How to extract text from an image
- Choose the language of the text in your image. English is recognized automatically as well.
- Add your image. Click Choose files, drag images into the box, or simply paste a screenshot with Ctrl+V (⌘+V on Mac).
- Wait a few seconds while the text is recognized. The first run downloads the OCR engine and language data.
- Copy the text or download it as a
.txtfile.
You can add several images at once — each one gets its own result box, newest first.
What is OCR?
OCR (optical character recognition) is the technology that turns pictures of text into real, editable text. Instead of retyping a document, you can copy the words straight out of a photo, scan or screenshot and paste them into an email, a document or a search box.
This tool uses Tesseract, a well-established open-source OCR engine, compiled to run inside your browser with WebAssembly. That’s what makes it possible to recognize text without sending your image to a server.
Getting the best results
OCR accuracy depends mostly on the image you give it:
- Use a sharp image. Blurry or shaky photos are the most common cause of mistakes.
- Photograph documents straight on, with even light and no shadows across the text.
- Crop to the text you need. Removing busy backgrounds and unrelated areas helps — use Crop Image first.
- Make sure the text is upright. Fix sideways photos with Rotate Image.
- Pick the right language. Recognizing Korean text with the English model, for example, gives poor results.
- Higher contrast is better. Dark text on a light background works best.
Small images are automatically enlarged before recognition, because the engine works best when letters are at least 20 pixels tall.
Understanding the confidence score
Each result shows a confidence percentage — the engine’s own estimate of how sure it is about the text. Above 85% usually means clean, reliable output. Below 60% suggests the image is blurry, low contrast or in a different language than the one selected, and you should check the text carefully or try a better image.
Common uses
- Copy text from screenshots of apps, videos or social media posts
- Digitize printed documents, letters and receipts
- Grab a quote or reference from a book page
- Extract error messages from a screenshot to search for them
- Capture contact details from a photo of a business card or poster
Limitations
OCR works line by line, so complex layouts such as multi-column pages, tables and forms may come out in an unexpected order — you may need to rearrange the text afterwards. Handwriting, decorative fonts, very small print and text on top of photos are much harder to recognize than plain printed text.
OCR for screenshots
Screenshots are usually the easiest images to read, because the text is perfectly sharp and straight. The fastest workflow is to take a screenshot (Win+Shift+S on Windows, Cmd+Shift+4 on Mac) and press Ctrl+V or ⌘+V on this page — no need to save the file first. On phones, take a screenshot and choose it from your photo library. If the screenshot contains several text areas, crop it to the part you need first so the result isn’t mixed with menus and buttons. For long documents photographed page by page, add all the photos at once; each page gets its own result box, which you can copy one after another into a single document.
Private by design
Screenshots and document photos often contain personal information. Here, recognition happens entirely on your device. The only things downloaded are the OCR engine (from this site) and the language data (from the public jsDelivr CDN) — your images and the recognized text are never sent anywhere.
Further reading
- How to check whether an online file converter uploads your files — a record of the requests this tool actually makes