PDFJadoo › Scan & OCR › OCR Image
The PDFJadoo genie with his golden lamp, rising in a swirl of gold magic

Lift the text out of any image

A photo of a page, a screenshot, a scanned letter — the words are trapped as pixels. OCR reads them and hands you real, editable, copyable text.

Runs 100% in your browserFiles never uploadFree · no sign-up
Drop an image here

or click to choose a .jpg, .png or .webp

Honest notes: the recognition engine loads once (~a few MB) and then works fully offline. Pick the document's language above — 124 languages are available, and each one's data (about 0.3–5.7 MB) downloads once, only the first time you use it, and then works offline. The text is returned exactly as the engine reads it — nothing is silently rewritten. Results are best on straight, well-lit, high-contrast pages; joined scripts such as Urdu come out best from clean printed pages, and we don't pretend otherwise. You can edit the recognised text in the box before downloading.

🔒 Private by design: this tool runs entirely in your browser. Your file is never uploaded, never stored, and works even with the internet off.

How to OCR an image

  1. Drop your image into the circle.
  2. Press Read the text.
  3. Wait while the engine reads — progress shows as it goes.
  4. Correct any slips right in the text box.
  5. Copy the text or download it as a .txt file.

This is the same Tesseract OCR engine used by libraries and archives, compiled to run in your browser. A photo of a confidential document is recognised on your own device — the image never travels to a recognition server.

Related wishes

Or browse the full deck of 130 free PDF tools — or fix a blurry or shadowed photo first, free, before it goes any further.

OCR Image — fair questions

How accurate is it?

On clean printed pages, very — with Enhanced accuracy on, our receipt tests read at 100%. Handwriting, glare, skew and tiny text all reduce accuracy — the Scan Documents straightening helps first. The text is returned exactly as read; nothing is silently “corrected”.

Which languages?

124 languages — the full Tesseract set, from English, Urdu, Hindi and Arabic to Chinese, Bengali, Turkish and 117 more. Mixed pages work too: pick “English + Urdu”, “English + Arabic” or “English + Hindi” to read both scripts together. Each language's data (about 0.3–5.7 MB) downloads once — only when you pick that language — and then works offline. Nothing you scan is ever sent anywhere.

Can it do a whole scanned PDF?

Use OCR PDF — it reads every page and returns a searchable document.

Back to All Tools