File Limits & Why They Exist
To keep things fast and stable, each scan is capped at 1 PDF, 50 MB per file, and 200 pages.
This tool runs 100% inside your browser — your file is never uploaded to any server. That privacy comes with a trade-off: all the work uses your device's memory. A 50 MB PDF is far larger than 50 MB in use once its pages are rendered to images, thumbnails are built, and each page is handed to the OCR engine — it can easily balloon to several hundred MB.
The limits are a safety ceiling that prevents the browser tab from freezing, failing to export, or crashing — which is especially important on phones and tablets with less memory. If your file is larger, split it into smaller parts first, or use a desktop browser for the best headroom.
Tip: For very long documents, scan a page range at a time (e.g. 1-50, then 51-100) to keep memory low.
Getting Started
1
Upload a PDF — Drag & drop a scanned PDF onto the upload area, or click "Select File" to browse. Your file stays 100% on your device; nothing is uploaded to any server.
2
The workspace opens with 3 panels: Pages (thumbnails on the left), Preview (center), and Options (right). On mobile, tap the bottom tabs to switch panels.
Toolbar Buttons
- — Go back to the upload screen
- / — Previous / Next page (or use the ← → arrow keys)
- / — Zoom out / Zoom in (50%–300%)
- — Fit Width (fill the preview area)
- — Reset zoom to 100% (Fit Page)
- — Toggle dark mode (top-right corner)
- — Open this guide
How to Run OCR
1
Pick a Text Source — Auto (default) instantly extracts a page's existing text when the PDF already has a selectable text layer, and only falls back to OCR for scanned pages. Choose Always OCR to force image recognition on every page.
2
Set Quality — Choose Fast for quick results, Balanced (default, recommended), or Best for maximum accuracy on fine text.
3
Choose Language Mode — Auto Text Extract + English OCR Fallback is on by default. The tool extracts selectable text directly from digital PDFs; for scanned/image pages it uses English OCR as the fallback. This is not automatic language detection — for scanned non-English pages, switch to Manual OCR Language Selection and tick the languages present in your PDF for accurate results.
61 OCR languages are available: 31 popular languages are always visible, plus 30 advanced languages (open Advanced Languages). Indian languages include Assamese, Odia, Punjabi and Sanskrit alongside Hindi, Bengali, Tamil, Telugu, Kannada, Malayalam, Marathi, Gujarati, Nepali and Urdu. Note: some Indian languages (Bodo, Dogri, Kashmiri, Konkani, Maithili, Manipuri) appear in the list but are disabled because no Tesseract OCR model is currently available for them. Mixed-language presets make it easy to select English + another language.
Examples:
• English + Hindi for bilingual Indian documents
• English + Arabic for invoices or forms with both scripts
• English + Chinese for documents containing English and Simplified or Traditional Chinese
4
Optional: Set Page Range — Leave blank to scan all pages, or type specific pages.
Example: 1, 3-5, 10 → scans pages 1, 3, 4, 5, and 10 only.
5
Click "Scan This Page" to process the currently previewed page, or "Scan All Pages" to batch-process. A progress bar shows status; press ESC or the Cancel button to stop.
6
View results — Extracted text appears in the "Text Output" box. Processed pages get a green border and ✓ badge in the Pages panel. Click any thumbnail to view that page's text.
Exporting Results
- Copy Text — Copies the current page's extracted text to your clipboard (if the current page has none, all scanned pages are copied with page markers).
- TXT — Downloads all extracted text as a plain-text file, organized by page.
- CSV — Downloads a spreadsheet-ready file with columns: Page Number, Extracted Text.
- JSON — Downloads structured JSON data, useful for developers.
Example TXT output:
[ PAGE 1 ]
Invoice #12345 — Date: 01/15/2026…
[ PAGE 2 ]
Item descriptions and totals…
Tips & Best Practices
- Best quality gives the most accurate results but takes longer — use it for small or fine-print documents.
- Previously scanned pages are cached — clicking a scanned page shows results instantly without re-processing.
- Your files are processed entirely in your browser and are never uploaded to any server. Internet may be needed to load the OCR engine and language data on first use.
- Use "Clear All" to reset everything and start with a new PDF.
- Dark mode is remembered across sessions.
- On mobile, swipe between Pages, Preview, and Tools tabs at the bottom.