Toolpit
Loading your page…
Toolpit
Loading your page…
Rebuild a PDF that will not open.
No file to hand?(sample-damaged.pdf · made-up data)
The file is re-read as leniently as possible and written out with a fresh structure. That fixes a broken cross-reference table, a truncated trailer and objects a strict viewer refuses. Content that is genuinely missing cannot be rebuilt — the result will say if so.
Most “corrupt” PDFs are structurally broken rather than lost — a damaged cross-reference table, a truncated trailer, objects a strict viewer refuses. This reads the file as leniently as possible and writes it out again with a clean structure, which is what fixes the common failures. It is honest about the rest: content that is genuinely missing cannot be rebuilt, and the result says so.
Drop in the PDF that will not open.
Click Repair — the file is re-read as leniently as possible and written out with a fresh structure.
Download the rebuilt file, and read the notes about anything that could not be saved.
No. Toolpit does not upload your file for processing — the tool runs as code in this browser tab, works on the copy your browser opened, and forgets it when you close the tab. You can check this yourself: open your browser’s network panel and run the tool. You will see the page, its scripts and the ads on the site, and — when you save, if you are signed in — one small request that charges 10 tokens to your account. None of them carries your file, its name or anything read from it, because file processing stays on your device.
Most PDFs that refuse to open are not missing their content — they are missing the map to it. A PDF ends with a cross-reference table listing where every object sits in the file, and if that table is damaged, truncated or disagrees with reality, a strict reader gives up even though the pages are all still there. Repair ignores the broken map, scans the file for the objects themselves, and writes a new document with a correct one.
Not completely, and no tool honestly can — the bytes that never arrived do not exist anywhere on your device. What happens instead is that whatever did arrive is salvaged: if the structure is too damaged to rewrite, the pages that can still be drawn are recovered as images, so you keep the ones that came through. The result says so plainly rather than pretending the file is whole.
That is the fallback, and it only happens when the document is past a structural rewrite. When the usual repair cannot parse the file at all, the pages are rendered and rebuilt as pictures — they look the same and they print the same, but the text layer is gone, so you cannot select or search it. Running the result through Image to text will give you the words back if you need them.
Every one of these runs in your browser too — no upload, no watermark.