optical character recognition
Technology that turns images of text — scans, photos, PDFs — into editable, searchable data.
Optical character recognition, or OCR, is the technology that reads text from images — a scanned page, a photo of a receipt, a PDF — and turns it into digital text a computer can search and process. It has existed for decades, but modern AI has made it far more capable.
Today’s AI-powered OCR does more than copy out words. The best models return a structured map of a document: where each block sits, whether it is a title, table or signature, and how confident the model is. Some can read dozens of pages in a single pass.
OCR is one of AI’s most useful and least hyped applications for business: turning invoices, contracts and archives into structured data. Crucially, several strong OCR models can run locally, so sensitive documents never have to leave your own systems.