Repair PDF

Add one damaged file or up to ten at once. Before anything is uploaded, your browser takes a quick look at each file and tells you whether it opens here and how many pages it has, or that it does not open at all. Then the files go to our server over an encrypted HTTPS connection, one after the other, where the repair runs.

Each file comes back as a new PDF with a report: how many pages were recovered, which method did the work, what problems were found and how the size changed. Your files are deleted from the server right after. The aim is simple: repair PDF documents that matter to you, honestly, and tell you plainly when there is nothing left to save.

A lot of repair tools hand you a file and say nothing. This pdf repair tool online writes a report for every file, so you know what you got. The first line reads like this: 12 of 12 pages recovered, Structure rebuilt (MuPDF), 2.4 MB to 2.1 MB.

Under it is the list of problems that were found. The common ones are: no PDF header, end of file missing (the file looks cut off), cross-reference table missing, broken cross-reference table, and a number of pages that could not be drawn. If the file turned out to be fine, the report says no damage found; the file was cleaned and rebuilt.

The method line uses four plain names: Structure rebuilt (MuPDF), Structure rebuilt (qpdf), Re-made from the page content (Ghostscript) and Pages rescued one by one. That tells you at a glance whether your links and form fields are likely to have survived.

A PDF is not one block of data. It is a set of numbered objects, such as pages, fonts, pictures and text streams, plus a table at the end of the file called the cross-reference table. That table lists where each object starts. A PDF reader jumps to the end of the file first, reads the table and then fetches the pages.

Most corruption hits exactly that weak point. When the end of the file is cut off, or the byte positions in the table no longer match, a strict reader gives up straight away, even though every page may still be in the file. To repair corrupted pdf files of this kind, the tool scans the whole file for objects, writes a fresh table and saves a clean copy. This is what Quick fix does, and it is the first step of Automatic.

Other damage sits inside the objects themselves: a compressed text stream with a few wrong bytes, a font that is half missing, a picture that ends early. A new table does not help there. Deep repair and Rescue pages are made for those cases.

Sometimes a file opens, but some pages are blank, show an error, or stop the reader when you scroll to them. The page objects are there, but their content cannot be drawn. The browser check and the server health check both count how many pages draw without an error, and the report lists how many could not be drawn before the repair.

To repair damaged pdf pages, Automatic first tries the structure rebuild. If some pages still fail, Ghostscript re-makes the file from what it can read, and Rescue pages copies every page that can be copied one at a time. A page whose content is broken beyond copying is saved as a picture of what it shows, so at least you can read it again.

If you already know which kind of damage you have, pick the method yourself. Deep repair suits files with font errors and garbled text; Rescue pages suits files where a few pages are lost and you simply want everything else back.

Some files do not open at all. Your reader shows an error, a blank window or nothing, and the file looks useless. The browser check on this page tells you the same thing honestly: Does not open, often with a hint such as no PDF header or end of file missing.

A file that does not open in a reader is very often still repairable, because readers are strict and repair engines are patient. MuPDF and qpdf read the file from the start, find the objects that are still intact and build a working file around them. So the first step to repair pdf that won't open is simply to add it here and leave the method on Automatic.

When the server cannot get a single page out of the file, you will see Not repaired with the message This PDF could not be repaired. That means the data needed for the pages is not in the file. The section below explains why that happens.

A PDF can look fine on screen and still fail at the printer. The print job stops on one page, pages come out blank, fonts turn into boxes, or the printer reports an error and prints nothing. Printers and print shops often use stricter software than your screen reader, so small structural errors show up there first.

To repair pdf for printing, try Deep repair. Ghostscript draws every page again into a brand-new PDF, with the fonts and pictures written out cleanly, which is very close to what a print workflow wants. Check the result on screen, then print. If you only need to send the file to a printer, losing form fields usually does not matter.

If the problem is only one bad page, Automatic or Rescue pages may give you a cleaner result. Delete Pages can then remove a page you do not need, and Compress PDF can make a very large print file easier to send.

Repair can rebuild structure, but it cannot invent data. If a download stopped at 40 percent, the last 60 percent of the file never reached your device. The pages stored in that part are gone, and no tool can bring them back, because the information is not in the file anymore. The repaired copy will contain the pages that were in the part that arrived.

The same is true for files that are mostly zeros, which is what you often get after a failed copy to a USB stick or a damaged disk. A file can have the right size and name and still hold nothing but empty bytes. The result box then says These files could not be repaired, and the best fix is to get a fresh copy from the sender, the original website or a backup.

Deep repair also has a price: it re-makes the file from its page content, so form fields and some links are lost. Bookmarks may not survive either. If those matter, try Quick fix first and compare.

When Rescue pages, or Automatic, finds a page whose content cannot be copied into the new file, it does not just drop the page. If the page can still be drawn, it is saved as a sharp picture of what it shows. You can read it, print it and send it, but you cannot select or search the text on that page.

The report tells you when this has happened. To make those pages searchable again, run them through PDF to Text, which reads the words from the page with OCR. OCR is good but not perfect, so check names, numbers and amounts against the picture.

A PDF with a password to open it cannot be repaired until it is unlocked, because the content is encrypted and the repair engines cannot read it. The browser check shows Password-protected for such files, and the server answers with: This PDF has a password. Unlock it first with Unlock PDF, then repair it.

So use Unlock PDF with the password you know, download the unlocked copy, and add that copy here. When the repaired file is ready, Protect PDF puts a password back on it if you need one.

Every method on this page is free: Automatic, Quick fix, Deep repair and Rescue pages, with the full report. There is no paid plan that unlocks better repairs and no watermark on the result. You can repair pdf online free as often as you need, up to 10 files of up to 200 MB each per batch.

The heavy work runs on our server, not on your device, so you can repair PDF files of any size even from an older phone. You only wait for the upload, the repair and the download.

There is no account, no e-mail address and no trial period. Open the page, add the files, press the button and download. That is how a repair pdf free no sign up service should work: you came to fix a file, not to fill in a form.

Your files travel to our server over an encrypted HTTPS connection. They are repaired there and deleted right after, together with every working copy the server made. We do not keep them, read them or use them for anything else, and the repaired PDF goes straight back to your browser.

On Windows, Mac, Linux or a Chromebook, open the page in any up-to-date browser and drag files from File Explorer or Finder onto the box. The repaired copies land in your Downloads folder. Because the server does the work, it is just as quick to repair pdf file online on an old laptop as on a new one.

On an iPhone or iPad, tap Select PDF files and choose the file from Files, iCloud Drive or a mail attachment you saved there. On Android, the picker shows Downloads, Drive and other apps. The repaired PDF is saved to Downloads, where you can open it or share it.

Files without a .pdf ending can be added too. A damaged attachment sometimes arrives with a strange name or type, and the picker accepts it so the server can take a look.

Wait until a download has finished before you open or move the file, and compare its size with the size on the website if one is shown. In a browser, a file still marked as downloading is not complete.

Use Eject or Safely remove before you pull out a USB stick or memory card. Save to a local folder first and copy to the stick afterwards, rather than saving straight onto it.

Send large PDFs as a link from a cloud drive instead of an email attachment, or make them smaller with Compress PDF first. Keep a backup of important documents in a second place, so a broken copy is never your only copy.

With a working file again, Compress PDF makes it small enough to e-mail, and Protect PDF adds a password. Merge PDF joins it with other documents, and Split PDF takes out the pages you want to send on their own.

If you want to check that nothing changed in the text, Compare PDF shows two versions side by side with the differences marked.