PDF Tools
Repair PDF
Try to recover a damaged or corrupt PDF — rebuild its structure, or as a last resort re-render readable pages into a new PDF.
How to use
- 1Add the damaged PDF.
- 2Press Repair. If a quick rebuild isn't possible, readable pages are re-rendered as images.
Frequently asked questions
Can every PDF be repaired?
No — if the file is badly truncated or overwritten, the missing parts can't be recovered.
Why is the repaired file not searchable?
If the deep repair had to re-render pages as images, text can't be selected. Use OCR PDF to make it searchable again.
Is my file uploaded?
No — repair runs in your browser.
Two different repairs, tried in order
This tool doesn't have one repair method — it tries the cheap fix first and only falls back to the expensive one if that fails. Step one loads the file leniently (tolerating invalid objects and a broken cross-reference table) and immediately re-saves it, which rewrites the internal object table and xref from scratch. A surprising number of 'broken' PDFs — the ones that fail to open in a strict viewer but are only cosmetically damaged — are fixed by this alone, and the output keeps real, selectable text because nothing was re-rendered. If that load throws, the tool falls back to deep repair: it opens the file with pdf.js page by page, renders whatever pages are still readable to a canvas, and rebuilds a brand new PDF out of those canvas images.
Why a deep-repaired file can't be searched
The deep-repair path only has pixels to work with — it never recovers the original text objects, so the new PDF it builds is a stack of JPEG images, one per recoverable page, compressed at quality 0.9. That means every page that made it through will look right, but none of the text is selectable, searchable or copyable. The tool's own result note says so directly, and the fix is a second step: run the output through OCR PDF afterward to add a searchable text layer back on top of the images.
What repair can't bring back
Both repair paths depend on the file having enough intact structure to parse at all. If the file is truncated mid-write, has had chunks overwritten, or the central directory is gone entirely, pdf.js can't open it and there's nothing to render — the tool reports that no readable pages were found. Partial damage is different: if 8 of 12 pages are readable and 4 are corrupted beyond recovery, the deep repair returns a 8-page PDF and tells you '8 of 12 pages recovered' — it does not fail the whole job for a few bad pages, but it also can't reconstruct content that isn't there.
Example: a contract PDF that won't open
Say a signed 12-page contract was emailed as an attachment, got corrupted in transit, and now shows a 'failed to load' error in every PDF reader you try. Dropping it into Repair PDF first attempts the structure rebuild — if the file's actual page data is intact and only the xref table is damaged (a common cause of 'file could not be opened' errors), you get back a 12-page PDF in seconds, with all text still selectable. If the structure rebuild fails because the damage runs deeper, the tool automatically retries with the deep repair, rendering each of the 12 pages as images; you might end up with '10 of 12 pages recovered' if two pages were unreadable, giving you back most of the contract as a flattened, non-searchable PDF you can still read, print and sign.
Related tools
- Straighten & Clean Scanned PDF (Deskew)
Automatically straighten crooked scanned pages and whiten the paper background — fix tilted scans, remove gray noise and shrink the file with grayscale. Runs in your browser.
- PDF Revision History Viewer
Find the earlier versions hidden inside a PDF that was edited or signed, see what changed in each (text, pages, comments, form values) and which version each signature covers, and download any version.
- Rotate PDF
Turn every page, or just the ones you choose, by 90°, 180° or 270° and save the fixed PDF.
- Unlock PDF (Remove Password)
Remove the password from a PDF you have access to — enter the password once and download a copy that opens without it.
- Email to PDF Converter (EML, MSG)
Open .eml and Outlook .msg emails without Outlook and save them as PDF: sender, recipients, date and subject at the top, the message with its inline pictures, and the attachments embedded in the PDF. Convert hundreds at once into one bookmarked PDF.
- Jupyter Notebook to HTML & PDF
Turn a Jupyter notebook (.ipynb) into a clean web page or PDF without installing Jupyter: Markdown with typeset maths, highlighted code, tables, plots and coloured error messages. Hide the code for a report-style version.