Docs · guide

What is OCR, and what it can and cannot read

By Alberto Gulotta · Updated 7 September 2026 · 10 min read

What is ocr? The letters stand for optical character recognition, and it is the step that turns a photograph of a page into something a computer can read. It is the hidden requirement behind half the complaints in this subject: the converter that produced gibberish, the PDF that will not search, the form that could not be filled.

It is also, quietly, one of the great free wins of the last few years. The recognition built into an ordinary phone is now good enough for most everyday jobs, costs nothing, and never sends the page anywhere.

The words people use track how much they already know. Image to text is what somebody types the first time, usually holding a phone. Ocr pdf and pdf ocr come from people who have met a scanned document and understood that it is a picture. Pdf to text converter means they want a plain file rather than a document, and searchable pdf is the most precise of all — that person knows exactly what is missing. Ocr software is a purchase decision, and text recognition is the name of the field itself.

Copying text on a Windows PC

Two of the routes above are Apple’s and one is Google’s, and the gap that leaves is Windows. On Windows the answer is not a program you install: it is two things already sitting on the machine, and which of them you want depends on one question. Is the text on the screen in front of you right now, or is it inside a picture you have already put into a note?

1. Put the text on screen and press Windows logo key + Shift + S. That is Microsoft’s own shortcut for the capture overlay, given on the Snipping Tool page for Windows 11 and Windows 10. It does not matter what the words are sitting in — a PDF that refuses to select, a paused video, a photograph, a program with no copy command anywhere in its menus.
2. Drag a box round the part you want. Only what you enclose is captured, so a single paragraph out of a crowded page is a better starting point than the whole window.
3. Select Text actions. In Microsoft’s words: “Once you’ve captured a snip, select the Text actions button to activate the Optical Character Recognition (OCR) feature.”
4. Take the words out. Microsoft again: “From here, you have the option to either select and copy specific text or use the tools to Copy all text or to Quick redact any email addresses or phone numbers in the snip.” The redaction is worth remembering — it is the fastest way to share a screenshot without the address in it.

One sentence on that page matters more than the procedure, and it is the reason this route is the right default for anything private. Microsoft states plainly that “All text recognition processes are performed locally on your device.” A payslip, a passport page or a letter from a solicitor never leaves the computer, which is the opposite of what happens when the same picture is dropped into a converter on the web.

The second route is for a picture that is already inside a document rather than on the screen, and it is in a program almost nobody thinks of as an OCR tool. Microsoft: “OneNote supports Optical Character Recognition (OCR), a tool that lets you copy text from a picture or file printout and paste it in your notes so you can make changes to the words.” The whole operation is two clicks — “Right-click the picture, and click Copy Text from Picture”, then “Click where you’d like to paste the copied text, and then press Ctrl+V.” Microsoft lists it for OneNote for Microsoft 365, OneNote 2024, OneNote 2021 and OneNote 2016.

What OneNote can do that the capture route cannot is a stack of pages in one command. When the picture is part of a file printout — a PDF sent to OneNote rather than pasted in — the same menu offers two commands instead of one: “Copy Text from this Page of the Printout” for the image you clicked, and “Copy Text from All the Pages of the Printout” for every page at once. A forty-page scan becomes forty pages of text without opening any of them.

Two warnings come from the same page, and both are worth knowing before you decide the feature is broken. The command does not always appear straight away: it is missing while OneNote is still reading the image, and Microsoft adds that “Some results may take up to 24-48 hours to become available.” The other is the sentence that ends the page, and it is the same warning this whole subject rests on — “The effectiveness of Optical Character Recognition depends on the quality of the image you’re working with.”

So: the capture route for anything you can see, because it takes about six seconds and the recognition happens on your own machine; OneNote when the material is already filed, or when there is more of it than you want to handle one screen at a time. Neither costs anything, neither needs an account beyond the one you already have, and neither uploads the page.

The guides below cover what is already on your devices, how to make a scanned PDF properly searchable, the specific things recognition still gets wrong, and the cases — handwriting, bad light, unusual alphabets — where expectations need adjusting.

Not covered here. It will not tell you recognition is accurate. It is a transcription made by a machine that cannot read, and on numbers it is wrong often enough to matter. Every guide here ends by telling you what to check.

It will not rank the ocr software yet. That verdict needs the same set of pages — clean, awkward and terrible — run through each program with the results published, which is a test.

And it will not send a private document to a website when the phone in your pocket does the job offline. On this guide more than any other, the free built-in answer is also the private one. What holds instead is simple: the recognition on your own phone is free, offline, and enough for most of this guide.