Compression v2 gives you four quality tiers plus a target-size mode that searches settings until your KB/MB goal is met — and tells you honestly when it cannot be reached.
How it works
- Upload a PDF — you’ll see how much of it is embedded photos vs text.
- Pick a tier (Balanced is lossless), or switch to “Fit to a size target” and enter a goal.
- Compress, then compare the before/after sizes in the result panel.
Good to know
- Only embedded photos are re-encoded. Text, vectors and fonts are never touched, so text stays razor sharp at every tier.
- Strong and higher tiers also downscale photos whose effective resolution exceeds the tier’s DPI cap (200/150/110 DPI).
- PDFs made mostly of scanned text or already-compressed images may barely shrink — the composition meter warns you first.
- If a size target is impossible without destroying readability, you get the closest achievable file and a clear message instead of a fake “success”.
- Every step runs locally in your browser using Web Workers — your file is never uploaded to any server.
Frequently asked questions
Why didn’t my PDF get much smaller?
Mostly-text PDFs and PDFs whose images are already heavily compressed have little fat to trim. The Balanced tier still strips metadata and unused objects, but photo re-encoding is where big savings come from.
Can I compress without making text blurry?
Yes. Text and vector graphics are never re-encoded — only embedded photographs are. Even Extreme keeps text crisp.
What does target-size mode do?
It tries combinations of JPEG quality and image scaling, highest fidelity first, and stops at the first combination that fits your goal. If nothing fits, you get the smallest safe result and an honest note.