Copy Text from Images with PowerToys Text Extractor

A line of text inside an image or video can be hard to reuse when you can’t select it normally. PowerToys Text Extractor uses optical character recognition (OCR) to read a chosen part of the screen and copy the result to your clipboard.

This guide uses Microsoft documentation revisions from 2025, before 26 March 2026. It is a documentation-based walkthrough, not a test of OCR accuracy on every app or language.

When Text Extractor is useful

Use Text Extractor when the words are visible but ordinary selection and copy do not work—for example, a label inside an image, a line in a video frame, or text in an app that offers no copy command. If the original text is selectable, using the app’s normal copy command is usually simpler. A quick choice: copy selectable text in its app, use Text Extractor for visible words inside an image or video, and use Snipping Tool when you need a screenshot.

Text Extractor reads text from a screen region; it does not decide whether the result is accurate or trustworthy. The recognized words are sent to the clipboard, so check them before you paste or act on them. This matters for names, numbers, web addresses and commands, where one misread character can change the meaning.

If your goal is to save a screenshot, use Snipping Tool instead. Microsoft’s Text Extractor guidance recommends Snipping Tool for capturing screenshots; Text Extractor’s job is to copy recognized text.

Install PowerToys and check the utility

If PowerToys is not installed, use Microsoft Store or the official PowerToys releases on GitHub. Microsoft’s installation guide also documents WinGet for people who prefer a command line: winget install --id Microsoft.PowerToys --source winget. Avoid third-party download sites.

After installation, open PowerToys Settings and make sure Text Extractor is enabled. The Home view in Settings can turn individual utilities on or off. Open the Text Extractor settings if you need to change its activation shortcut or preferred OCR language.

Capture and copy text

1. Bring the image, video frame or other content to the front so the text is visible. Pause a video on the frame you want to read.
2. Press Windows key + Shift + T. This is Text Extractor’s default shortcut. If you changed it, use the shortcut shown in Text Extractor settings.
3. Click and hold the primary mouse button, then drag a rectangle around the words you need. Keep the selection tight enough to avoid unrelated labels, but include the full lines you want.
4. Release the mouse button. Text Extractor recognizes the selection and copies the result to the clipboard. The capture mode closes after recognition; press Esc to close it sooner.
5. Paste with Ctrl + V into the destination app. Read through the result before sending, saving or running it.

If the selection is in the wrong place, hold Shift while moving it. Releasing Shift lets you resize the rectangle again. This can help when the overlay is close to the right area but covers a neighboring line.

Fix a shortcut or language problem

If Windows key + Shift + T does nothing, first check that PowerToys is running and Text Extractor is enabled in Settings. Then look at the activation shortcut in Text Extractor settings. It is configurable, so you can choose another key combination if the default is unavailable on your PC.

If the tool opens but returns no text or the wrong characters, check the preferred language in its settings. The default follows Windows’ language and keyboard settings. Text Extractor can recognize only languages with the corresponding Windows OCR language pack installed. Microsoft’s documentation explains how to check the available OCR languages and add a pack; the example uses Windows PowerShell and administrator access, so follow those steps only if you are comfortable making that system change.

OCR is a best effort. A result can contain missing punctuation, lookalike letters or incorrect digits. Select the words again if the region was too large or missed an edge, then proofread the output. Don’t assume a command, URL, account number or other important string is correct just because it looks plausible.

For a dense page, try a paragraph or one column at a time instead of selecting the whole screen. If the text is difficult to read at its current size, zoom in using the source app before you capture it. These are practical ways to make the selection easier to inspect; they do not guarantee a perfect OCR result. When the formatting matters, paste into a plain-text editor and compare the line order with the original before reusing it.

What it does—and what it doesn’t

Text Extractor is a quick bridge from visible words to editable text. It can save retyping when the source is an image, video or app window, and its keyboard shortcut avoids opening a separate OCR editor. The resulting text goes to your clipboard for pasting.

It is not a translation tool, a screenshot saver or a source of verification. If you need a picture of the screen, use Snipping Tool. If you need to rely on the recognized text, compare it with the original before you use it.

Sources and further reading

• Microsoft Learn, PowerToys Text Extractor utility — source revision from 28 August 2025; covers the shortcut, selection, clipboard, settings and OCR language packs.
• Microsoft Learn, Install PowerToys on Windows 10 and Windows 11 — source revision from 19 November 2025; covers the Microsoft Store, GitHub releases, WinGet and utility toggles in Settings.