Occasionally a PDF goes through the PDF Compressor and comes out barely smaller than it went in. This is not a fault in the tool — it is a sign of what is inside the file. A handful of specific situations cause this, and each has a different fix.
This guide walks through the usual causes in the order to check them, so you spend your time on the fix that will actually work rather than repeating the same compression pass and hoping for a different result.
Cause 1: it has already been compressed
If a PDF was generated by a scanner app or another tool that already compressed the images, there is little redundancy left to remove. Running it through compression again mostly just re-encodes already-compressed data, which does not shrink much and can slightly reduce quality for no gain. If you have access to the original, uncompressed source (the raw scan, or the document before a previous export), start from that instead.
Cause 2: the content is mostly text, not images
Compression targets embedded raster images. A PDF that is genuinely text — exported straight from Word or a website — is usually already small, and there is nothing large to compress away. If a text-only PDF is unexpectedly big, the more likely fix is elsewhere: check for an embedded font set that includes every character and weight when only a couple are used, or a background image or watermark placed behind the text on every page.
Cause 3: the file uses vector graphics or complex layers
PDFs exported from design tools (posters, brochures, CAD-adjacent documents) can contain complex vector paths, gradients, transparency layers and embedded fonts rather than simple raster images. These do not respond to image compression the way a scanned photo does. For files like this, re-exporting from the original design software at a lower quality setting, or flattening the document to a simpler image-based PDF first, usually works better than compressing the vector PDF directly.
Cause 4: it is one enormous image, not a real document
Sometimes a “PDF” is actually a single huge image (a photographed document, or a scan saved at an unnecessarily high resolution) wrapped in a PDF container with almost no compression applied in the first place, and yet it still resists further compression because the image format inside is not one that responds well to re-encoding. Converting the page to a standard JPG with PDF to Image, adjusting it, and rebuilding the PDF with Image to PDF can succeed where directly compressing the original does not.
Cause 5: the file is encrypted or has restricted permissions
A password-protected or permission-restricted PDF often cannot be re-processed until the protection is removed, because the content itself is locked. If the file will not compress at all and you have the password, remove it, compress, then reapply protection with Protect PDF if needed.
A practical checklist
- Confirm you are starting from the original file, not an already-compressed copy.
- Check whether the PDF is genuinely text-based — if so, look at fonts and backgrounds rather than expecting image compression to help.
- If it came from design software, try re-exporting at a lower quality instead of compressing the finished PDF.
- If it is essentially one image per page, convert to images, adjust, and rebuild.
- Remove any password or permission lock before compressing, and reapply it afterwards.
What a stubborn file usually has in common
Across most of the causes above, the underlying pattern is the same: compression can only remove data that is redundant, and a file has already run out of redundancy either because something already removed it, or because the content genuinely does not have much to begin with (crisp vector shapes, dense text, already-optimised images). Recognising which of those applies to your file is more useful than repeatedly trying every setting on the same tool, because it points you straight at a fix that changes the actual amount of information in the document rather than just re-processing the same bytes.
When the honest answer is “split it”
If none of the above moves the number and the file still needs to be smaller, the most reliable remaining option is simply to send less of it. Split PDF to keep only the pages that are actually required removes size in direct proportion to the pages you drop, which is a guarantee that no compression setting can match for a file that has genuinely hit its floor.
Frequently asked questions
Why does the file size barely change even at the strongest setting?
Usually because there is little redundant image data left to remove — the file is already compressed, is mostly text, or is vector-based rather than raster.
Does converting to images and back lose quality?
There is some quality trade-off, but it can still end up smaller and more consistent than the stubborn original, especially for scan-like PDFs.
Is it worth compressing more than twice?
Rarely. If two passes have not moved the number meaningfully, the fix is elsewhere on this list, not a third pass.
Can a corrupted PDF cause this too?
Occasionally. If a file behaves strangely across several tools, try re-saving or re-exporting it from its original source before troubleshooting compression further.
Is the tool free and private?
Yes. No sign-up, no watermark, and files are removed from the server automatically about an hour after processing.