Microsoft’s Snipping Tool in Windows 11 can now pull text from screenshots without ever touching the internet. A text-extraction feature called Text actions, highlighted in a late update to Paul Thurrott’s Windows 11 Field Guide on July 12, uses on-device optical character recognition to let you copy words out of error dialogs, PDFs, video frames, and apps that normally block selection. It also has a redaction tool that can scrub email addresses and phone numbers before you paste.
For millions of Windows users, this removes the need to install a separate OCR utility or upload sensitive captures to a cloud service. Everything happens locally, making it safe for IT pros parsing internal build numbers and for anyone who’s ever stared at an error message on a remote desktop session, wondering how to get that text into a ticket.
The feature isn’t exactly new—Microsoft’s documentation confirms it’s been part of Snipping Tool for several Windows 11 releases—but its quiet arrival and privacy-first design have kept it under most radars. Here’s how it works, why it matters, and how to make it part of your daily workflow.
What the Snipping Tool’s Text Actions Actually Do
Open a screenshot in the Snipping Tool, and you’ll spot a button labeled “Text actions.” Click it, and the app scans the image for words, highlighting them in blue. You can then drag to select specific phrases or hit “Copy all text” to grab everything. The extracted text lands on your clipboard, ready to paste into documents, scripts, or search bars.
Behind that button is a local OCR engine that reads whatever is visible in the capture area: system dialogs, video subtitles, command-line output, even handwritten notes if they’re clear enough. Microsoft’s documentation notes that the feature can also redact detected email addresses and phone numbers automatically—a thoughtful safety net when screenshots contain contact information.
This is a departure from older approaches to grabbing on-screen text. Before Text actions, you might have used a phone camera to snap a photo of an error code and then manually typed it out. Or you might have turned to Microsoft’s own PowerToys utility, which offers a similar feature via the Win+Shift+T shortcut. The Snipping Tool version differs in one critical way: it is fully self-contained. No cloud processing, no network calls, no ambiguity about where your data is going.
For anyone handling internal error messages, server names, or proprietary configuration text, that local-only guarantee removes a compliance headache. And because it’s built into Windows 11 by default, no extra installation is needed.
Why This Matters for Your Privacy and Workflow
The most consequential word in Microsoft’s documentation is “locally.” All recognition runs on your device. That means you can grab text from a screenshot of a confidential document, a password-protected screen, or a medical record and not worry about a third-party service seeing it.
This contrasts sharply with many third-party OCR tools that send images to the cloud for analysis. Even some built-in smartphone features require an internet connection. With the Snipping Tool, you can be completely offline—useful if you’re troubleshooting a network issue and need to copy an IP address or error message from a control panel that doesn’t support highlighting.
Workflow gains are just as concrete. IT administrators who frequently document problems can snip a dialog, extract the text, and paste it into a ticket in seconds. Writers who need to quote text from a locked PDF or an image-heavy whitepaper can bypass the usual reformatting pain. Students can lift equations or citations from lecture slides without retyping.
The redaction feature also solves a real-world problem. Imagine taking a screenshot of a settings screen that includes an email address you didn’t notice. Text actions can automatically black it out before you forward the image. That’s not just convenient—it’s a potential privacy save.
How Text Extraction Made Its Way into the Snipping Tool
The Snipping Tool that shipped with early Windows 11 builds was a modest upgrade over the old Windows 10 version. It gained a modern interface and basic annotation tools but nothing approaching OCR. The real OCR muscle lived in PowerToys, Microsoft’s open-source utility collection. PowerToys’ Text Extractor module, launched in 2021, introduced a Win+Shift+T hotkey that let you select any portion of the screen and instantly copy the recognized text to the clipboard.
That tool was embraced by power users. It worked anywhere—even over videos and games—and could handle multiple languages via Windows OCR language packs. But it required that you download and keep PowerToys updated, something casual users rarely do.
Sometime in late 2024 or early 2025, Microsoft began baking text recognition directly into the Snipping Tool. The feature first appeared as a quiet roll-out in Windows 11 version 23H2, then became more prominent with the 24H2 update. By mid-2026, according to Thurrott’s updated guide, Microsoft was comfortable recommending the Snipping Tool’s version over the PowerToys alternative for most screenshot-based tasks.
That recommendation is telling. It signals that the built-in tool has reached a maturity level where it can serve as the default for everyday OCR needs. PowerToys remains available for niche cases—like when you want to skip the step of saving or opening a screenshot—but the simpler path now lives inside a tool every Windows 11 user already has.
