One document, two local paths
Searchable pages keep their embedded text and layout. Image-only pages use the same private PP-OCRv6 engine as BrowserOCR’s image tool.
Private PDF OCR
Extract native text instantly and recognize scanned pages only when needed—all locally, without uploading your document.
Text and Markdown stay on this device.
Searchable pages keep their embedded text and layout. Image-only pages use the same private PP-OCRv6 engine as BrowserOCR’s image tool.
Process the whole PDF or select pages such as 1-20,35. Large-document guidance helps mobile browsers stay within their memory limits.
Common questions
No. PDF inspection, native text extraction, scanned-page OCR, passwords, and results stay inside your browser.
Yes. Native pages use their embedded text while scanned pages are rendered and recognized locally with PP-OCRv6.
Yes, after this page, its PDF decoders, and the selected OCR models have loaded and been cached once. A new decoder or model still needs a connection the first time.
Choose one PDF up to 100 MB and 300 pages. You can process all pages or enter a smaller page range.
Yes. The password is held only in the PDF worker memory for the current document and is never stored.