Recover what is still there.

Diagnose suspicious PDF structure, try a semantics-preserving rewrite first, then fall back to page-by-page visual salvage when necessary. Every output is reopened, sampled and verified before it is offered for download.

Drop damaged PDFs here

Up to 20 PDFs · 500 MB/file browser safety limit · source documents stay in this browser

Documents

Select to repair

Recovery strategy

Auto · recommendedTry structure-preserving rewrite first, then visual salvage if necessary. Structural rewriteRe-serialize readable PDF objects while retaining searchable/vector content and semantics. Visual salvageRender readable pages and rebuild a clean image-based PDF when structure cannot be trusted.

Try stripping non-PDF bytes before a displaced PDF header, only in the fallback candidate. Trim obvious bytes after the final EOF marker in the fallback candidate. Append a missing EOF marker to the fallback candidate.

Salvage quality

Page encoding

JPEG quality 88%

Grayscale salvaged pages

Allow partial recovery if individual pages cannot render. Failed pages are listed in the receipt.

Original files are never overwritten. A repair receipt is generated for every successful output.