Image to Text (OCR)
Have a picture with words in it — a photo of a page, a screenshot, a receipt or a sign? Read the text out of it and download it as a plain text file, or as a searchable PDF that keeps the image and makes its words selectable.
Files are never uploadedHow to use it
Add one or more images
Drop your pictures onto the box above, or click to browse. You can add several at once; each one becomes its own text file.
Choose the text layout
Leave it on Automatic for most images. Use "A single line" for a screenshot of one row, or "Sparse text" when the words are scattered across a sign or a label.
Let it read the words
The tool looks at the pixels and recognises letters, numbers and punctuation. The progress bar moves as each image is processed.
Check the result
Open a text file and compare it with the picture. Small or blurred text and fancy fonts are where mistakes happen, so check anything you will rely on.
Download the text files
Save the files individually, or use Download all when you added several images. Tick “Also save a searchable PDF” if you want the picture back with selectable text on top.
Common issues
Blur and glare defeat recognition
A photo taken in a hurry is often slightly out of focus or has a bright reflection across the words. Retake it square-on, in even light, and run it again.
Rotated photos read badly
Text that is sideways or upside down is misread or missed. Straighten the image before adding it — the recogniser expects the text roughly level.
Numbers and similar-looking letters get swapped
Characters such as 0 and O, 1 and l, 5 and S are easily confused. Always proofread figures you intend to use.
Plain text keeps no layout at all
The .txt output carries words only — no fonts, colours, columns or images from the picture. If you want the picture preserved, tick the searchable PDF option, which keeps the image and hides the text behind it.
Frequently asked questions
What images work best?
What is the searchable PDF option for?
Which languages are supported?
Are my images uploaded?
Can it read handwriting?
Why is a small logo or a stylised heading missed?
Can I process several images at once?
Does the text keep the arrangement of the words?
What if no text is found?
Related tools
Send us feedback
Something broken, or unclear? Tell us and we will look at it.