Repair PDF
"The file is damaged and could not be repaired" usually means the index broke, not that the pages are gone. This rebuilds the index in your browser — the file is not uploaded.
How to repair a PDF
- Choose the damaged PDF.
- Read what it found: the index, the ending, the header.
- Press the button and the repaired file downloads.
Why a PDF stops opening
A PDF keeps a table at the end saying at which byte every object starts, and readers go straight to it. Break that table and the reader is lost, although every page is still inside the file. It breaks when a download stops early, when a program crashes while saving, when a file is copied from a phone that ran out of space, or when a tool moves things and forgets to write the new positions.
What repairing actually does
It walks the whole file looking for the objects themselves, writes the table again from what it finds, and points it at the document's catalogue. Objects packed inside compressed bundles are unpacked so the new table can reach them. Nothing is re-encoded: the pages, the text and the images are copied exactly as they are, so a repaired file is the same document, not a photograph of it.
When the pages can only be saved as images
If the file is damaged past rebuilding but its pages can still be drawn on screen, they are drawn into a new PDF. The text stops being selectable, which is a real loss, and it is said plainly when it happens — a document you can open beats one you cannot. A file encrypted and damaged usually ends up here, because the pages cannot be read without undoing the encryption first.
Is it private?
Yes. Everything runs inside your web browser, using your own device. Nothing is uploaded to any server, so your documents stay completely private. How does this work without uploading?
Frequently asked questions
Are my files uploaded?
No. The repair runs in your browser. You can disconnect from the internet after the page loads and it still works.
Will the text still be selectable?
Yes, whenever the file can be rebuilt — that path copies the objects untouched. It is only the last resort, redrawing the pages as images, that loses the text, and the page says so when that is what happened.
It says nothing can be recovered.
Then what arrived is probably not a PDF any more: a download that brought back an error page, or a file whose first part is missing. Try downloading it again from wherever it came from; half a PDF can be repaired, but a file with no objects left in it cannot.
Does it work with a password-protected file?
Partly. Restrictions are removed as part of the rebuild, but a file that needs a password to open cannot be read without it, so at best its pages are saved as images.