Drop your source. We'll handle the rest.
No choosing what to buy before you've shown us your paper. Drop the source first — we'll read it, suggest what'll work best, and only then ask for the card.
What works beautifully
- arXiv PDFs · Anthropic / OpenAI essays
- Substack, LessWrong, personal blogs (full text)
- Book chapter PDFs · whitepapers · industry reports
- Long-form articles · ≥ 1,500 words
- All PDFs · ≤ 25 MB
What we'll send back (with a hint)
- Google Scholar pages → paste the [PDF] link
- Amazon · Goodreads book pages → upload the PDF
- Paywalled or login-gated articles → upload the PDF
- Twitter threads · YouTube · podcast pages
- Aggregator threads (HN, Reddit) → open the linked source
Still being studied. We're brewing toward these — submit them anyway and we'll write when v2 lands.
- Video · audio · podcasts
- Fiction · novels · poetry
- Scanned or handwritten PDFs (OCR)
- Multi-source synthesis (multiple papers → one card set)
- Books over 600 pages (long-form distillation)