Save articles from URLs or paste text directly. Read them in a clean, distraction-free view. Listen to them with AI-generated text-to-speech.

• Save from URL — extracts article content using Mozilla Readability
• Paste text — supports markdown and plain text
• Text-to-Speech — generates audio using Kokoro-82M (runs locally, no API keys)
• Code blocks — syntax highlighted with Shiki
• Search, filter, sort — find articles by title, filter by read status or favourites

• Docker and Docker Compose
• pnpm (if running outside Docker)

docker compose up -d --build

Or via pnpm:

pnpm docker:start

Open localhost:3000. PostgreSQL, the Next.js app, and TTS all run from this single command.

cp .env.example .env
docker compose up -d postgres       # start just the database
pnpm install
pnpm prisma generate
pnpm prisma migrate deploy
pnpm dev

1. User submits a URL or pastes text
2. URLs are fetched, cleaned (popup removal, paywall detection), and parsed with Readability
3. Content is split into typed sections (paragraphs, headings, code blocks, tables, lists, images, videos, blockquotes)
4. Sections are stored as JSON in PostgreSQL
5. TTS spawns a Node.js child process that runs Kokoro-82M locally — the ONNX model (~92MB) auto-downloads on first use
6. Audio is saved as WAV and served statically with sentence-level alignment for highlighted playback

Set these environment variables to customize TTS (all optional):