PDFGym
EN

Repair a PDF

or drop one here
Choose a PDF file first.

Choose a PDF that will not open and get back a fresh copy with -repaired in its name, written from whatever the library can still read. The rebuilt file is reopened and its pages, sizes and text are compared with the readable source before the download appears. Truncated, encrypted and non-PDF files are refused with a plain explanation instead of a fake fix.

The attachment that arrived broken

This page rebuilds structure, not content: a file whose pages survive but whose index broke. If the reader reports damage on a complete download, the bytes may still be fine.

Say a client emails Invoice-Q3-final.pdf on Monday: 2 MB, looks complete, but your reader calls it damaged and refuses to draw a page. The usual cause sits at the end of the file, where the index table maps every object to its byte offset; one bad entry and the reader has no map. You drop the file here and press Repair PDF. The library opens what it can, writes a fresh copy with a correct index, and the reopen check compares pages, sizes and text before Download repaired PDF appears.

Read the symptom before you repair:

Symptoms and what they mean
SymptomLikely causeMove
Reader reports damage, size looks rightBroken index tableRepair here
File tiny, download stopped earlyTruncated bytesDownload again first
Opens in one reader, not anotherStrict structure checksRepair, retry the strict one
Password prompt, then failureEncryption, not damageUnlock PDF first

Two refusals are final in this tab. Truncated files end mid-object, so there is nothing to rebuild: fetch the complete file first. Encrypted files scramble the structures the repair reads, so open them with Unlock PDF before repairing. Non-PDF uploads are turned away at the door.

Four steps from broken file to fresh copy

  1. Choose the damaged PDF

    Press Choose a PDF or drop the file onto the page. The original is only read, never changed.

  2. Press Repair PDF

    One button starts the rebuild: the library opens the damaged file and writes a fresh copy of whatever still opens.

  3. Wait for the output check

    The rebuilt file is opened again and compared with the readable source: page count, page sizes and selectable text.

  4. Download the repaired PDF

    Download repaired PDF appears only after the checks pass. The copy carries -repaired in its name.

Repairs that fail, and what they mean

This PDF is too damaged to open means truncated or gutted bytes, not a wrong click. Re-downloading fixes most of these; a second identical failure means the source itself is broken, so ask the sender to re-export.

A repair that succeeds but shows fewer pages than expected recovered only what was readable. Compare against the sender's page count before trusting the copy, and keep the original until the rebuilt file proves itself in your reader.

Repairing the wrong problem wastes a run: garbled text from broken fonts and missing characters are content damage, which a structural rebuild cannot fix. Fix PDF fonts, missing text, wrong characters covers those symptoms. Locked files stop before the rebuild with Unlock this PDF before repairing it; open them with Unlock PDF first.

When repair is not the fix

A file that opens but misbehaves, wrong glyphs, blank pages, missing images, needs content surgery, not a fresh index. Repair the structure first if readers complain, then chase the content fault with the guide above.

When the original source still exists, re-export beats repair: a fresh export from the word processor or the scanner carries no damage at all. After a good rebuild, clean the file with Remove PDF metadata and shrink it with Compress PDF before it travels on.

Short answers about repairing PDFs

Will repair change my pages?
No. It rewrites the file structure around the content it can read; pages, text and images pass through. The reopen check compares all three with the readable source.
The repaired file is smaller. Did I lose pages?
Not necessarily: a fresh index drops dead bytes the old file carried. Compare the page count with the source; if it matches, the diet was dead weight.
Can it fix a file that is half downloaded?
No. Truncated files end mid-object with bytes missing, and no tool invents them. Finish the download first, then repair the complete copy if needed.
Does it touch the original?
Never. The original is read only; you download a separate -repaired copy. Keep both until the new file proves itself.
My file opens but shows wrong characters. Will this help?
No, that is font or content damage inside readable pages. See Fix PDF fonts, missing text, wrong characters instead; structure is already fine.

Tools for the file after the rescue

Each tool below runs in the browser with no account and checks its own output first. Rescue here, then clean, shrink or unlock the fresh copy.