opnai / Image to Text (OCR)
Image to Text (OCR)
Files are deleted automatically after 24 hours. We do not log your content or use it for AI training.
Turn the words in a picture into text you can select, edit and search. Upload a photo of a page, a screenshot, a receipt or a sign and the OCR engine reads it and returns plain text with a Copy button.
Recognition runs on our server, which handles larger images and mixed layouts better than most in-browser scripts. Your image is deleted automatically after 24 hours and is never used to train a model.
How to use Image to Text (OCR)
- Drop a PNG, JPG or WebP image onto the box, choose a file, or paste a screenshot with Ctrl+V. The limit is 10 MB.
- Press Extract text. The page shows when your job is queued and when recognition is running.
- Read the result in the text box, fix any slips, and press Copy to put it on your clipboard.
- Press Delete now if you want the uploaded image removed immediately. Otherwise it expires after 24 hours.
Features
The recognized text appears in a box beside the preview and can be copied with a single button.
Works on straight scans as well as phone photos, within the limits of image quality.
You get clean text without fonts or styles, ready to paste into notes, email or a document.
Open the page, upload and copy. There is no sign-up and no watermark.
No content logging, no model training on uploads, automatic deletion after 24 hours.
Getting accurate text from a photo
OCR quality depends mostly on the picture. Hold the camera parallel to the page, avoid shadows and glare, and fill the frame with the text. A resolution where letters are at least about 20 pixels tall gives the engine enough to work with.
If the first result has errors, upscale the image with the image upscaler and try again. A cleaner source often fixes confusion between similar characters such as O and 0 or l and 1.
Common uses
People use image to text for quoting from a book page, saving the details from a screenshot, digitizing a printed form, copying a code from a photo, or making a scanned note searchable. Always proofread numbers, names and addresses before relying on them.
Frequently asked questions
What kind of images give the best OCR results?
Sharp, evenly lit images with dark text on a light background. Straight-on scans and screenshots are best. Blurry, tilted, low contrast or handwritten text is much less accurate.
Which languages are supported?
The engine reads printed text in many Latin-script languages and some others. Accuracy varies by language and font, so check important results.
Can it read handwriting?
Neat printed-style handwriting sometimes works, but cursive and messy notes are unreliable. Treat any handwriting result as a rough draft.
Is there a limit on image size?
Yes, 10 MB per image in PNG, JPG or WebP. For long documents, process one page at a time.
Does the layout of columns and tables survive?
The output is plain text in reading order. Complex layouts, columns and tables may need tidying after copying. If you want a Word file instead, use the JPG to Word tool.
Is my image private?
Images are processed to produce the text and then deleted after 24 hours, or sooner with Delete now. We do not log content or use it for training.