v4.12.0 now available

Copy any text
you can see.

A lightning-fast OCR utility for Windows. Extract text from anywhere on your screen — instantly. The full experience, with the latest OCR models and local AI, lives on the Microsoft Store.

4.7k
GitHub Stars
4.6/5
Store Rating
v4.12.0
Latest

Three steps to any text

No setup. No accounts. No cloud. Just the text you need, right now.

Press your hotkey

Hit your configured shortcut from anywhere in Windows — no need to switch apps.

Select a region

Draw a box around any text on screen — a photo, video, app, PDF, anything.

Text in clipboard

The recognized text lands instantly in your clipboard, ready to paste anywhere.

Built for every workflow

From quick one-off grabs to power-user editing — Text Grab has a mode for it.

Fullscreen Grab

Draw a region around the text you need and it's in your clipboard instantly — or click a single word to grab just that. Works on any app, browser, game, PDF, or video. It's also the basis of the PowerToys Text Extractor.

Grab Frame

Float a transparent overlay on top of any window. Text updates live as content changes, with built-in search so you can find exactly what you need.

Edit Text Window

Turn raw OCR into clean, usable content. Work in plain-text, markdown, or spreadsheet mode with regex extraction, find & replace, case conversion, a live Calc Pane, and batch image scanning.

Quick Simple Lookup

Your personal hotkey-activated text snippet dictionary. Store frequently used phrases, codes, or templates and paste them in a flash.

Tools that work together

Text Grab doesn't stop at capture. Patterns, position, and automation turn one good grab into a repeatable workflow — so you can capture text, find what matters, clean it up, and move it to the next step without switching apps.

Capture Clean Extract Structure Reuse
Multiple capture engines

Pick the engine that fits your machine and source: Windows AI on the NPU of a Copilot+ PC, the built-in Windows OCR, open-source Tesseract, or Direct Text that reads native UI elements with no OCR at all.

Post-grab actions

Clean and route text the instant it's captured. Trim each line, remove duplicates, fix GUID formatting, insert it into the active app, or open a web search — then save the combination as a reusable template.

Patterns & RegEx

Find and extract structured values — IDs, dates, prices, emails, phone numbers — even when the surrounding text is messy. Select a result and Text Grab suggests a pattern you can tune and save for next time.

Grab Templates

Combine named capture regions, pattern placeholders, and output formatting into one reusable extraction. Run it on every similar screenshot, PDF, form, or invoice instead of rebuilding the steps each time.

Calc Pane

Evaluate math, functions, unit conversions, and date math line by line right beside your text — and get sum, average, median, or min/max straight from spreadsheet selections, no second calculator needed.

Bulk folder OCR

Point Text Grab at a folder and extract text from every image or PDF in a single pass — recursively, with templates applied — then finish the combined results in the Edit Text Window.

Grab tables & lists — with formatting intact

Copying structured data off a web page usually means a mangled mess. The Text Grab browser extension makes pulling tables and lists feel effortless and natural — the rows, columns, and structure come along for the ride, so it pastes exactly the way it looked. Works on any Chromium browser, on any operating system.

Tables stay tables. Lists stay lists.

Select a table or list and it lands on your clipboard with its formatting preserved — ready to paste as a real table, with the rows, columns, and merged cells (colspan / rowspan) intact. The extension writes both polished HTML and tab-separated text, so it drops in clean wherever you need it. No reformatting, no cleanup, no copy-paste gymnastics.

Works great on its own
Straight into your spreadsheet

No desktop app required. Grab a table or list right in your browser and paste it directly into Excel, Google Sheets, LibreOffice Calc, Notion, or your docs — formatting and all. The extension alone is a complete copy-anything-structured tool.

Even faster with Text Grab
Hand off to the desktop app

Pair it with the Text Grab desktop app and capturing data gets even quicker — send a region or image straight into spreadsheet mode, run patterns and templates, OCR what isn't selectable, and pull from PDFs. The browser and desktop work as one fast pipeline.

Select Region

Draw a box on any page and pick how to capture it: Table pulls structured data straight out (even detecting repeating layouts and grids), Direct Text copies the underlying text, and Screenshot sends it to Text Grab for OCR with scrolling and stitching for tall regions.

PDFs & images

An integrated pdf.js viewer adds region capture to PDFs, and a right-click option sends any image to Text Grab for OCR — so even tables and text locked inside graphics become selectable.

Send to Text Grab

One click hands a capture off to the desktop app for further refinement — patterns, templates, and the full Edit Text Window. Built with vanilla JS and isolated-world injection, so it even works on strict-CSP sites.

Free & open source · works on any Chromium browser (Chrome, Edge, Brave, Vivaldi, Arc and more) on any OS. View the source on GitHub →

The right tool for the job

Designed from the ground up for Windows power users who value speed, privacy, and simplicity.

Private by default

All OCR runs locally via the Windows OCR API. No cloud processing, no data sent anywhere, ever. Your screen contents stay on your machine.

Blazing fast

From hotkey to clipboard in under a second. Zero startup time, zero friction. Integrates invisibly into your existing workflow.

