Screenshot to Text

Turn a screenshot to text: select an area to extract text from images and PDFs with on-device OCR, and the text is copied for you.

Copy text you can see but cannot select. Screenshot to Text uses OCR to turn an image to text: select an area of your browser tab and read the result in a small editable card. Select, read, paste • Press Alt+Shift+O or click the extension's toolbar icon, then drag over the text. • The recognized text is copied automatically by default. Turn this off in Settings if you prefer. • Check the result in the card, correct any mistakes, and press Copy again after editing. • Right-click an image and choose “Copy text from image” to read its visible part. If the image cannot be identified directly, select the area yourself. Read what is on screen Use it on text in images, scanned documents, web pages and e-book readers. It also captures selected areas of PDFs open in Chrome's browser viewer, including scanned PDFs. If the page cannot show the selection overlay, a captured view opens in a separate tab for selection. You can select text on a paused video frame too, but subtitles and complex backgrounds can produce mistakes. It reads the visible area you select, not a whole document or a scrolling page. Edit and choose line breaks The card lets you edit recognized text before copying it again. Paragraph lines are joined by default; switch on “Original lines” to keep line breaks as they appear in the image. Tables remain plain text, not spreadsheet cells. Drag the card by its header to move it out of the way. One or two recognition languages Choose the language of the text in the card or in Settings. Use “+ Add language” for text containing two languages. When you choose a non-Latin script and a slot is available, English is added automatically; you can remove it. These are recognition languages: the extension does not translate text. Changing languages reads the same selection again. Your previous result stays available while it works. After a language change, press Copy to copy the new result; it is not copied automatically. Successful language choices are remembered for later captures. On-device recognition, offline when the language is available Text recognition (OCR) runs on your computer. The extension does not upload your screenshot or recognized text. Eight languages are bundled and work offline immediately. Other languages download on demand from screentotext.com, with cdn.jsdelivr.net as a public fallback, and work offline while their downloaded files remain cached. A first download needs an internet connection. Download servers receive your IP address and the requested language, not your screen content. If a language needs downloading, the card shows progress. You can cancel a download, then change or remove a language in the card. No account is required, and there are no capture quotas or credits. For clearer results Small text is automatically enlarged for another recognition pass when suitable. This does not recover missing detail: zoom in before selecting tiny or blurry text, and always check the result. Handwriting, decorative fonts, slides, multiple columns, subtitles and complex layouts may need substantial corrections. Accuracy varies by language and image quality. Chrome's own pages and the Chrome Web Store cannot be read by extensions. To read files opened from your computer, open chrome://extensions, click “Details” on Screenshot to Text and turn on “Allow access to file URLs”. When access is blocked, the extension shows these steps with a button that opens the page. Press Esc to cancel a selection or a read, dismiss the language picker, or close the card. You can change the capture shortcut at chrome://extensions/shortcuts.