FileTools

Image to Text (OCR)

Extract editable text from images and screenshots, right in your browser.

Choose files

or drag and drop · or paste a screenshot (Ctrl/⌘+V)

Your files are processed on your device and never uploaded.

How to extract text from an image

  1. Choose the language of the text in your image. English is recognized automatically as well.
  2. Add your image. Click Choose files, drag images into the box, or simply paste a screenshot with Ctrl+V (⌘+V on Mac).
  3. Wait a few seconds while the text is recognized. The first run downloads the OCR engine and language data.
  4. Copy the text or download it as a .txt file.

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:

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

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

Frequently asked questions

Is this OCR tool free?

Yes. It's completely free, with no sign-up and no limit on the number of images.

Are my images uploaded to a server?

No. Text recognition runs inside your browser. Only the OCR language data is downloaded the first time you use a language; your images never leave your device.

Which languages are supported?

Twenty languages, including English, Korean, Japanese, Simplified and Traditional Chinese, Spanish, French, German, Portuguese, Russian, Arabic, Hindi, Thai and Vietnamese. English is always recognized alongside the language you pick.

Can it read handwriting?

It's designed for printed text. Neat block handwriting sometimes works, but cursive and messy handwriting usually won't be recognized reliably.

Can I extract text from a PDF?

Convert the PDF pages to images with PDF to PNG first, then add the images here.

Why is the first recognition slow?

The OCR engine and the language data (a few megabytes per language) are downloaded once. After that they're cached by your browser and recognition is much faster.

Related tools