Summarize a PDF with real page citations
The PDF's text layer is extracted in your browser with its page map; the AI must cite only pages that actually exist. Scanned PDFs without a text layer are reported honestly — not OCR'd in this first version.
How to use it
- Sign in, then add one text-based PDF (first 20 pages are used).
- Optionally name a focus, then run the task (1 trial).
- Check the cited pages against the document, edit, and download the notes as Markdown.
Limits & formats
Text-based PDFs up to 30 MB; the first 20 pages per task in this first version. Empty or image-only pages are skipped and counted. Citations are constrained to real page numbers, but re-check anything you rely on — an AI summary can still misread emphasis. Password-protected files are not supported.
Privacy
Text extracted in your browser is sent through YYSI's server to our AI provider (DeepSeek) as the task input. YYSI stores page counts, token usage and the summary — not your file. Results are kept for 24 hours, then become unavailable.
Related tools
Feedback on this tool
Did it work well? Any improvement ideas? Leave a note — it helps YYSI improve the toolbox.
Notes are anonymous and reviewed manually: only you see yours right away (in this browser); others see them after approval. Do not include personal data or contact details.