Recognize text from scanned, image-based PDFs locally in your browser — multi-language, page range selection, TXT export. No upload to any server, completely free, no signup.
How to use
1
Upload a scanned PDF
Upload a scanned or image-based PDF, then pick the language and page range
2
Local recognition
Pages are rendered and OCR'd one by one in your browser — nothing is uploaded, progress is live
3
Copy or download
Results stream into the preview — copy or download as TXT, no login needed
FAQ
Why does each page take 5-15 seconds?
The OCR engine runs entirely inside your browser (no server involved). Each page needs rendering, preprocessing and recognition, so speed depends on your device. The first run for a language also downloads its language data once (cached afterwards).
How is this different from PDF to Text?
PDF to Text reads the file's built-in text layer — instant, but scans don't have one. This tool reads the page images via OCR, built for scanned and photographed documents. The two tools complement each other.
How accurate is it?
Clean printed text recognizes well; handwriting, blurry photos and complex tables reduce accuracy. Pick the right language, scan at 300 DPI, and proofread key numbers and names afterwards.
Are my files uploaded to a server?
No. Rendering and recognition happen entirely in your browser. Files never leave your device and memory is cleared on refresh. Limits: 30 MB and 30 pages per file.
You decide what tools we build nextSubmit wish →