How to use: Image to text (OCR)
- 1Choose one or more images, or a scanned PDF.
- 2Pick the language in the image - English, Hindi, or both.
- 3Press "Extract text". The first run downloads the recognition engine once, later runs are faster.
- 4Copy the text or save it as a .txt file.
Good uses for OCR
- Copy text from a screenshot or photo instead of typing it out.
- Make a scanned document searchable, or paste its text into an email or form.
- Digitise printed notices, receipts and letters in Hindi or English.
- Pull a paragraph out of a scanned book for notes or quotes.
Getting the best results
- Shoot straight on with the page filling the frame, in good light without glare.
- Higher resolution helps small print; avoid screenshots of tiny thumbnails.
- Crop to the text so backgrounds and photos don’t confuse the reader.
- Rotate sideways scans first with Rotate PDF; upright text reads best.
Privacy
ID cards, letters and bills often contain personal details. Unlike most online OCR services, NapTools never uploads your files, so there is nothing for anyone else to store or read.
Frequently asked questions
Is it accurate?
For clear, well-lit printed text, accuracy is usually very high. Handwriting, very small print, heavy shadows or blurry photos lower accuracy. Always check names, numbers and dates before using the text.
Does it work with Hindi?
Yes. Choose Hindi for Devanagari text, or Both for documents that mix Hindi and English, which is slower but reads both scripts.
Can it read a whole scanned PDF?
Yes. Each page is rendered and read in turn, and the text of every page is combined into one file, with a separator between pages.
Are my documents uploaded?
No. Text recognition runs inside your browser using an open-source engine (Tesseract). Your images and PDFs are never sent to a server.
Why is the first run slow?
The recognition engine and language model (a few megabytes) download the first time you use the tool. They are then cached by your browser.