Image to Text (OCR)

Extract text from a photo, screenshot, or scanned document using optical character recognition, running entirely in your browser with no image upload to a server. Digitize a printed page, pull text from a receipt, or recover a quote from an old image instantly, with support for multiple languages, completely free.

Drop an image here, or click to choose one

JPG, PNG, WebP, and most other image formats

Usage examples

Copying text from a screenshot

Pull the text out of a screenshot of an article or chat so you can edit or search it, instead of retyping it by hand.

Digitizing a photo of a printed page

Take a photo of a book page, sign, or printed document and turn it into text you can copy and paste.

Extracting text in another language

Pick the matching language before extracting to get better accuracy on non-English text.

Pulling text out of a scanned receipt photo

Photograph a paper receipt and extract its text so you can copy the total or item list into an expense tracker without retyping it.

Recovering text from an old meme or quote image

Extract the caption text baked into an image file so you can quote, translate, or search for it as plain text.

Frequently asked questions

Is my uploaded image sent to a server?

No. Recognition happens entirely in your browser, and the image is never uploaded. The one exception is that the recognition model file for your chosen language is downloaded once from a public CDN on first use — this is just a generic program file, unrelated to your image's content.

How accurate is the recognition?

It depends heavily on the image: high-resolution, well-lit, straight-on photos of clear printed text recognize best. Handwriting, blurry or low-resolution images, skewed angles, and unusual fonts all reduce accuracy.

Why isn't my language in the list getting recognized well?

Make sure you've selected the matching language before extracting — the recognition model is tuned per language, and picking the wrong one will produce poor results even for clear images.

Can I extract text from a PDF instead?

Yes — use the separate PDF OCR tool, which is built specifically for scanned PDF pages.

Does it work well on handwriting?

Not reliably. This tool is built around recognizing printed and typed text; cursive or handwritten text has far more variation and typically produces inaccurate or garbled results, even on clear photos.

Can I extract text in more than one language at once?

No, one language model is used per extraction. If an image mixes two languages, run the extraction once per language and compare which result captures each part more accurately.

Why does the first extraction take longer than the next one?

The first run for a given language has to download that language's recognition model before it can start, which takes a few seconds depending on your connection. Subsequent extractions in the same language reuse the already-downloaded model and start recognizing immediately.