Image to text converter

Upload a screenshot or a scanned page to convert the text in it into editable text that you can copy. The tool reads English, simplified Chinese and traditional Chinese. Printed text and clear screenshots are recognised most accurately. Handwriting and blurred photos are often recognised poorly.

Upload image

Language of the text

You can also drag an image here.

After taking a screenshot, press Ctrl+V on this page to paste it.

JPG, PNG, WebP, GIF, BMP or HEIC, not more than 20 MB.

Extracted text

The extracted text will be shown here after you upload an image.

How to use

  1. Please upload an image with text. Select “Choose image”, drag the image into the box, or take a screenshot and press Ctrl+V on this page. Then choose the language of the text.
  2. Results depend on the image. Printed text in screenshots and scans is read well. Photos taken at an angle, uneven light, very small text and handwriting cause many more mistakes. Please check the result against the image, and crop the image to the text if necessary.
  3. The average confidence is shown under the result. Below 70%, the result may have many mistakes, so please check it line by line.
  4. The first scan downloads the program and the language files from this website: about 7 MB for English and about 9 MB for Chinese and English. The text is then read in your browser. Your image and the text are not uploaded.

Text is read with tesseract.js (Apache-2.0 licence) and the Tesseract language files (Apache-2.0 licence). They are downloaded from this website to your browser the first time you use the tool. Your image and text are not uploaded.

Frequently asked questions

Are the layout, tables and columns kept?
No. The result is plain text, line by line. Tables, columns and font styles are not kept.
Why are there mistakes in the result?
This tool uses the open-source Tesseract engine in your browser. It is less accurate than the text recognition built into phones and paid services, especially for Chinese. Clearer images with larger text give fewer mistakes.