How to Copy Text From a Photo or Screenshot (Without Retyping a Single Word)

You can see the text — a receipt, a slide, a page of a book, a screenshot — but you can't select it, because it's a picture. Here's how to pull real, editable text out of any image in about 20 seconds, no retyping required.

The Text You Can See but Can't Touch

You have a photo of a whiteboard after a meeting. A screenshot of a paragraph someone sent you. A picture of a recipe card, a Wi-Fi password taped to a router, a slide from a webinar, a page of a textbook. The words are right there, perfectly legible. You just need to copy a few lines of them into an email, a doc, or a search bar.

So you try to select the text — and nothing happens. No blinking cursor, no highlight. Your finger or mouse drags uselessly across the image. That's because, as far as your computer is concerned, there is no text in that file at all. There's just a grid of colored dots that *happens* to look like words to your eyes.

The usual fallback is to squint at the image and retype everything by hand, which is slow, error-prone, and genuinely one of the most tedious things a computer should have made obsolete decades ago. The good news: it did. The process is called OCR, it takes about twenty seconds, and you never have to retype a word.

Why You Can't Select Text in an Image

A photo or screenshot is a raster image — a rectangle of pixels. When you point your camera at a page or hit the screenshot key, your device records what the words *look* like, not what they *say*. The letter "A" isn't stored as the character A; it's stored as a few hundred dark pixels arranged in a triangle-ish shape. Beautiful to a human, meaningless to a text cursor.

This is the same reason you can't select the text in a scanned PDF — a scan is just a photo of a page wrapped in a PDF. Until something actually *reads* the shapes and figures out which letters they represent, the words stay locked inside the picture.

That "something" is OCR.

What OCR Actually Does

OCR stands for Optical Character Recognition. It's software that looks at an image, finds the shapes that look like letters and numbers, and converts them into real, editable text characters you can copy, paste, edit, and search.

Modern OCR is dramatically better than the flaky scanner software you might remember from the 2000s. It handles:

Printed text of almost any font, in dozens of languages

Photos taken at an angle or with uneven lighting

Screenshots of apps, slides, chat messages, and error dialogs

Multi-column layouts, keeping paragraphs roughly in order

Mixed content — pulling the text out and ignoring the logos and photos around it

What it's *not* magic at: messy handwriting, heavily stylized display fonts, extremely low-resolution images, and text photographed so blurry that even you have to guess at it. More on that below.

The 20-Second Fix

Here's the whole process using the Image to Text tool:

Upload the image. Drag in your JPG, PNG, screenshot, or photo. No account, no app install.
Let it read. The tool scans the image and extracts every character it can recognize.
Copy the text. You get back plain, selectable text. Copy the part you need, or grab all of it.
Paste it anywhere. Email, Google Docs, a spreadsheet, a search bar — it's real text now.

That's it. A receipt becomes a line you can paste into your expense sheet. A slide becomes notes. A photographed paragraph becomes a quote you can actually quote.

Getting the Image Into the Right Shape First

OCR works best on a clean, standard image file. A couple of quick prep steps dramatically improve the result:

If it's an iPhone photo, it's probably a HEIC. Some OCR tools choke on Apple's HEIC format. Run it through HEIC to JPG first and you'll have a universally readable image.

If it's a giant PNG screenshot, shrink it to JPG. A full-resolution PNG isn't more readable to OCR — it's just slower to upload. Converting with PNG to JPG keeps the legibility and drops the file size.

Crop to just the text. The less clutter in the frame, the fewer mistakes OCR makes. Crop out the background, the margins, and anything that isn't the words you want.

What If It's a Scanned PDF, Not an Image?

Half the time, the "image" you're fighting with is actually a PDF — a scanned contract, a faxed form, a document someone photographed and saved as a PDF. The words still aren't selectable, for the exact same reason.

The fix is the same idea, pointed at a PDF. Run the file through PDF to Text to pull out the raw words, or PDF to Word if you want an editable document that keeps the layout. Both read the page and hand you back real, editable text instead of a frozen picture.

When OCR Gets It Wrong (and How to Help It)

OCR is good, not clairvoyant. When the output comes back garbled, it's almost always one of these:

The image is too low-resolution. If the text is only a few dozen pixels tall, there simply isn't enough detail to tell an "e" from a "c." Take a new photo closer up, or find a higher-resolution version of the image.

The photo is blurry or badly lit. Motion blur and deep shadows are OCR's kryptonite. Retake the shot with steady hands and even lighting — natural daylight on a flat surface beats a dim desk lamp at an angle every time.

The font is decorative. Wedding-invitation script, heavy display fonts, and distorted logo text confuse the recognizer. Standard body text reads nearly perfectly; fancy headlines may not.

It's handwriting. Neat printing sometimes works; cursive and doctor-scrawl usually don't. For handwriting, expect to proofread.

In every case, the move is the same: give OCR a clearer image. A sharp, well-lit, straight-on photo of printed text will come back almost flawless. Always skim the result for the classic mix-ups — "0" vs "O", "1" vs "l", "rn" read as "m" — before you rely on it.

A Few Things This Quietly Solves

Once you know OCR is a 20-second step and not a chore, a lot of small annoyances disappear:

Copying a Wi-Fi password off the back of a router instead of spelling it out loud

Turning a photographed slide deck into searchable notes

Grabbing a quote from a book without transcribing the whole page

Pulling the line items off a paper receipt into a spreadsheet

Getting the exact wording of an error message you screenshotted, so you can search it

Any time you catch yourself about to retype what's in a picture, that's the signal to run it through OCR instead.

Bottom Line

The reason you can't select text in a photo or screenshot isn't a glitch — it's that the file is a picture of words, not the words themselves. Your device recorded what the text looked like, never what it said.

OCR closes that gap. It reads the shapes, figures out the letters, and hands you back real, editable, copy-pasteable text. Feed it a clean, sharp image — convert a HEIC to JPG or shrink a PNG to JPG first if you need to — drop it into Image to Text, and copy what comes out.

The next time the words you need are trapped inside a picture, don't reach for your keyboard. Reach for OCR, and let the computer do the typing it should have been doing all along.