Skip to content
Hawwil

Extract text from images and PDFs (OCR)

Photographed a document, or have a scanned PDF and want the text? Convert it here to text you can copy and edit. Arabic and English are supported, and everything happens inside your browser.

Processed on your device· Your files are never uploaded to any server

Drag and drop your files here

or

Accepted formats: PDF, JPG, PNG, WebP, HEIC — max 50 MB per file

The first use downloads the language model (about 5 MB) and stores it in your browser.

How to use this tool

  1. Drag and drop images or scanned PDF files into the box, or click "Choose files".
  2. Choose the text language. The default "Arabic and English" suits most documents.
  3. Click "Extract text" and wait. The first time we load the language model, which takes a little while; after that it is faster and works offline.
  4. Copy the text or download it as a TXT file. Review it before relying on it, because automatic recognition can make mistakes.

Frequently asked questions

Does it support Arabic?

Yes. It supports Arabic and English, and you can choose "Arabic and English" for mixed documents. Accuracy is higher with clear printed fonts, and lower with handwriting or decorative fonts.

Are my images uploaded to a server?

No. Text recognition happens inside your browser, so you can use it for official and sensitive documents. We even host the language model ourselves and do not load it from an outside site.

Why does the first time take longer?

Because the language model (about 5 MB) is only downloaded the first time and stored in your browser, so later runs are faster and work even without internet.

How do I get the best result?

Use a clear image with good lighting, straight and not upside down text, and high resolution. Avoid shadows and reflections.

Does it support scanned PDF files?

Yes. We convert each page to an image and extract its text. The maximum is 50 pages at a time.

Why did some words come out wrong?

Automatic recognition is never 100% accurate, especially with numbers, small fonts or shaky images. Review the text and correct it before using it.