Pull the text out of a photo or scanned image.
OCR works best on clear, high-contrast, horizontally-aligned text — printed pages, screenshots, and signs work well. Handwriting, blurry photos, or heavily stylized fonts may not extract accurately. The first run downloads a small language model to your browser, which can take a few seconds.
Optical character recognition (OCR) looks at the shapes in a picture and turns them into editable text. Use it to copy text from a photo of a page, a screenshot, a scanned receipt, a sign, a slide or a whiteboard. The text recognition runs on your device using the open-source Tesseract.js engine, so the image is not sent to a server.
The first time you use it, the browser may need to download the language data, which can take a few seconds.
OCR quality depends mostly on the picture.
Always proofread the result. Numbers and similar-looking characters such as 0 and O, or 1 and l, are common mistakes, so double-check amounts, dates and codes.
Because recognition is done locally, photos of IDs, invoices, medical notes and contracts stay on your device. If you only need to copy a few words from a screenshot, your phone's built-in text selection may be quicker; use this tool for whole pages and images that do not offer selectable text.
How accurate is the text extraction?
Accuracy depends heavily on image quality — clear, well-lit, high-contrast text (like a clean document scan) extracts nearly perfectly, while blurry photos or unusual fonts and handwriting are much less reliable.
Does this work with handwriting?
OCR is built and tuned for printed text. It can sometimes read very clear handwriting but isn't reliable for cursive or messy handwriting.
Is my image uploaded anywhere for processing?
No — recognition runs entirely on-device in your browser, so the image and any text in it never leave your computer.