Can I copy text from a picture for free?
Yes. Browser OCR (optical character recognition) reads printed text from screenshots, photos, and scans directly on your device — no app, no account, no upload. Open the image to text tool, select your PNG or JPG, choose the language, and copy or download the recognized text. Because the engine (Tesseract) runs in your browser, a photo of a contract or a receipt never reaches a server — unlike most “free OCR” websites.
What OCR reads well — and what it doesn’t
- Great: screenshots of articles, slides, error messages; clean printed pages.
- Good: straight-on photos of books and receipts in even lighting.
- Hit-or-miss: skewed photos, glare, small fonts, colored backgrounds.
- Poor: cursive handwriting, decorative fonts, very low-resolution images.
5 tricks that double OCR accuracy
- Shoot straight-on, not at an angle — perspective distortion confuses character shapes.
- Crop to just the text region before running OCR.
- Use daylight or a desk lamp; avoid flash glare on glossy paper.
- Pick the correct language — English models misread accented characters.
- For blurry photos, retake rather than upscale; resolution you don’t capture can’t be recovered.
Image OCR vs PDF to text — use the right tool
If your “image” is actually a digital PDF (you can select text in a viewer), skip OCR entirely — the PDF text extractor reads embedded text with 100% accuracy. Use image OCR only for true pictures: screenshots, camera photos, and scans saved as PNG/JPG. For scanned multi-page PDFs, convert pages to images with PDF to JPG first, then OCR the pages you need. Full comparison in best free PDF extractors 2026.