Image to text (OCR)

The tool uses optical character recognition (OCR), powered by the open-source Tesseract engine, running directly in your browser. It finds lines of text in the image, recognizes the letters and gives you plain, editable text.

The first time you use a language, the browser downloads its recognition data (a few megabytes), which can take a few seconds. Later runs are faster. Your image itself is never uploaded: the recognition happens on your device, so it is safe for receipts, letters and other personal documents.

You can search 120 languages and scripts, from English, Spanish and German to Japanese (horizontal and vertical), Chinese, Arabic, Hebrew, Hindi and Cyrillic, and pick up to 3 at once for mixed documents. Choosing the right language matters, because it lets the engine recognize accented letters such as é, ñ or ü correctly.

• Several images in a row: the text from each is collected into one document, separated by file name.
• Select a region: drag a rectangle on the first image to read only that part, such as one column of a table.
• Download as Word (DOCX) or TXT, or copy to the clipboard.
• Paste: drop a screenshot straight in with Ctrl+V.

Language data downloads once per language (1–15 MB) and is cached by your browser afterwards.

OCR is only as good as the image. For clean text:

• Shoot straight on. Hold the phone parallel to the page, not at an angle.
• Use even light. Avoid shadows across the page and glare on glossy paper.
• Fill the frame with the text; crop away the table or background with Crop image.
• Keep it sharp. Tap to focus and hold still; blurry photos cause many errors.
• Higher resolution helps with small print. Screenshots are usually recognized very accurately.
• One column at a time. Complex layouts with columns, tables or text over pictures can come out in a mixed-up order.

Always proofread the result, especially numbers, names and email addresses, where a single wrong character matters. Common confusions include the letter O and the digit 0, lowercase l and the digit 1, and rn read as m.

Formatting such as bold, font sizes and columns is not kept; you get plain text. If you need a formatted, editable document, use JPG to Word, which rebuilds paragraphs and headings in a .docx file.

Many OCR websites upload your image to their servers. Here, the Tesseract engine runs inside your browser tab, so the picture and the recognized text stay on your device. That makes it a good fit for bank statements, medical letters, IDs and other documents you would rather not upload. When you close the tab, nothing is kept.

• Copying text from a screenshot of an app, a video frame or a website that does not allow selecting text.
• Digitizing printed documents: letters, recipes, notes and old typewritten pages.
• Receipts and invoices for expense reports.
• Quotes from books for study notes or citations.
• Slides and whiteboards photographed in class or meetings.
• Accessibility: turning text in images into text that screen readers and translation tools can handle.

Note that recent phones can also select text in photos (Live Text on iPhone, Google Lens on Android), which is handy for quick copying. This tool is useful on computers, for longer texts and when you want to keep the processing on your own device.