Classic OCR is built for printed type and turns handwriting into gibberish. Here an AI model reads the whole line the way a person would, so joined-up writing, block capitals and scribbles in the margin come out as words.
Each page costs 1 credit, and you can add up to 10 photos at once — a whole set of lecture notes, say. Pages it couldn't read aren't charged.
The model doesn't invent words it can't make out: it marks them [?], so you know exactly what to check against the original.
