Image to Text (OCR)

Why on-device OCR is the point

Think about what people actually run through an OCR tool: an ID card, a receipt, a bank letter, a medical note, a page of a contract, a screenshot of a private message. Nearly every free OCR site uploads that image to its servers to process it. Here the engine is downloaded into your browser and the picture stays on your machine. There is no account, nothing is stored, and nothing is transmitted.

Thirteen languages — pick the right one

English, Spanish, French, German, Italian, Portuguese, Dutch, Russian, Chinese (Simplified), Japanese, Korean, Arabic and Hindi are all supported. Choose before you run it, because the tool works in one language at a time and the choice makes a real difference — an Arabic document read with the English model returns nonsense, not slightly worse text. If a page mixes two languages, pick the one most of the text is in and correct the rest by hand.

Getting a good result

OCR accuracy depends almost entirely on the image. A flat, sharp, evenly lit photo of printed text is read very well. Things that degrade it: shadows across the page, a photo taken at an angle so the lines curve, low resolution, busy backgrounds, and text over a picture. If a result is poor, re-photographing the page straight on in better light helps more than anything the tool can do.

Two limits worth knowing. Handwriting is not reliably recognised — this engine is built for printed text, and cursive in particular will not work. And layout is not preserved: you get the words, not the columns, tables or formatting they sat in.

Copy it or download it

The recognised text appears in a box with a character and word count, so you can read it before doing anything. Copy it to the clipboard, or download it as a .txt file. A progress percentage runs while it works, which is useful because OCR on a large image takes a few seconds — and the first run takes longer, since the engine and your chosen language data are downloaded then. Later runs in the same language start immediately.

Related tools

Crop Image is worth using first on a photo where only part of the frame has text — cropping to the text region improves accuracy and speed. Extract Text from PDF pulls text straight out of a PDF that already has real text in it, which is faster and more accurate than OCR when the PDF is not a scan.