX files)

  brew install libreoffice
  

Steps

1. Clone the repository:

   git clone https://github.com/richardr1126/OpenReader-WebUI.git
   cd OpenReader-WebUI
   

2. Install dependencies:

   With pnpm (recommended):

   pnpm i # or npm i
   

3. Configure the environment:

   cp template.env .env
   # Edit .env with your configuration settings
   

   Note: The base URL for the TTS API should be accessible and relative to the Next.js server

4. Start the development server:

   With pnpm (recommended):

   pnpm dev # or npm run dev
   

   or build and run the production server:

   With pnpm:

   pnpm build # or npm run build
   pnpm start # or npm start
   

   Visit http://localhost:3003 to run the app.

💡 Feature requests

For feature requests or ideas you have for the project, please use the Discussions tab.

🙋‍♂️ Support and issues

If you encounter issues, please open an issue on GitHub following the template (which is very light).

👥 Contributing

Contributions are welcome! Fork the repository and submit a pull request with your changes.

❤️ Acknowledgements

This project would not be possible without standing on the shoulders of these giants:

• Kokoro-82M model
• Orpheus-TTS model
• Kokoro-FastAPI
• Orpheus-FastAPI
• react-pdf npm package
• react-reader npm package

Docker Supported Architectures

• linux/amd64 (x86_64)
• linux/arm64 (Apple Silicon, Raspberry Pi, SBCs, etc.)

Stack

• Framework: Next.js (React)
• Containerization: Docker
• Storage: Dexie + IndexedDB (in-browser local database)
• PDF:
  • react-pdf
  • pdf.js
• EPUB:
  • react-reader
  • epubjs
• Markdown/Text:
  • react-markdown
  • remark-gfm
• UI:
  • Tailwind CSS
  • Headless UI
  • @tailwindcss/typography
• TTS: (tested on)
  • Deepinfra API (Kokoro-82M, Orpheus-3B, Sesame-1B)
  • Kokoro FastAPI TTS
  • Orpheus FastAPI TTS
• NLP: compromise NLP library for sentence splitting

License

This project is licensed under the MIT License.