Product · 2026 · In development
Papers
Listen to research papers, read aloud with the citations handled and the math in words.
- Next.js
- TypeScript
- SQLite
- LLMs
- Text-to-speech
- Kind
- Product
- Year
- 2026
- Status
- In development
Add an arXiv id, a PDF, or a link. Papers turns it into a read-aloud script, reads it with a natural voice, and shows a clean reader that highlights the sentence you are hearing.
The pipeline
- Ingest. arXiv HTML, web articles, or PDF pages read by a vision model become one structured document.
- Script. A language model rewrites it for the ear: citations and references handled, equations spoken in words.
- Voice. One audio clip per sentence, generated in order, so playback starts before the whole paper is done.
- Read along. A shared library with progress and highlights for each listener.