Text recognition
Extract text from screenshots or image pins, check and edit OCR results, copy all or selected text, and adjust layout, language, and window behavior.
When you cannot select text in an image, use text recognition to turn it into editable, copyable text. Check the result against the original image before copying it to a document, chat, or another app.
Start from a screenshot or image pin
Extract text while taking a screenshot
Press Ctrl+1⌃ Control+1 to start a screenshot and select the area containing the text, keeping lines and paragraphs complete where possible.
Click Ocr and copy on the screenshot toolbar, or press Shift+C⇧ Shift+C.
The text recognition window opens by default. Wait for the result, check the text, and click Copy text.
Use an existing image pin
Focus the image pin and press Shift+C⇧ Shift+C. If text has not been recognized, the recognition window opens. If a result is available, the selected text is copied, or all text if nothing is selected. Your settings determine whether a confirmation window appears first.
You can also right-click the pin and choose Recognition → Detect Text. Once a recognition result is available, the menu changes to Copy All Text.
Check against the image, then copy
The window shows Original image on the left and editable Recognition result on the right. Correct mistakes, add missing text, or remove extra content directly. To focus on the text, turn off Show original image at the top.
| What to copy | Action |
|---|---|
| The entire result | Clear the text selection, then click Copy text |
| A short passage | Drag to select the text on the right, then click Copy text |
| A passage in the image | Drag over text on the left; double-click to select its text block, then copy |
You can also press Ctrl+C⌘ Command+C to copy. Selections on the two sides are mutually exclusive: a new selection replaces the previous one. Selecting text on the right highlights matching content in the original image where possible, making it easier to check.
After editing the result, copy the corrected text from the right. Selecting on the left still uses the original recognized content.
By default, copying all text closes the window, while copying a selection keeps it open so you can take several passages.
Adjust paragraphs, spaces, and punctuation
Click Layout at the top to arrange text for your destination document.
| Option | Use |
|---|---|
| Smart paragraphs | Organizes text by recognized paragraphs; the default for general articles |
| Merge line breaks | Joins breaks between recognized lines, reducing sentences split across multiple lines after copying |
| Keep line breaks | Preserves recognized line divisions for lists and line-by-line material |
| Remove spaces | Removes spaces, tabs, and other horizontal whitespace while keeping line breaks; off by default |
| Punctuation | Keeps punctuation unchanged or converts it to English or CJK punctuation |
| Vertical text order | Sets right-to-left or left-to-right reading order for vertical text |
Leave Remove spaces off when you need spaces between English words. Text size changes only the display size in the window, not the copied text. Highlight selection in image controls the corresponding highlight in the original image.
Changing paragraphs, spaces, punctuation, or vertical text order regenerates text from the original recognition result. Set the layout before proofreading. If it overwrites manual edits, press Ctrl+Z⌘ Command+Z in the right-hand editor to undo the replacement.
Choose a recognition language and retry if needed
Recognition language at the bottom selects the language of the text in the image. Changing it recognizes the current image again. Follow software uses the app's interface language; it does not automatically detect the language in the image.
Uninstalled languages are marked Not installed, select to download. Select one and wait for the language pack to finish downloading before recognition. Canceling the download keeps the previous selection. Changing language in this window applies to the current image. To change the default for later recognition, set the language under System → OCR.
If you see No text detected or Recognition did not complete, the original image is retained. Check that the text is clear, the selection is complete, and the recognition language matches, then use the window's retry action.
Set copying and closing behavior
Settings at the top contains Close after copying all text, Close after copying a selection, and Copy text when closing. Copying on close is off by default; keep that setting when you want to check each result.
When recognizing a screenshot, enable Copy directly next time at the bottom. Later operations of the same kind skip the window and copy text as soon as recognition finishes. Selected text and all text from pins each remember this preference separately.
To restore the confirmation window, open System → OCR → Show dialog when and reselect the corresponding Screenshot, Copy selected text, or Copy all text option.
Related articles
Eraser
Erase parts of annotations with a pencil or rectangle, adjust width, enable rectangle editing, and distinguish undoable erasing from clearing all marks.
Table recognition
Convert tables in screenshots or image pins into editable cells, check and sort rows, copy all or selected cells, and save XLSX or CSV files.