Repair PDF: Fix Corrupted PDF Files Online Free
Rebuild broken cross-reference tables with qpdf, fall back to raster salvage when structure is beyond saving, and read a recovery report that measures pages and objects before and after.
or click to browse
How to Repair a Corrupted PDF
One file, one button, two repair strategies:
- Drop the damaged file - The tool scans the header, cross-reference pointer, page tree, and object count straight from the bytes before touching anything.
- Press Repair PDF - First ladder: qpdf rebuilds the cross-reference table and re-emits a normalized file, keeping text and links. If qpdf cannot parse it, the second ladder rasterizes every page pdf.js can still read.
- Read the report and download - Pages detected and recovered, object counts before and after, xref status, and outline status, followed by the repaired download.
Rebuild, Raster Salvage, and What Each Preserves
Different damage calls for different repairs. Knowing which ladder ran tells you what the output can still do:
- Cross-reference rebuild - Fixes the table of contents of the file: object offsets, trailer, startxref pointer. Text stays selectable, links stay clickable, vectors stay crisp. This is the normal outcome for interrupted downloads.
- Raster salvage - For files whose structure cannot be parsed at all: each readable page renders to an image and becomes a new page. Content returns; the text layer does not (run OCR afterward if you need search).
- Object counts - The report counts every object marker such as 12 0 obj before and after, so you can see whether normalization consolidated or expanded the structure.
- Page accounting - Detected pages come from a tolerant pdf.js read of the original; recovered pages are what the output actually contains. A mismatch is called out in the report.
- Outline status - Bookmarks are checked in both files, so a dropped outline is reported instead of silently lost.
- Honest limits - Content deleted by the damage itself cannot be invented back. The report separates structural damage (repairable) from missing data (not).
Fix PDF Files Without Uploading
A file that will not open is stressful enough without handing it to a website. This page keeps the whole recovery local: qpdf compiled to WebAssembly rebuilds structure, pdf.js provides the tolerant read and raster fallback, and pdf-lib assembles salvage pages. When the file opens again, continue with Unlock PDF if it was protected, PDF Tools for splitting and merging, or PDF to JPG for a visual copy. For site-wide checks, run the Free SEO Audit Tool or book our SEO audit services.
Frequently Asked Questions
How do I repair a corrupted PDF?
Drop the damaged file and press Repair PDF. The tool rebuilds the cross-reference table with qpdf; if that cannot read the pages, it falls back to raster salvage, which preserves every readable page as an image. The recovery report tells you exactly which path ran.
What does the recovery report show?
Which repair ladder was used, how many pages were detected versus recovered, object counts before and after, whether the cross-reference table was broken, and the status of the outline. Everything is measured from the actual bytes, not guessed.
What is the difference between rebuild and raster salvage?
A rebuild restores the document's structure (cross-reference table, object offsets) while keeping text, links, and vectors intact. Raster salvage renders each readable page to an image and rebuilds the PDF from those images: content comes back, but the searchable text layer does not.
Why does a PDF become corrupted?
Common causes are an interrupted download, a storage device pulled mid-write, a broken email attachment transfer, or software that crashed while saving. The header and cross-reference table are the usual casualties.
Is my damaged file uploaded?
No. Both the qpdf rebuild and the pdf.js fallback run inside this browser tab. The file never leaves your device, and the repaired copy downloads directly.
Will the repaired PDF keep its text and links?
On the rebuild ladder, yes: text, vector content, and links survive because only structure is repaired. On raster salvage, pages become images, so text is no longer selectable unless you run OCR afterward.
Can this repair password-protected PDFs?
If the file opens, it can be repaired. If it needs a password, unlock it first with the Unlock PDF tool, then repair the decrypted copy here.