Turn images and PDFs into text (OCR)
OCR for photos, screenshots and PDFs. Free, with no limits and no watermark. Everything happens inside your browser, so your files are never uploaded.
Reading the file…
This file
Your file is never sent anywhere. The reading happens inside this browser.
One-off setup
This loads the reading engine (about 40 MB) into your browser. On a slow connection it can take a minute. It is only needed once.
Pressing "Read the text" starts this download too.
Keep this page open. Switching to another app stops the work.
Done
Lines the reader was unsure about are marked. Check numbers and names against the original.
Keeps the original look, and the text can be searched and copied — in Japanese too.
Why this tool
- No retyping. Paper documents, photos, anything on your screen — straight into text.
- Once set up, images are read with the network switched off. (Loading a PDF, and saving a searchable PDF, each fetch one more component.)
- A searchable PDF that really is searchable. It looks exactly like the original, but you can search it and copy from it — including Japanese and other non-Latin text, something many tools get wrong.
- PDFs go straight in — no converting to images first (up to 20 pages).
- Paste a screenshot. Take one and press Ctrl+V on this page.
- iPhone photos (HEIC) open in Safari only; elsewhere, convert them to JPEG first.
- Uncertain lines are marked, so you know where to check.
How to use it
- Press "Get ready" once (about 40 MB; a minute on a slow connection).
- Drop an image or a PDF in, or paste a screenshot with Ctrl+V.
- Press "Read the text".
- Copy the text, or save it as .txt or as a searchable PDF.
Questions
Is my file uploaded?
No — our server has no way to receive files at all. The only traffic is fetching the reading engine.Once that is done, images are read with the network switched off; try it if you want to be sure. (Loading a PDF and saving a searchable PDF each fetch one more component.)
What is a "searchable PDF"?
A PDF that looks like the original scan, with a layer of text sitting on top of it. You can search the document and copy from it — useful for filing paper you may need to find again later.
How accurate is the OCR?
On a clean document it is close to perfect, digits included. Accuracy drops on photos that aretilted, dark or blurred, and similar-looking characters can be confused.Check amounts, dates and names against the original.
Are there any limits?
Files up to 30 MB and PDFs up to 20 pages, at roughly 10 seconds per page. For longer documents, take the pages you need first with Extract PDF pages.
Does it read handwriting?
It is built for printed text. Handwriting sometimes works, but do not count on it.