It's worth understanding exactly what this tool does before you rely on it: it re-saves your PDF using object streams — a standard PDF feature that groups the document's internal objects (fonts, page dictionaries, form field definitions, annotations) together and compresses that group, instead of storing each one separately. This is a real, safe optimization, and it never touches a single pixel of an embedded image or a glyph of text — but it also means it does not re-encode or downsample embedded images, which is where most of a scanned PDF's file size actually lives.
What this means for your result
- Text-based PDFs (exported from Word, Google Docs, a code editor) tend to have a lot of small internal objects — form fields, font subsets, structural metadata — so object-stream packing can give a modest, genuinely useful reduction here.
- Scanned or image-heavy PDFs won't shrink much through this tool, because the images themselves — usually the overwhelming majority of the file's bytes — are copied through untouched. If you need a scanned document to go from, say, 8MB down to 1MB, this specific optimization won't get you there.
- Nothing about image or text quality changes, in either case — since no image is ever re-encoded, there's zero risk of visible quality loss, at the cost of a smaller size reduction than an image-recompressing tool would achieve.
We go deeper into the general theory of PDF compression (including what a full image-recompressing pipeline looks like) in PDF Compression Explained — that article covers compression more broadly, not just what this specific tool implements today.
Common uses
- Shaving overhead off a PDF that already has lots of form fields, annotations, or embedded fonts
- A quick, zero-risk pass before sharing a document, when you don't need dramatic size reduction
- Reducing size on documents where you specifically cannot afford any change to image fidelity
If you need a scanned document to shrink dramatically
That requires actual image recompression, which this tool doesn't do yet. For now, the most reliable path is reducing the scan resolution/quality before it becomes a PDF in the first place (most scanning apps let you pick a lower DPI or "web quality" setting) — or converting each page to an image, running it through Image Compressor, and rebuilding the PDF with Image to PDF.
FAQ
Why did my scanned PDF barely shrink?
This tool optimizes the PDF's internal object structure — it doesn't re-encode embedded images, which is almost always where a scanned document's size actually comes from. A small reduction here is expected and normal, not a sign anything went wrong.
Will this ever make my images or text blurry?
No, and it can't — no image is re-encoded and no text is touched. Every byte you get back that isn't structural overhead is byte-for-byte identical to your source file.
What's the largest file I can compress?
100MB on the free plan, 500MB on Premium, processed entirely in your browser regardless of size.
Should I compress before or after merging multiple PDFs?
Before, if you're going to run this at all — merging several files first just means the object-stream optimization applies once to a bigger combined file instead of once per source file, with no real advantage either way given how modest the effect is.