How to Copy Text from Screenshots on Mac: OCR Methods Explained

The screenshot has the text you need. Now what?

You screenshot an error message from your browser. You screenshot a code snippet from a tutorial video. You screenshot a Slack message with a stack trace. The text is right there in the image, but you can't select it, copy it, or paste it anywhere. You end up retyping it by hand — and introducing typos along the way.

OCR (optical character recognition) solves this by extracting selectable text from images. macOS has built-in OCR that most people don't know about, and there are third-party tools that go further. This guide covers every method available on Mac in 2026, from zero-install options to dedicated OCR apps.

Method 1: Live Text in Preview (built-in, no setup)

Since macOS Monterey (12.0), Apple has included Live Text recognition across the system. The simplest way to use it: open any screenshot in Preview, hover over text in the image, and your cursor changes to a text selection cursor. Click and drag to select the text, then Cmd+C to copy.

Live Text in Preview works well for:

• Clean, high-contrast text (error dialogs, terminal output, web pages)
• Standard fonts at readable sizes
• Single-language text (English and most Latin-script languages)

It struggles with:

• Small text or low-resolution images
• Text with complex backgrounds or overlapping UI elements
• Code with syntax highlighting (it sometimes misreads colored characters)
• Handwritten text or stylized fonts

The key limitation: you have to open the file in Preview first. If you're working fast and just captured a screenshot to clipboard, Live Text in Preview requires saving the file, opening it, selecting the text, copying, then switching back. That's five steps when you just wanted the text.

Method 2: Live Text in Quick Look

A faster version of the Preview method. Select a screenshot file in Finder and press Space to open Quick Look. If Live Text detects readable text in the image, you can select and copy it directly from the Quick Look preview without fully opening the file.

This saves one step compared to Preview and is the fastest built-in method when you already have a screenshot saved as a file. It uses the same recognition engine, so the accuracy is identical.

Method 3: Live Text system-wide (macOS Ventura and later)

Starting with macOS Ventura, Live Text works in more places than just Preview and Quick Look. You can select text in images displayed in Safari, Messages, Photos, and some third-party apps that use Apple's standard image views. If you see an image with text in any of these apps, try hovering over the text — the selection cursor may appear.

macOS Sequoia expanded this further. You can now right-click an image in many contexts and see a "Copy All Text" option that extracts every line of text from the image at once, without having to manually select regions.

Method 4: Shortcuts app for batch OCR

If you regularly need to extract text from screenshots, you can build a Shortcut that automates the process. The Shortcuts app includes an "Extract Text from Image" action that uses Apple's OCR engine.

A useful workflow:

1. Open the Shortcuts app and create a new shortcut
2. Add "Receive Input" set to accept images
3. Add "Extract Text from Image"
4. Add "Copy to Clipboard"
5. Save and assign a keyboard shortcut in System Settings > Keyboard > Keyboard Shortcuts > App Shortcuts

Now you can right-click any image file, choose Quick Actions, and run your shortcut to extract text in one step. Or trigger it with the keyboard shortcut you assigned. This is significantly faster than the Preview method for repetitive use.

Method 5: Third-party OCR tools

When the built-in OCR isn't enough, third-party tools offer better accuracy, more features, and faster workflows.

Shottr includes OCR as part of its screenshot tool. Capture a screenshot with Shottr, then use its text recognition to select and copy text from the capture. It handles code and terminal output better than Apple's Live Text because it's optimized for developer-oriented content.

TextSniper is a dedicated OCR app for Mac. Press a keyboard shortcut, select a region of your screen, and the text is instantly copied to your clipboard. No screenshot file is created — it goes directly from screen to clipboard as text. It supports QR codes and barcodes too.

macOS's built-in screencapture + Shortcuts can approximate the TextSniper workflow without a third-party app: capture to clipboard, run the "Extract Text from Image" shortcut on the clipboard contents, and the extracted text replaces the image in your clipboard. It requires some initial setup but works without installing anything.

Comparison: which method to use

OCR tips for better results

Capture at the highest resolution possible. Retina screenshots (2x) give OCR engines twice the pixel data to work with. If you're capturing from a non-Retina display or a scaled resolution, text recognition accuracy drops noticeably on small text.

Crop before running OCR. The less noise in the image, the better the recognition. A tightly cropped screenshot of just the error message will give better results than a full-screen capture where the error is a small portion of the image.

Watch out for syntax highlighting. Code editors display text in multiple colors with varying font weights. OCR engines sometimes misread characters when the color contrast between text and background varies within the same line. If you need to extract code, consider switching your editor to a high-contrast theme or capturing from a plain text view.

Dark mode can reduce accuracy. Light text on dark backgrounds is harder for some OCR engines than dark text on light backgrounds. If you're getting poor results from a dark mode screenshot, try inverting the colors in Preview (Tools > Adjust Color, or use a Shortcut with color inversion) before running OCR.

When you need the screenshot and the text

OCR extracts text from screenshots, but it discards the visual context. Sometimes you need both — the screenshot for showing what you see, and the ability to reference specific text within it. This is especially common when working with AI coding assistants: you want to show the error and have the AI read the exact error message.

LazyScreenshots captures your screen and auto-pastes it directly into Claude, Cursor, ChatGPT, or any other app — in one shortcut. The AI assistant reads the text from the image using its own vision capabilities, so you get the visual context and the text recognition without any extra steps. No OCR tool needed, no clipboard juggling, no file management.