How to Start Using It Today
Getting started takes three steps:
- Capture a screen region. Press Win+Shift+S (or Print Screen if you’ve set that key to open Snipping Tool). Your screen dims and a toolbar appears. Choose the rectangular snip, freeform, window, or full-screen option. For text extraction, a rectangular snip works best because you can tightly frame the text.
- Open the notification. When the capture completes, a toast notification pops up in the lower-right corner. Click it—that opens the snip inside the Snipping Tool editor.
- Activate Text actions. In the editor toolbar, look for the button labeled “Text actions.” Click it. After a short processing pause (usually under a second), recognized text appears highlighted in your image. You can now select and copy specific passages, or use the “Copy all text” button. To redact sensitive content, enable the redaction toggle before selecting text.
A few practical tips:
- Clearer captures yield better results. For error dialogs and command windows, keep the capture area tight around the text. Avoid complex backgrounds like wallpaper or gradients.
- Oversized text or tiny fonts can trip the OCR. If extraction fails, try enlarging the source on screen (by zooming in a browser or scaling a PDF) before capturing.
- Tables and multi-column layouts may jumble the output. In those cases, consider extracting one column at a time.
- The feature respects language settings. It uses the Windows OCR language packs already installed on your system, so German, Japanese, and other supported scripts should work out of the box if the corresponding language is added in Windows Settings.
If you’re moving between many machines, note that the feature is tied to the Snipping Tool version. On Windows 11 23H2 and newer, it should be present by default. On older builds, you might not see the Text actions button. Running Windows Update and ensuring you have the latest Snipping Tool app update from the Microsoft Store is the best way to get it.
When PowerToys Still Wins
For all the convenience of the built-in tool, PowerToys Text Extractor has a couple of remaining advantages. First, it skips the screenshot altogether. You press Win+Shift+T, draw a rectangle around the text you want, and the recognized text lands on your clipboard immediately. No images are saved, no editor opens. That speed is hard to beat when you’re grabbing dozens of serial numbers or log entries in rapid succession.
Second, PowerToys can extract text that the Snipping Tool never sees—such as text inside a live video call, a game HUD, or a DirectX overlay. Since the Snipping Tool works from a static captured image, it misses dynamically generated content that hasn’t been captured yet.
For most scenarios, however, the Snipping Tool’s Text actions are the more approachable choice. The ability to review and edit the extracted text before copying, combined with redaction and an image archive, suits the way many people work. And because it doesn’t require installing anything, it’s a natural pick for less technical users or in locked-down corporate environments where PowerToys might not be approved.
Microsoft’s own recommendation tilts toward the Snipping Tool for screenshot capture—a sensible default from the company that made both tools.
The Fine Print: OCR Isn’t Magic
No OCR engine is flawless, and the one inside Snipping Tool is no exception. Unusual fonts, low-contrast text, heavy JPEG compression, and text laid over patterned backgrounds can produce garbled results. A common pitfall is capturing a window that includes both text and iconography; the OCR may try to interpret icons as characters and insert nonsense symbols.
In testing with common error dialogs, the tool performed reliably with standard system fonts like Segoe UI and Consolas. But screenshots of stylized PDFs or web pages with decorative fonts sometimes dropped letters or swapped similar characters (for example, “0” for “O”). The rule of thumb: always proofread before you paste something into a production script or a command line. A misread “rm -rf /” versus “rm -rf ./” is not a lesson you want to learn the hard way.
Tables and multicolumn layouts are another weak point. The OCR reads left to right, top to bottom, so a two-column PDF might come out as one garbled paragraph. If the layout matters, extract columns individually.
Despite these caveats, for the sort of text most people need—error codes, UI labels, plain paragraphs—the accuracy is high enough to be genuinely useful. And because the recognition happens instantly with no upload delay, you can often re-snip a cleaner capture in seconds if the first pass isn’t perfect.
What’s Next for Screen Capture and OCR in Windows
The integration of local OCR into the Snipping Tool is part of a broader trend at Microsoft: moving smarts into the operating system without always phoning home. The same philosophy powers features like local transcription in Windows Studio Effects and on-device AI models in Copilot+ PCs. Expect future Snipping Tool updates to refine the OCR engine, possibly adding language packs automatically based on your display language settings.
There’s also room for tighter integration with the clipboard manager. Imagine a future where you press Win+V and see the extracted text alongside your image captures, ready to paste without opening the editor. That would further blur the line between PowerToys’ quick-copy model and the Snipping Tool’s screenshot-archive model.
For now, the practical takeaway is simple: the next time you need to copy text from something that doesn’t let you select it, try Win+Shift+S before you reach for your phone or a third-party utility. The tool you already have might just do the job—securely, offline, and without adding another app to your PC.