Question 1

Does my scanned document get uploaded anywhere?

Accepted Answer

No. The OCR engine is downloaded to your browser the first time you use the tool and runs entirely on your own device. Your document never leaves your computer, which is different from most free OCR services that process files on their servers.

Question 2

Why does it take a few seconds to start the first time?

Accepted Answer

The recognition engine and the language data are fetched once on first use, then cached by your browser. Later runs start immediately. This is also why the tool costs nothing to use and has no page limit.

Question 3

Can I extract text from more than one page at once?

Accepted Answer

Yes — switch to "Page range" and pick up to 20 pages at a time. Each page is processed one after another, with progress shown as it goes, and the result shows exactly which pages returned text. Larger ranges are split into batches of 20 on purpose, since each page takes several seconds and an unbounded run would tie up your browser tab for a long time with no way to check progress.

Question 4

Which languages can it read?

Accepted Answer

English, Hindi, Assamese and Bengali. Pick the language that matches your document before running it — accuracy drops sharply if the wrong script is selected.

Question 5

How accurate is it on a typical scanned form?

Accepted Answer

Clean, straight scans of printed text read very well. Accuracy falls on photos taken at an angle, low-contrast faxes, or handwriting — handwriting in particular is not reliably supported. If the result looks wrong, rescan at a higher resolution and make sure the page is flat and evenly lit.

Extract Text (OCR)