A scanned PDF is just photos of text: unsearchable, uncopyable, invisible to Ctrl+F. OCR (optical character recognition) reads the pictures and returns real text.
This guide shows how to run OCR in 14 languages entirely on your device — no uploads, no per-page fees.
Try it now: 👁 OCR — Scan to Text · 📋 Extract Text / Markdown · 📡 Scan to PDF
Match the language, double the accuracy
Pick your document's language before running — recognition quality jumps dramatically with the right language pack (especially Arabic, Hindi, and CJK scripts). First run downloads the pack; later runs are instant and offline.
What you get back (and proofreading)
You receive two files: the raw recognized text and a rebuilt PDF. Proofread critical numbers — OCR confuses 0/O, 1/l, and 5/S on rough scans. For archival scans, keep the original alongside the text version.
Steps
- Upload the scan or photo of the document
- Select the document language
- Click Run and wait for recognition
- Download the text file, the PDF, or both
Pro tips
- 300 DPI straight scans beat phone photos of curved pages.
- Batch-unlock first: OCR needs readable, unprotected files.
Frequently asked questions
Which languages work?
English, Spanish, French, German, Italian, Portuguese, Dutch, Russian, Arabic, Hindi, Chinese (Simplified), Japanese, Korean, and Turkish.
Does OCR work without internet?
After the first use, the engine and language packs cache locally — then yes, fully offline.
My scan is skewed or faint. Will it work?
Clean, straight, high-contrast scans recognize far better. Re-scan crooked pages if accuracy matters.