The short answer
If a PDF stops getting smaller no matter how aggressive you get, you’ve most likely hit that file’s compression floor, not a limitation of the tool. A well-structured, mostly-text PDF simply doesn’t have much left to remove once its structure is cleaned up; a scanned or image-heavy one can keep shrinking, but only by trading away visual quality. Try compressing yours to see which case you’re in; the measured result will tell you.
The usual causes, most common first
The document is already lean. A PDF that’s mostly typed text with simple layout — no large images, not heavily edited or merged from other files — doesn’t have much slack to begin with. Structural cleanup (removing unused objects, repacking streams more efficiently) can only remove what’s there, so a file with almost no waste will barely move. That’s expected, not a failure.
Images are already compressed. Re-encoding a JPEG at a similar quality barely changes its size; it only gets much smaller at a noticeably lower quality. If your PDF’s images were already well compressed, a loose target won’t move the file much.
Fonts are fully embedded. Embedding keeps a document looking correct on any device, but embedded font data doesn’t compress the way image data does — it’s often already fairly efficient, so it contributes to the file’s size without giving compression much to work with.
The file carries hidden application data. PDFs exported from design tools like Illustrator or Photoshop with an option like “preserve editing capabilities” checked can embed a full copy of the original source file inside the PDF. You can’t see it on the page, but it can be a large chunk of the file. This tool doesn’t strip that data, so re-exporting is the fix.
It’s a scanned document at very high resolution. This one can still shrink — but only by re-encoding the scanned image at a lower quality, which is a visible trade-off. If a scan won’t shrink without becoming hard to read, that’s the practical limit for that scan at that resolution.
How to tell a file’s limit from a tool giving up
A good compressor tells you when it couldn’t reach a target and shows the smallest size it got to, rather than handing back a similar-sized file and calling it done. If it turned pages into images to hit an aggressive target, it should say so too, since the text is no longer selectable. Those two disclosures tell you whether you’ve hit the file’s limit or a tool that stopped early.
What to try
- Start with a looser target first. If a loose target (say, half your current size) is reached almost instantly, the file had room to shrink, and anything smaller will cost image quality. If even a loose target barely moves the needle, the file was already lean.
- If it’s a scan, expect some loss of sharpness. That’s the cost of a smaller scanned file, not a bug. See compressing a scanned PDF for what to expect.
- If it came from Illustrator or Photoshop, re-export without “preserve editing capabilities” (or the equivalent option) before compressing — that alone can remove a large embedded source file the compressor otherwise can’t touch.
- If several documents need to fit a portal’s per-file limit, splitting a large PDF into smaller files can help you meet that limit — but it reduces how the content is packaged, not the total amount of data, so it’s a workaround for a per-file cap, not a way to make the content smaller.
FAQ
Is there a hard limit to how much a PDF can shrink?
Yes, in the sense that every file has a floor set by what’s inside it. A tool can get you to that floor; nothing can get you meaningfully past it without removing content.
My compressed file came back the exact same size — is that a bug?
Not necessarily. If the file was already at or under your target size, this tool tells you it’s already small enough and doesn’t reprocess it. If you set a target smaller than the file already is and it still came back unchanged, that’s the file hitting its floor, not a malfunction.
Will converting to black-and-white or grayscale help?
It can, for image-heavy or scanned documents specifically: color image data takes more space than the same image in grayscale. It won’t do much for a mostly-text document. This tool doesn’t convert to grayscale, so you’d do that in your scanner settings or the program that made the file.
Should I try a different compressor if mine won’t shrink the file further?
Only if you suspect the first tool gave up early rather than reaching the file’s floor. If the first tool clearly showed a best-effort result with a measured size, a second tool working on the same file has the same content to work with and will likely land in the same place.