Repair PDF

A broken file often cannot be previewed, so the repair screen does not pretend otherwise. Instead of a page preview you get the file itself listed with its name and size, the OCR rebuilt pages switch, the language picker that goes with it, and one clear Repair PDF button. That is the whole decision: repair it as it is, or ask for OCR on any page that has to be rebuilt as a picture.

Take the OCR choice seriously before you press the button, because it is the one setting that changes what you end up with. Leave it off and a rebuilt page comes back as an image you can read but not search. Turn it on and the words in that page are read on your device and written back as real text. OCR takes longer, since every rebuilt page has to be re-rendered and read, so switch it on when the words matter and leave it off when you only need the pages back.

The tool works from the gentlest fix to the most thorough, stopping as soon as one succeeds so it keeps as much of your original file as possible:

1. Structural rewrite rebuilds the file's internal index so a readable PDF opens again, with all real text intact.
2. Deep recovery digs into the objects when the structure alone is not enough.
3. Stream recovery decodes and repairs damaged internal streams.
4. Page-by-page rebuild is the last resort: it reconstructs the file one page at a time, keeping real text where it can and re-rendering only the pages it must.

While it works, a status line keeps you posted, ending with something like Recovered 3 of 3 pages as the download starts. The panel that replaces it is the repair report, and it is deliberately short. It repeats how many pages came back out of how many it found, names the fixed file, and tells you two things that matter: whether your real text survived, shown as Text kept, and the size of the repaired file.

The fixed copy downloads on its own, and the name always gains a -repaired suffix, so report.pdf is saved as report-repaired.pdf and your original is never overwritten. If the download did not reach your device, Download again saves it a second time without redoing the repair, and Share passes it to another app on a phone or tablet.

From there the same panel can hand the recovered file straight to another tool without adding it again. Compress PDF brings the size back down after a page rebuild, Organize pages drops any page that came back empty, Merge PDF joins it to a good copy of the missing part, Rotate PDF fixes a page that came back sideways, and Protect PDF puts a password back on if the original had one.

A PDF holds your words as text plus an index that says where everything lives. Most damage hits the index, not the words. When that is all that is wrong, the fix is to write a fresh index around your untouched content, and you get a file that is as searchable and selectable as the day it was made. That is the repair you want, and it is why the report says Text kept.

Sometimes the words themselves are damaged. Then the only way to save a page is to draw whatever can still be read and store it as a picture. You get the page back and it looks right, but the text inside it is no longer text, so searching will not find it. Switch on OCR and those pictures are read back into real text, which is the closest you can get to the original.

So a repair that keeps your text is a light repair, and a repair that rebuilds pages means the damage went deeper. The report tells you which one you got, and that is the single most useful line on the screen.

• Broken or missing cross-reference tables
• Files that will not open at all
• Interrupted or cut-off downloads
• Junk bytes before the header
• Missing end-of-file markers
• Damaged internal streams

These two problems look alike and need different tools, so it is worth telling them apart before you start. A damaged file usually fails on its own terms: your reader shows an error, opens a blank grey page, or says the file cannot be opened because it is not a supported format. A locked file opens far enough to ask you for something, either a password box before you can see any page, or a warning that you cannot print or copy from it.

If you are asked for a password and you know it, Unlock PDF is the right tool. If the file will not open at all, or the pages are there but scrambled, repair is the right tool. One more clue: a file that used to open on this same device and now does not is almost always damaged rather than locked.

Repair recovers what is still inside the file. It cannot recreate data that is no longer there:

• Content that was overwritten or truncated away when the file broke
• Pages whose data was fully deleted, not just mis-indexed
• A password you have lost (use Unlock PDF for a known password)

The repair report always tells you exactly what was saved, so you are never left guessing.

Sometimes the answer is that the data is gone, and the tool says so plainly rather than handing you an empty file. If that happens, a few things are still worth trying, in this order.

• Download or copy the file again from wherever it came from. A cut-off transfer is the most common cause of a PDF that will not open, and a second copy is often whole.
• Check the file size against the source. A file that is far smaller than it should be lost its ending, and the missing part was never saved to your device.
• Look for the program's own recovery copy. Editors and scanners often leave an autosave or a temporary file beside the original.
• Check your email, cloud folder, or phone gallery for an earlier version. A slightly older copy that opens beats a newer one that does not.
• Try the repair once more on a computer if the first try was on a phone. A big rebuild needs memory, and a phone can run short on a very large file.

If none of that works, the file is very likely beyond saving. That is a hard answer, but it is an honest one, and it saves you paying a site that promises otherwise.

When a page is too damaged to read normally, it is rebuilt as an image so you at least get it back visually. Turn on OCR rebuilt pages before you repair, pick your languages, and the tool reads the words in those rebuilt pages on your device, so the recovered text stays selectable and searchable rather than locked in a picture. Leave it off for the fastest repair when a plain visual recovery is enough.

Every repair ends with a short, plain-English report. It shows the recovery method that finally worked, how many pages were recovered, and the size of the fixed file. That way you know whether you got a clean structural fix with all your text intact, or a page-by-page rebuild where some pages came back as images, and you can decide what to do next.

A repaired PDF opens, which is a relief, but opening is not the same as being complete. Keep the broken original until you have checked four things, and the whole check takes about a minute.

• Compare the page count in your reader with the count in the report, and with what the document should have.
• Look at the last page. Damage that came from a cut-off download shows up at the end of a file first.
• Search for a word you know is in there with Ctrl+F, or Command+F on a Mac. If it is found, your text layer survived.
• Skim for pages that came back blank or sideways. Organize pages removes the blank ones and Rotate PDF straightens the rest.

Only when all four look right is it safe to throw the broken copy away. Because the repaired file is saved under a new name, you can keep both side by side while you decide.

Unlike the converters on frisqoo, Repair PDF takes a single file per run, and that is on purpose. Every broken file is broken in its own way, so each one needs its own path through the recovery steps, its own decision about OCR, and its own report. Batching them would hide exactly the information you came for.

Repairing several is still quick. Add the next file when the last one has downloaded and the screen resets itself, ready to go again. Nothing is uploaded on any run, so a folder of damaged files can be worked through one by one without a single byte leaving your device.

Yes. Many repair sites ask you to upload your broken file to their servers, which is risky when that file may hold sensitive data. This tool does all the work inside your browser, so nothing is uploaded, stored, or seen by anyone else. It downloads the fixed copy with a -repaired suffix, so your original is never touched.

• Opening a file that suddenly shows a blank page or an error.
• Rescuing a download that was cut off partway and will not open.
• Recovering an important document a program crashed while saving.
• Getting a readable copy of a scan that another tool corrupted.
• Salvaging as much as possible of an old, partly damaged archive.