Image to Text (OCR)
Extract editable text from photos, scans and screenshots in 15 languages — recognition runs right in your browser.
OCR settings
About this tool
A photographed page of notes, a screenshot of an error message, a receipt, a book excerpt, a sign in another language — retyping is slow and error-prone. This free image to text converter uses optical character recognition (OCR) to pull the words out of JPG, PNG, WebP and BMP images and hand them back as editable text you can copy or save as a .txt file. Fifteen languages are supported, from English and Spanish to Arabic, Hindi, Japanese, Chinese and Korean, and line breaks are preserved by default so lists and addresses keep their shape.
Unlike most online OCR services, the recognition runs inside your browser using the Tesseract engine compiled to WebAssembly. The engine and language data are fetched from a CDN on first use and cached, but your image is read on your own machine and never sent anywhere — which matters when the image is a contract, an ID or a medical document. For the best result, use a sharp, well-lit image at a decent resolution with the text roughly horizontal; a good scan or screenshot comes out close to perfect, while a blurry phone photo of a curved page will not.
How to extract text from an image
- Choose the language of the text and whether to keep line breaks.
- Drop one or more images, or paste a screenshot with Ctrl+V — recognition starts automatically and a progress bar tracks it.
- Edit the text in place if needed, then Copy it or Download it as .txt.
Frequently asked questions
Is my image uploaded for OCR?
No. The recognition engine (Tesseract compiled to WebAssembly) runs inside your browser, so the image itself never leaves your device. What is downloaded, once, is the engine and the language file you pick — a few megabytes fetched from a public CDN and then cached by your browser.
Why is the extracted text wrong or garbled?
OCR needs clear, reasonably large letters on an even background. Blurry phone photos, tiny screenshots, skewed pages, handwriting and decorative fonts all hurt accuracy. Choosing the wrong language is another common cause — the engine guesses characters from the alphabet it was given.
Which languages are supported?
Fifteen out of the box: English, Spanish, French, German, Italian, Portuguese, Dutch, Polish, Turkish, Russian, Arabic, Hindi, Japanese, Simplified Chinese and Korean. Each language has its own data file, downloaded the first time you select it, so switching languages triggers a short extra load.
Can I convert several images at once?
Yes. Drop a batch of images and they are recognized one after another, each with its own text box, copy button and .txt download. Download all saves a single text file with the output of every image separated by headings.
Does it work on scanned PDFs?
This tool takes image files only. For a scanned PDF, first convert its pages to JPG with the PDF to JPG tool, then drop those images here. If your PDF already contains real text rather than scans, the PDF to Text tool extracts it directly and far more accurately.