Retyping text you can see in a picture is a waste of time. Optical character recognition (OCR) reads the text for you and turns a screenshot, photo or scan into editable, copyable words — and it runs instantly, so your images are never stored.
Pick a screenshot, photo or scan that contains text.
Select the language of the text in the image.
Click Extract — recognition runs securely.
The recognised text appears in a box — copy it anywhere with one click.
Getting accurate results
OCR is good but it isn’t magic — clear input gives clean output. For the best accuracy, use a sharp, well-lit image where the text is large and level, avoid heavy backgrounds behind the words, and pick the correct language so the engine knows which characters to expect. Straightening a tilted photo before you upload helps a lot.
Handy uses for OCR
- Copying text from a screenshot or photo of a slide.
- Digitising a receipt, label or business card.
- Grabbing a quote from an image instead of retyping it.
- Pulling text out of a scanned document.
Tip: The first run downloads the language model, then caches it — so the next extraction in the same language starts instantly.
Extract text from an image now
Turn a screenshot, photo or scan into editable text — free, private OCR.
Open the Image to Text tool →Frequently Asked Questions
Does the image get uploaded?
No. OCR runs securely; your image stays private.
Which languages are supported?
Common languages including English, Spanish, French, German, Hindi, Arabic and Chinese.
Why is the first run slower?
The language model downloads once and is then cached before recognition starts.
Is it free?
Yes — completely free, with no sign-up or limits.