OCR

Drop in a photo or screenshot and get its text back verbatim, structured as Markdown — ready to copy and paste. Nothing is stored: the image goes straight to the model and is discarded once the text comes back.