AI-ready

Choose the right capture engine for the job — Windows AI on a Copilot+ PC's NPU for the highest accuracy, the built-in Windows OCR, Tesseract, or Direct Text. Translation and local AI tools are exclusive to the Microsoft Store version.

Open source

The source code is fully open on GitHub — audit it, fork it, or contribute. A free build is available for developers. The full-featured release with latest OCR and AI is on the Microsoft Store.

Built around how you work

However you wrangle text — collecting it, cleaning it, digging through logs, or racing the clock — there's a way Text Grab fits into your day.

The Information Worker

Bouncing between tools all day

Lives between a dozen browser tabs, a spreadsheet, and a chat window. The browser extension pulls tables and lists straight off the page with their formatting intact, and Text Grab cleans up whatever comes along for the ride — so the data you collect lands ready to use instead of needing a reformat.

  • Browser extension
  • Grab tables & lists
  • Edit Text Window
The Researcher

Turning stacks of images into data

Has folders full of scanned pages, screenshots, and figures that all need to become searchable text. Point Text Grab at the folder and bulk OCR runs the whole set in a single pass — entirely on-device, so sensitive source material never leaves the machine.

  • Bulk folder OCR
  • On-device & private
  • Choice of engines
The IT Pro

Finding the signal in the logs

Scans walls of log output and config dumps for the one GUID, IP address, or error code that actually matters. Patterns & RegEx pull the structured values out of the noise, and Quick Simple Lookup keeps commands, connection strings, and boilerplate a single hotkey away.

  • Patterns & RegEx
  • Quick Simple Lookup
  • Direct Text

What sets Text Grab apart

Plenty of tools can pull text off your screen. Few do it this fast, this privately, or with this much control.

Feature Text Grab PowerToys
Text Extractor
Windows
Snipping Tool
Cloud OCR
apps
Runs fully offline, on-device
Choice of OCR engines
Windows AI, WinRT, Tesseract
Edit & clean up captured text in place
Grab tables & lists with formatting intact
Reads QR codes
Browser extension included
Open source

Comparison reflects typical default behavior; third-party tools vary. Got a tool we should add? Let us know.

How do I get this on my computer?

Pick the track that matches your machine. Whether it's your own PC or a locked-down work laptop, there's a path that fits.

Your work PC

On a managed or locked-down machine? Try these in order — stop at the first one your IT policy allows.

If the Microsoft Store is allowed

Install it from the Store just like any other approved app. You get the full experience with automatic updates, and nothing extra to configure.

Open in Microsoft Store
If you can run signed apps

Download the digitally signed build and run it — the Authenticode signature is what SmartScreen and AppLocker check, so no admin install is required on many managed PCs.

Get the signed build
If everything's locked down

Ask for a signed .msix package your IT team can review and deploy through their managed channels. Reach out and Joe will get one to you.

Request a signed MSIX

Frequently asked questions

Everything you might be wondering before you grab your first bit of text.

Does Text Grab send my data off device?

No. Text Grab runs completely locally and works fully offline. OCR is performed on your machine using the built-in Windows OCR API (or Windows AI on a Copilot+ PC) — nothing about your screen or captured text is ever uploaded to a server.

What do I need to run it?

Text Grab is a Windows desktop app and runs on Windows 10 and Windows 11. The browser extension works on any Chromium browser — Chrome, Edge, Brave, Vivaldi, Arc and more — on any operating system.

How accurate is the text recognition?

Very accurate for clear, reasonably sized text. You can pick the engine that fits the job — Windows AI on a Copilot+ PC's NPU for the highest accuracy, the built-in Windows OCR, Tesseract, or Direct Text. Results depend on the resolution and clarity of what's on screen, so zooming in a little helps with small fonts.

Which OCR models does Text Grab use?

Text Grab picks the best available engine for your hardware, and you can switch between them anytime:

  • Windows AI models — the newest, most accurate option, running on the NPU of supported Copilot+ PCs for top-tier recognition.
  • Windows OCR (WinRT) models — the built-in OCR available on all Windows 10 and 11 devices. Fast, reliable, and the default when Windows AI isn't supported.
  • Tesseract — an optional open-source engine you can add if you want it, useful for extra languages and specialized recognition.

Whichever model runs, everything happens locally on your device — no captured text or images are sent anywhere.

Can it grab text from images, PDFs, and videos?

Yes. If you can see it on your screen, Text Grab can grab it — including text inside images, PDFs, paused videos, error dialogs, and other apps that normally won't let you select text.

Does it work with languages other than English?

Yes. Text Grab uses the OCR language packs installed on your system, so it recognizes many languages. You can add more languages through Windows settings. Translation tools are also available in the Microsoft Store version.

Is Text Grab open source?

Completely. The full source code is available on GitHub — you're welcome to audit it, fork it, or contribute. The browser extension's source is open too.

How do I report a bug or request a feature?

Open an issue on GitHub, or email joe@textgrab.net. Feedback and feature ideas are always welcome.