Repair PDF

Try to fix a damaged PDF or recover its readable pages.

Enable JavaScript in your browser to use this tool.

Repair PDF help

Try to repair a damaged PDF or recover pages that can still be read. Results depend on which parts of the file survive; missing content cannot always be reconstructed.

How to use Repair PDF

Choose the damaged PDF and read the page-recovery information.
Select Repair PDF to try a structural repair while preserving readable page content.
If needed, select Recover pages as pictures, then Recover readable pages; review any omitted pages.
Select Download PDF when a result is available and inspect every recovered page.

Options explained

Repair the PDF: Try a structural repair first when the file can be parsed; inspect the result rather than assuming all content was restored.
Recover pages as pictures: Build an image-only copy from readable pages if repair is unsuitable. Unreadable pages are reported and omitted.
Save an edited copy. Its digital signature will no longer be valid.: Shown for signed PDFs. Saving changes invalidates the original digital signature; keep the signed source separately.
Process all files: Choose multiple inputs for repair pdf. This runs the settings above separately for each file, shows individual results and failures, and offers Cancel file, Cancel all files and Download batch as ZIP. Check that page positions or form fields suit every input.

Limits and supported files

PDF documents. An opening password is required when the file needs one to be read.

Local tests used synthetic 10, 50, 100 and 200 MB PDFs containing large images. Chromium, Firefox and WebKit: completed 10, 50, 100, 200 MB.

Repair supports documents with up to 300 pages. Chromium, Firefox and WebKit: completed 50, 300 text pages; 1000 text pages did not complete.

Repair PDF was tested on a 16 GB Mac with Chromium, Firefox and WebKit. “Did not complete” includes input limits, errors and test timeouts; try a smaller selection. These are observed examples, not guaranteed limits on your device. WebKit tests do not replace a check in Safari on your own device.

Up to 100 input files and 512 MB total; up to 200 output files or 512 MB. ZIPs over 128 MB need Chrome/Edge direct disk saving; use individual downloads in other browsers. The Workflow stress test completed 100 synthetic files in all three engines; this does not guarantee the same capacity for every conversion.

Questions

Can every damaged PDF be repaired?

No. A file with missing page data or an ambiguous page order may not be recoverable. Keep another copy if available. The tool reports pages it cannot reliably recover instead of promising a complete document.

What does recovering pages as pictures mean?

That option creates a new PDF from readable page images. Selectable text, interactive forms, links and bookmarks are not retained, and unreadable pages are omitted as reported. Review the recovered-page list before using the result.

Should I try repair before recovering pages as pictures?

Try a structural repair first when the file can be parsed; inspect the result rather than assuming all content was restored.

What happens if some pages cannot be recovered?

Build an image-only copy from readable pages if repair is unsuitable. Unreadable pages are reported and omitted.

Your files are processed on your device. Verify it yourself. Read our privacy policy for details about website analytics and saved settings.