Compress a PDF - Free, No Upload, No Watermark
Shrink a scanned document or image-heavy PDF by re-encoding its embedded photos at a lower quality, without touching the text layer.
Maintained by Hirantha Sandaruwan Lakshan. Updated 31 July 2026.
About this tool
Everything happens in your browser tab - the file is never uploaded anywhere.
How it works
Every image embedded in the PDF is found and re-encoded at your chosen quality level, then swapped in place of the original - the page layout, text, and structure are left untouched.
A PDF that's mostly text rarely has much to gain this way - structural cleanup alone only returns a percent or two, so this tool is honest when a file is already about as small as it can get, rather than reporting a trivial saving as a win.
Each image is only replaced if the re-encoded version is actually smaller - if a photo is already efficiently compressed, the original is kept rather than risking a worse result.
Worked examples
A scanned document email attachment rejects
Multi-page scans are usually just a sequence of full-page photos - re-encoding them at Balanced or Smallest quality can cut the file size dramatically with little visible difference.
A PDF export full of screenshots or photos
Reports and slide exports with embedded photos often carry them at unnecessarily high quality - recompressing shrinks the file without touching the surrounding text.
A text-only PDF that's already small
Contracts, forms, and other text-heavy PDFs typically have little or nothing to gain here - the tool says so directly instead of claiming a saving that isn't meaningful.
Quality presets at a glance
| Preset | JPEG quality | Typical saving on a scan | Use it for |
|---|---|---|---|
| High quality | 82% | Modest | Documents that still need to look sharp when zoomed or printed |
| Balanced | 60% | Substantial | The default - a clear size drop that still reads cleanly on screen |
| Smallest file | 40% | Largest | Archives and email attachments where getting under a limit matters most |
Frequently asked questions
Will this make my text blurry?
No - only embedded images are re-encoded. Text drawn directly on the page (the normal case for anything not scanned) is untouched.
Why didn't my PDF get any smaller?
If a PDF has no embedded images, or its images are already efficiently compressed, there's little to save - the result says so rather than reporting a negligible change as a success.
Does this work on password-protected PDFs?
Yes - enter the password when prompted and the file is unlocked before its images are recompressed.
Which images actually get recompressed?
Images already stored as JPEG inside the PDF. Some scanners and tools store images in ways this can't safely re-encode, in which case they're left as-is rather than risking a broken result.
How much smaller will my PDF get?
On a scan or a photo-heavy export, commonly 60-90%. On a text-only PDF, close to nothing - and the result says so plainly rather than dressing a 1% change up as a success.
Does this reduce the page dimensions?
No. Pages keep their size and layout exactly; only the images embedded in them are re-encoded.
Can I compress a scanned document?
That's the case it's built for. A scan is effectively a stack of full-page photographs, which is precisely where re-encoding pays off most.
Is there a page or size limit?
No fixed limit, but every image is decoded and re-encoded in turn, so a very long scanned document takes proportionally longer and is heavier on memory.
Sources & references
Updated Jul 31, 2026