1. The quickest way — image to text in your browser
- Open the Image to Text (OCR) tool.
- Drop in your JPG or PNG — or a scanned PDF.
- Pick the language of the text.
- Copy the recognised text, or download it as a .txt file.
The OCR runs on your device — your image is never uploaded, so it's fine for receipts, IDs and anything private.
Extract your text nowFree OCR in your browser — screenshots, photos and scans, multiple languages, no upload.
Open Image to Text →2. Why you can't just copy text from an image
To your eyes a screenshot full of text and a web page full of text look the same — but to your computer they're completely different. The web page stores actual characters; the screenshot stores only pixels, coloured dots that happen to be arranged in letter shapes. There's nothing to select. OCR bridges the gap: it examines those pixel patterns, recognises each character, and reconstructs the words — turning a picture of text back into text.
3. Getting the most accurate results
OCR is very good on clean input and struggles on messy input. A minute of preparation beats ten minutes of fixing errors:
- Shoot straight-on — hold the camera flat over the page; skewed angles distort letter shapes.
- Light it evenly — shadows and glare across the text are the biggest accuracy killers.
- Sharp beats big — a focused photo at normal resolution beats a blurry high-resolution one.
- Crop to the text — removing busy backgrounds gives the engine less to misread.
- Proofread names and numbers — OCR fixes itself on dictionary words but can't guess a phone number, so check digits by eye.
Printed text works far better than handwriting — neat block capitals sometimes recognise, but cursive is still beyond most OCR engines.
4. Other languages
The language picker isn't decoration — it loads a model trained on that language's characters and vocabulary. Run German text through an English model and you'll get mangled umlauts and guessed words. Pick the right language first and accuracy jumps. Once the text is out, you can even run it through the Translate tool — photo of a foreign menu to English text in two steps.
5. Screenshots, photos and scanned PDFs
The same technique covers more than photos. Screenshots — of an error message, a quote, a code snippet — are ideal OCR input: perfectly flat, perfectly lit. Scanned PDFs are just images wrapped in PDF pages, and the tool handles those directly too. If you want the scan itself to become searchable — with an invisible text layer over each page rather than a separate .txt file — that's a slightly different job, covered in our guide to making a scanned PDF searchable. And if your PDF already has real text (you can select it in a viewer), skip OCR entirely and use the faster extract text from PDF method.
Need to OCR a whole scanned PDF?PDFNest has 32 free tools that all run in your browser.
Explore PDFNest →Frequently asked questions
How do I extract text from an image for free?
Open the Image to Text tool, drop in your JPG or PNG, pick the language, and copy or download the recognised text. It runs in your browser — nothing is uploaded.
Can I copy text from a screenshot?
Yes — a screenshot's text isn't selectable because it's just pixels, but OCR reads it in seconds.
How accurate is OCR on photos?
Very accurate on clear, flat, well-lit photos and screenshots; less so with blur, glare, odd angles or handwriting. Always skim the output for errors in names and numbers.
Does OCR work in other languages?
Yes — pick the correct language before running OCR and accuracy improves dramatically.