From paper to AI in three steps<\/h2>\n1. Scan the pages<\/h3>\n
Open Scan on your phone and point it at the page. It finds the edges of the document live and straightens the shot for you. Two modes:<\/p>\n
- \n
- Document<\/strong> \u2014 for a single page or handout.<\/li>\n
- Book<\/strong> \u2014 for an open book. It detects the two-page spread and splits it into two clean pages automatically<\/strong>, so you can scan a whole chapter spread-by-spread without stopping.<\/li>\n<\/ul>\n
Choose an enhancement \u2014 Original, Auto, Gray, or a high-contrast Scan<\/strong> look \u2014 flip on the torch in dim light, and tap the shutter. Captured pages stack up; you can reorder them, re-crop edges, or rotate any page before saving.<\/p>\n
2. It reads the text<\/h3>\n
When you hit Save & learn<\/strong>, on-device OCR reads every page and builds a searchable PDF<\/strong> \u2014 the scanned image with an invisible, correctly-positioned text layer underneath. That means the text is real and selectable, nothing was retyped, and the words are ready for AI to understand.<\/p>\n
3. Chat with it and search it<\/h3>\n
The finished document is added to your chosen knowledge base<\/strong> and learned like any other source. Once it’s ready, you can:<\/p>\n
- \n
- Chat with it<\/strong> \u2014 ask questions and get answers grounded in the scanned pages, with citations.<\/li>\n
- Search it by meaning<\/strong> \u2014 the scanned book shows up in semantic Search, so you can find exact passages across it and the rest of your library.<\/li>\n<\/ul>\n
A physical book you couldn’t search five minutes ago is now an AI you can interrogate.<\/p>\n
Who it’s for<\/h2>\n
- \n
- Students<\/strong> \u2014 scan textbook chapters and lecture handouts, then ask the material questions before an exam.<\/li>\n
- Researchers<\/strong> \u2014 pull an out-of-print book or archival document into a searchable knowledge base.<\/li>\n
- Anyone with a shelf of paper<\/strong> \u2014 manuals, recipe books, printed reports \u2014 turn them into something you can actually ask.<\/li>\n<\/ul>\n
Built to be quick and clean<\/h2>\n
- \n
- Book mode<\/strong> splits spreads for you \u2014 scan a chapter without splitting pages by hand.<\/li>\n
- Live edge detection<\/strong> and re-croppable corners keep every page straight.<\/li>\n
- On-device OCR<\/strong> means the text is real and selectable, not a flat photo.<\/li>\n
- Everything lands in the same knowledge base<\/strong> as your other sources, so chat and search work across all of it together.<\/li>\n<\/ul>\n
Try it<\/h2>\n
- Live edge detection<\/strong> and re-croppable corners keep every page straight.<\/li>\n
- Researchers<\/strong> \u2014 pull an out-of-print book or archival document into a searchable knowledge base.<\/li>\n
- Search it by meaning<\/strong> \u2014 the scanned book shows up in semantic Search, so you can find exact passages across it and the rest of your library.<\/li>\n<\/ul>\n
- Book<\/strong> \u2014 for an open book. It detects the two-page spread and splits it into two clean pages automatically<\/strong>, so you can scan a whole chapter spread-by-spread without stopping.<\/li>\n<\/ul>\n