RelicRELIC
Copying text from images and PDFs

How to copy text from an image, screenshot or PDF

Text trapped in a photo, a screenshot or a flattened PDF feels stuck, but it almost never is. The tool that frees it is called OCR, and it is already built into the phone and computer you own. Here is how to use it on every device.

Jordan Gibbs June 29, 2026 8 min readUpdated July 1, 2026
ImageOCREditable text
Optical character recognition reads the words inside an image and turns them into selectable, copyable text.

You have a receipt photo with a confirmation number on it, a screenshot of a recipe, or a scanned page where the words just will not highlight. The instinct is to retype it by hand. You do not have to. Every modern phone and computer can read the text inside a picture and hand it back to you as real, copyable text, usually in under a second and without sending the image anywhere. This is the overview guide for the whole topic. The deeper, device-specific walkthroughs are linked below as you need them.

Why the text will not select in the first place

A normal document, like a web page or a Word file, stores text as actual characters, so your cursor can grab them. A photo or a scan stores text as pixels that happen to look like letters. There is nothing underneath for the cursor to select, which is why a scanned PDF looks perfectly readable yet refuses to highlight. The job, then, is to turn those pixels back into characters.

The technology that does this is optical character recognition, or OCR. It studies the shapes in an image, recognises them as letters and words, and rebuilds editable text. It used to be slow, clunky and inaccurate. Today it runs right on your device, fast and quietly, which is why the feature has shown up everywhere without much fanfare.

On an iPhone or iPad

Apple calls its version Live Text, and it works on photos, screenshots and what you see through the camera. Open a picture in the Photos app, then press and hold on any text in it. After a moment the words become selectable, with the familiar grab handles. Drag to choose what you want, then tap Copy. You can also tap the small Live Text button (lines of text inside a frame) in the bottom corner to select everything at once.

It works live too. Point the Camera app at a sign, a label or a page, and when the Live Text button appears, tap it to capture the words without taking a photo at all. If you are mainly working from screen grabs, the focused walkthrough is how to extract text from a screenshot.

On an Android phone

Android leans on Google Lens, which is built into Google Photos and the Google app, and on Circle to Search on newer phones. Open a picture in Google Photos and tap the Lens icon, or take a screenshot and tap the Lens or search option that appears. Lens scans the image, then lets you tap Select text, drag over the part you want, and choose Copy. On phones with Circle to Search, press and hold the home button or navigation bar, then draw around the text to pull it out.

The result lands on your clipboard like any other copy, ready to paste into a note, a message or an email. For the version aimed specifically at photos rather than documents, see how to copy text from a picture.

On Windows

Windows has two strong options. The built-in Snipping Tool can recognise text: take a snip, then look for the text actions button in the toolbar, which lets you copy all the recognised text or just the part you select. The other is PowerToys Text Extractor, a free Microsoft utility. Once installed, press its shortcut (by default Win+Shift+T), drag a box over any text on screen, and it goes straight to your clipboard. Text Extractor is the faster choice when you do this often, because it works over anything visible, including a video frame or a window that blocks selection.

For PDFs specifically

Most PDFs are not images at all, and you can select and copy from them like any document. The trouble starts with scanned or locked files. Those need OCR or a small workaround, which is its own topic worth doing properly. The full method, including the awkward password-protected and read-only cases, is in how to copy text from a PDF, with the trickier files covered in how to copy text from a scanned or locked PDF.

A quick test for any PDF: try to drag-select a line of text. If the highlight appears, it is a real text PDF and you can copy normally. If nothing highlights, it is an image, and you need OCR.

On a Mac

macOS has Live Text too, and it behaves like the iPhone version. Open an image in Preview or Quick Look, then click and drag across the text as if it were a normal document. When the cursor turns into a text selector, you can highlight and press Cmd+C to copy. It also works on PDFs that are really scans, and on stills you pull from a video. No extra app, no upload, just select and copy.

When the text comes out wrong

OCR is excellent with clean, printed text and weaker with everything else. If the result is garbled, a few things usually help:

  • Get more resolution. Recognition works far better on a large, sharp image than a tiny or blurry one. Zoom in before you screenshot, or take a closer photo, rather than cropping a tiny region afterwards.
  • Straighten the shot. Text photographed at an angle or curved on a page throws the recognition off. Square it up and flatten the page where you can.
  • Improve the contrast. Dark text on a light background reads best. Glare, shadows and busy backgrounds all cost accuracy.
  • Proofread numbers and symbols. Letters tend to come out right, but a 0 can become an O, and a 1 an l. Always check codes, amounts and punctuation by eye before you trust them.

Handwriting is the hardest case. Neat printing has a fair chance, but cursive and messy notes are still hit and miss across every tool, so treat anything you get from handwriting as a draft to correct rather than a finished copy.

The one habit that makes this pay off

Pulling text out of an image is only half the value. The other half is being able to find it again. People screenshot things precisely because they matter later, the booking code, the address, the quote, and then the screenshot vanishes into a camera roll where nothing is searchable. The words are right there in the picture, but your device treats them as a flat image and forgets them.

That is the gap worth closing. If your tools read the text inside every image you save and keep it searchable, a screenshot stops being a dead end and becomes something you can actually look up. Copy it once, recognise it once, and the contents are yours to find whenever you need them, with no retyping and no scrolling back through the camera roll.

Frequently asked questions

What is OCR, in plain terms?

OCR stands for optical character recognition. It is software that looks at a picture of text, like a photo or a scan, and works out which letters and words are there, then hands you back real, selectable text. Modern phones and computers run it on the device in a fraction of a second, which is why you can now tap text inside a photo and copy it.

Why can I see the text in a PDF but not select it?

Because that PDF is really an image. Scanned documents and exported screenshots are flattened to a picture, so there are no actual characters underneath, just pixels that look like letters. Your cursor has nothing to grab. The fix is to run OCR over the page, which rebuilds the text from the image so you can select and copy it.

Is copying text from an image accurate?

It is very good now, especially with clear, printed text. Clean screenshots and typed documents come out almost perfect. Accuracy drops with handwriting, low resolution, odd fonts, busy backgrounds, or photos taken at an angle. Always glance over the result for stray characters, and watch numbers and punctuation, which are the usual places small errors hide.

Does the image need to be online for this to work?

No. On an iPhone, Android phone, or a recent Windows or Mac machine, the text recognition runs locally on the device. The picture does not get uploaded anywhere, which is faster and keeps the contents private. Web-based OCR tools do send your image to a server, so for anything sensitive, prefer the built-in tools on your own device.

Written by
Jordan GibbsFounder, Relic

Jordan Gibbs is the founder of Relic, an end-to-end encrypted, permanent, searchable memory for everything you copy. He writes widely about AI, agents, and practical tooling on Medium, where he is read by tens of thousands, and builds privacy-first software. Here he covers how everyday tools like the clipboard actually work, and how to use them without handing your data to someone else.

MediumGitHubLinkedIn
Part of
Copying text from images and PDFs
Keep reading
6 min

How to copy text from a PDF

A normal PDF lets you select and copy like any document. A scanned or locked one needs a different approach. Both kinds handled, on desktop and phone.

Read
6 min

How to copy text from a picture

Your phone can already read the words inside a photo, you just have to know where to tap. Copying text from a picture on iPhone, Android, Windows and Mac.

Read
6 min

How to copy text from a scanned or locked PDF

A scanned PDF is really a photo of a page, and a locked one blocks copy on purpose. Pulling the text out of each, using OCR and a couple of safe workarounds.

Read
ocrhow-toproductivity