Your first library in ten minutes

1. Point SCL at one folder of PDFs. It scans in place — nothing is moved, renamed or touched.
2. Watch the topic view appear. Your papers, organised by what they are about. Papers that are in the prepared corpus (3,000,000 and counting — most of arXiv) light up immediately with paraphrases, links and analysis.
3. Search inside everything. One box, full text, contents not filenames.
4. Open a paper. Read the paraphrase at the depth you want — Expert is the default — ask it questions in the Q&A panel, and follow the cross-links to related papers you already hold.
5. Queue the rest. Papers not yet in the corpus can be selected for processing; you’ll be told when each is ready.

Choosing your AI

• Local model — zero Privacy Gap, the recommended default.
• Your own key (OpenAI / Anthropic / Google, or local via Ollama / LM Studio) with a hard monthly cap you set.
• Anonymised access — reach a frontier model with your content stripped before anything leaves. Reduced exposure, not zero — we explain exactly what leaves.

If something breaks

Tell us: /contact. During the beta you will usually hear back from the founder the same day.