Image to Text – Extract Text from Photos and Screenshots (OCR)

How to extract text from an image

1. Choose an image — a photo, screenshot or scan.
2. Choose the language of the text.
3. Press Extract text. The first time you use a language, its data is downloaded (about 2–15 MB), then reading starts.
4. Edit, copy or download the text.

Tips for better results

• Photograph straight on, with the page flat and filling the frame.
• Use good, even light and avoid shadows and glare.
• Keep it sharp — tap to focus before taking the photo.
• Crop to the text with crop image to remove busy backgrounds.
• Choose the right language — it affects accents and letter shapes.
• Higher resolution helps small print; avoid heavily compressed images.

What you can use it for

• Copying text from screenshots of slides, videos or apps that don’t allow copying.
• Digitising printed documents, letters and notes.
• Receipts and invoices for expense reports.
• Business cards into your contacts.
• Quotes from books for study and research.
• Signs and menus — combine with a translation app when travelling.
• Accessibility — turn images of text into text that can be read aloud with text to speech.

Supported languages

English, French, Spanish, Portuguese, German, Italian, Dutch, Polish, Swahili, Afrikaans, Hindi, Bengali, Arabic and simplified Chinese, plus combined English + French and English + Hindi. Each language model is downloaded only when you first use it.

How it works

The tool uses Tesseract, a well-known open-source OCR engine originally developed at HP and later maintained with Google’s support, compiled to WebAssembly so it runs in your browser. It finds lines of text in the image, recognises characters with a neural network trained for each language, and returns the text with an overall confidence score.

Getting the text into other apps

After extraction, the text box is editable, so you can correct any misread characters before copying. Paste into Word, Google Docs or email, or download a .txt file. For tables, the text comes out line by line; paste it into a spreadsheet and use “split text to columns” to rebuild the table.

Numbers and special characters

OCR sometimes confuses similar characters, such as 0 and O, 1 and l, or 5 and S. Always check account numbers, amounts and codes against the original image before using them.

Privacy

Many OCR websites upload your image to a server. This one reads it on your device; only the language model is downloaded, from the jsDelivr content delivery network. That makes it suitable for ID documents, bank letters and medical forms.

To count the words you extract, use the word counter; for text-based PDFs, use PDF to text.