- Drag it in, choose it, or paste it. It is read here in the tab and never uploaded.
- The settings column holds 2 controls for this tool. The defaults suit most files.
- “Read the text” works on your device, then Original and Result sit side by side with a download.
About Image to text
Image to text reads the words in a photo, a screenshot or a scan and saves them as plain text you can copy, search and paste — one file for every picture, or all of them in one.
The recognizer is the one Make searchable uses, tesseract, run in your browser with its English model. It reads printed type well: screenshots, documents, signs, receipts. Handwriting, very small or blurred text, and languages other than English come out poorly or not at all.
Lines come out in reading order, and the recognizer’s own blocks — a column, a caption, a table — are kept apart by a blank line. Layout beyond that is not kept: a table comes out as lines of words, not as columns. For a scanned PDF, Make searchable puts the words back into the pages instead; PDF to Word rebuilds the layout.
A phone photo is read the way up it is shown, and drawn down to four thousand pixels on its long side first, which is more than the recognizer needs. The recognizer and its English model come to about seven megabytes, loaded from this site — not from anyone else’s — when the tool runs.