Drop an image here or
JPG, PNG, WebP, HEIC, PDF - up to 10 MB
How It Works
Upload your image
Drop any photo, scanned document, screenshot, or PDF. Supported formats include JPG, PNG, WebP, HEIC, TIFF, and PDF. Files up to 10 MB. Need to convert to PDF first? We've got that too.
OCR processes it
The engine analyses the image and extracts all visible text, preserving line breaks and structure. For difficult images - Handwriting, poor scans, tables - Upgrade to AI-powered extraction. You can also use the Photo Editor to crop or clean up your image before uploading.
Copy or download
The extracted text appears immediately. Copy it to your clipboard or download as a .txt file. Edit it directly in the text box before downloading.
What Is OCR?
OCR (Optical Character Recognition) is the technology that converts images of text into machine-readable, editable text. It works by analysing the shapes, patterns, and spatial relationships of characters in an image and mapping them to known character codes.
Modern OCR combines classical computer vision (binarisation, deskew, segmentation) with neural networks trained on millions of text samples. Our engine supports over 100 languages and handles a wide range of document types. If you need to work with the image itself before extracting text, try our Photo Editor to adjust contrast or crop the area you need - Or use the Background Remover to isolate text from noisy backgrounds.
Standard vs AI-Powered Extraction
| Image type | Standard (free) | AI-powered |
|---|---|---|
| Clean printed text, typed documents | Excellent | Excellent |
| Scanned books and forms | Good | Excellent |
| Screenshots, UI text | Good | Excellent |
| Complex multi-column layouts | Fair | Excellent |
| Tables and structured data | Fair | Excellent |
| Handwriting | Poor | Good |
| Stylised / decorative fonts | Poor | Good |
| Non-Latin scripts (Arabic, CJK) | Good (with language pack) | Excellent |
Tips for Best Results
Use high resolution
300 DPI or higher gives the engine the detail it needs for accurate character recognition. Avoid compressed or heavily pixelated images - If your file is too large, compress it first without losing legibility.
Keep it straight
The engine handles slight rotation but performs best on level, upright text. Severely skewed or rotated pages should be corrected before uploading - Use the Photo Editor to straighten and crop your image first.
High contrast helps
Dark text on a light background gives the clearest signal. Low-contrast or coloured backgrounds can reduce accuracy - consider using the Photo Editor to adjust contrast before extracting.