Extract Text from Image — Free OCR Tool
OCR (Optical Character Recognition) turns the text in an image into actual characters you can copy, edit, and search. It's how you get usable text from a screenshot, a photo of a document, a scanned page, or any other image that contains words. This tool runs Tesseract.js — a mature, open-source OCR engine — directly in your browser. No file upload, no account, support for 60+ languages.
What OCR Is and How It Works Here
Optical Character Recognition is the process of analysing an image pixel-by-pixel to identify the shapes of letters, digits, and punctuation. The engine compares detected shapes against trained character models to determine what each symbol is, then assembles them into words and lines of text.
This tool uses Tesseract.js, a WebAssembly port of Google's Tesseract OCR engine — one of the most widely used and tested OCR systems in existence, originally developed by HP and now maintained by Google as an open-source project. The engine and the language data files it needs are downloaded to your browser (typically a few MB per language). After that initial load, recognition happens entirely locally.
The result is text that matches the content of your image, laid out in reading order. You can then copy it, edit it, translate it, or paste it into any document.
How to Get the Best OCR Accuracy
OCR accuracy is almost entirely determined by the quality of the input image. These practices consistently improve results:
Use high-resolution images
Text in images below 100 DPI is difficult to resolve accurately. For scanned documents, 150–300 DPI gives the best balance of file size and recognition quality. Screenshots from modern screens are typically fine without any adjustment.
Ensure high contrast between text and background
Black text on white background is ideal. Light grey text on white, or coloured text on a patterned background, causes significantly more errors. If your image has low contrast, try increasing it in an image editor before running OCR.
Keep the image straight
Skewed or rotated text degrades accuracy. A document scanned at a slight angle may be significantly harder to read than the same document scanned level. Most document scanner apps have an auto-deskew feature worth enabling.
Select the correct language
Tesseract uses language-specific character frequency data to improve recognition. Running English OCR on a French document still works, but selecting French will produce noticeably better results, especially for accented characters.
Languages Supported
The tool supports over 60 languages. A selection of supported languages includes:
And many more. If your language is not listed above, select it from the language dropdown in the tool — the full list is available there.
Frequently Asked Questions
How accurate is the OCR text extraction?
Accuracy depends heavily on image quality. Clear, high-contrast printed text in good resolution (150 DPI or above) typically achieves 95–99% accuracy. Handwriting, stylized fonts, and low-quality scans produce lower accuracy. Always review the output for errors.
Does it work on screenshots?
Yes. Screenshots typically have high contrast and clean fonts, making them some of the best source material for OCR. Text in screenshots of websites, documents, and applications extracts very accurately.
Which languages are supported?
Over 60 languages are supported including English, Spanish, French, German, Portuguese, Italian, Dutch, Russian, Chinese (Simplified and Traditional), Japanese, Korean, Arabic, Hindi, and many more. Select the correct language before running OCR for best results.
Is my image uploaded to process the OCR?
No. The OCR runs entirely in your browser using Tesseract.js — an open-source OCR engine compiled to WebAssembly. Your image data never leaves your device.