Image to Word
This reads printed English text. It does not rebuild the original layout, columns, or tables. Handwriting and photos of screens are often incomplete. The image stays on your device. The recognition engine is downloaded once from the network.
Drop a photo of printed text here or click to choose
A straight, well-lit photo of a printed page works. A blurry screenshot usually does not.
Pull the words out of a picture and save them as plain text or a simple Word document. The Word file is a series of paragraphs. It is not a replica of the page.
How to use
- Photograph the page straight on, with the text large in the frame. Image to Scan can clean a gray page first if the photo is yellow or uneven. Then open that JPG here.
- Click Extract text. The first run downloads the English recognition engine. Later runs in the same browser are faster. Your image is not sent to a server.
- Check the text. Names, small print, and tight columns are the usual mistakes. Fix them in the box.
- Copy, download TXT, or download Word. The Word file uses one paragraph per line that was recognized.
What this will not do
- It will not keep fonts, columns, headers, or tables in place.
- It is English only on this page. Other languages are a different engine, not a setting we left out by accident.
- Handwriting, cursive, and text inside a photograph of a monitor are weak inputs.
- It is not a substitute for a certified translation or for an accessible PDF that already has a text layer.
Questions
Why does it need the network if the image stays local?
The recognition program and the English language data are too large to ship inside the page. They download from a public CDN when you click Extract. The picture itself is not part of that request.
The box is empty.
The page may be too skewed, too small, or not English. Crop closer to the paragraph and try again.