Get the Text Out of Any Photo or PDF — Free OCR That Works Offline

Free OCR on real pages: a crooked scan read at 87% confidence, a Spanish notice read as English (91%, accents wrong) and then as Spanish (95%, every accent right), a read with the internet switched off, and PDF to text on a text PDF and on a scanned one. Nothing is uploaded.

Getting the text out of a photo or a PDF, on real pages. A crooked scan of the plan’s first page reads at 87% confidence — the tool scores every read so you know how carefully to check it, and tables are always the hard part. A Spanish shop notice read with the language left on English comes out at 91% with the accents wrong (the ñ in "añade" becomes "fi"); set the language to Spanish and the same page reads at 95% with every accent right. Pick the language the page is written in: there are 54. The language data is fetched on the first read, and after that the reading happens on your computer — the video cuts the internet and reads again, at 95%. PDF to text reads a PDF’s own text layer exactly, table included, and when a scanned PDF has no text layer it reads the page like a photo instead. Nothing is uploaded.

Image Tool Hub — 200+ free tools for images, video, audio, PDFs and Office documents, in English, Spanish, French, German, Portuguese and Indonesian. Everything runs on your device.