A LITTLE HELP. A LOT DONE.
Good tools.
Great flow.

Your shortcut to the everyday. Create, convert, calculate,
and get back to what you love.

✓ Free to use✓ No sign-up✓ 500+ possibilities
PDF

PDF Repair

Best-effort recovery of corrupt or unreadable PDFs by re-parsing and re-saving

Loading interactive tool…

About this tool

PDF Repair attempts to recover a damaged or partially corrupt PDF that fails to open in normal readers. The tool loads your file via pdf-lib in lenient mode, traverses the object tree, resets the metadata, and writes a fresh, well-formed PDF — often re-emitting object streams to compact and de-duplicate. This works for common damage classes: truncated downloads, broken cross-reference tables (xref), partially-corrupted metadata, files saved by buggy generators, or PDFs edited by a tool that did not flush properly. It cannot recover files where the actual page content streams are missing or the PDF is encrypted with a lost password. Everything runs locally in the browser — no upload, no server. Always keep the original damaged file before downloading the repaired copy.

Why use it

  • Best-Effort Without Software Install: No need to install qpdf, MuPDF, or Ghostscript locally — everything runs in the browser.

  • Two-Stage Parse: Strict parse first for clean files, lenient fallback for damaged ones — gives every chance of recovery.

  • Object Stream Re-Emission: Compacts and rebuilds the cross-reference table, fixing the most common corruption class.

  • Safe by Default: Always keeps your original — the repaired file downloads as a separate 'repaired' copy.

  • Live Diagnostic Log: Every step is reported — header check, parse mode, page count, save size — so you understand what the tool actually did.

  • 100% Local: Confidential or sensitive damaged documents never leave your browser, even when broken.

How to use

  1. Upload the damaged or unreadable PDF — up to 50 MB, processed entirely in your browser.
  2. The tool sniffs the file header and reports basic findings in the log panel.
  3. Leave 'Ignore minor errors during parse' on (default) — it lets pdf-lib salvage objects with small inconsistencies.
  4. Leave 'Re-emit object streams' on — this compacts and de-duplicates the saved PDF.
  5. Click Repair PDF — the tool first tries strict parse, then falls back to lenient mode if needed.
  6. If repair succeeds the recovered file downloads automatically; verify by opening it in your normal PDF reader.

When it helps

  • When a PDF download was interrupted and the resulting file fails to open with 'corrupt' or 'damaged' errors.
  • When a tool's buggy export produces PDFs that crash Adobe Reader but show fine in other viewers.
  • When a cloud sync (Dropbox, Google Drive) flagged a PDF as broken and the original is no longer available.
  • When merging or editing a PDF in another tool produced a result that fails to open.
  • When you receive a PDF attachment that returns 'PDF structure error' in your reader.
  • Before reporting a bug to a PDF generator vendor — re-save through pdf-lib and see if the issue persists in the cleaned copy.

Examples

Truncated download

Input

report.pdf (1.2 MB) — interrupted Wi-Fi mid-download, fails to open in Acrobat

Expected result

report-repaired.pdf — pdf-lib re-emits the xref, file opens cleanly with all 14 pages intact

Buggy generator output

Input

invoice.pdf (88 KB) — produced by an old library, opens in Chrome but crashes Preview

Expected result

invoice-repaired.pdf — clean PDF that opens in every reader, identical visual content

Edited and resaved

Input

manual.pdf (3.4 MB) — edited in three different tools, now flagged as 'PDF structure error'

Expected result

manual-repaired.pdf — single clean object table, slightly smaller, all pages preserved

Tips

  • Always keep the original damaged PDF until you have verified the repaired copy works in your preferred reader.
  • If repair fails, run the file through qpdf locally: qpdf --check broken.pdf — it gives the most detailed structural diagnostic.
  • Files damaged at the page-content-stream level (missing graphics, missing text) cannot be recovered — only structural damage is repairable.
  • If the damaged file is encrypted, run PDF Unlock first if you have the password; this tool cannot guess passwords.
  • If the result is smaller than the original, the tool likely de-duplicated unused objects — that is normal and the content is still intact.

Frequently Asked Questions

Will every corrupt PDF be repairable?
No. PDFs with damaged page content streams (missing text, missing graphics) cannot be recovered — only structural damage like broken xref or metadata is repairable. Roughly 70–80% of 'PDF won't open' errors fall into the repairable category.
Is my file uploaded to your server?
Never. pdf-lib runs entirely in your browser, even on damaged files. UtilityKit cannot see the content of your PDF.
What if repair fails?
Try qpdf locally with qpdf --check broken.pdf for a detailed diagnostic. Some PDFs are damaged beyond browser-side recovery.
Will the repaired file look identical?
Yes — for repairable files, content is preserved byte-for-byte at the object level. Only the xref table and metadata are rebuilt.
Why is the repaired file smaller than the original?
The tool de-duplicates unused objects and re-emits compressed object streams — that is normal and the content is intact.
Does this work on encrypted PDFs?
Only if you have the password — run the file through PDF Unlock first, then through PDF Repair if the unlocked copy is also damaged.
Can I batch-repair many PDFs at once?
Not in the current UI — process them one at a time. For bulk repair, scripting via pdf-lib or qpdf locally is more efficient.
What does 'Ignore minor errors' actually do?
It tells pdf-lib to skip over malformed individual objects rather than aborting the whole parse — letting the rest of the PDF be salvaged when one object is broken.

Explore the category

Merge PDFs Merge multiple PDFs into one PDF Redact — Black Out Text & Regions Permanently redact sensitive text and regions in PDFs with black rectangles PDF Combine & Reorder Drop multiple PDFs, drag thumbnails to reorder every page, then save one combined file. Visual page-level control. PDF Organizer Reorder PDF pages by a custom comma-separated list Split PDF Split a PDF by page ranges Compress PDF Reduce PDF file size
PDF Accessibility: Making Documents Screen-Reader Friendly Most PDFs are accessibility nightmares — untagged content, no reading order, images without alt text. Making a PDF accessible is not difficu Apr 2026 · 8 min How PDF Works: Inside the World's Most Portable Document Format A PDF looks like a printed page but is actually a complex container format with embedded fonts, compressed image streams, and a cross-refere Apr 2026 · 9 min PDF vs DOCX: When to Use Each Document Format PDF and DOCX both contain text and images, but they are designed for fundamentally different purposes. Choosing the wrong format causes fric Apr 2026 · 7 min

Glossary

Cross-Reference Table (xref)
A table at the end of a PDF that maps object numbers to byte offsets — the most common corruption point.
Object Stream
A compressed container holding multiple PDF objects together — re-emitting it compacts the output.
Lenient Parse
Loading mode that ignores minor structural inconsistencies and tries to salvage usable objects — what this tool falls back to.
Page Content Stream
The drawing instructions inside a page — if these are corrupted, the visual content cannot be recovered.
Producer
Metadata field naming the tool that wrote the PDF — reset to 'UtilityKit PDF Repair' on the recovered copy.
Truncated File
A PDF that is missing bytes at the end (often the xref table) due to interrupted download — usually repairable.