Compress PDF to 100 KB
The tightest PDF target there is. It works for short documents — and says so clearly when it cannot.
Updated · Published by Novpix LLC
Drop a PDF here
or click to browse — target is 100 KB
Files are processed on your device and never uploaded.
Runs entirely in your browser — nothing is uploaded
How it works
- 1Drop your PDF onto the tool. A one- or two-page document is the realistic case at this target.
- 2Click Compress. The first pass renders at 150 dpi and measures the result, then predicts the resolution needed for 100 KB.
- 3Download if the target was met, or read the reported floor — the smallest size achieved and why it stopped there.
Frequently asked questions
- Can a PDF really be squeezed to 100 KB?
- It depends almost entirely on page count. Once a page becomes an image it costs a minimum number of bytes no matter how far the resolution drops, roughly 20–40 KB for anything still legible. One or two pages fit inside 100 KB comfortably; ten pages cannot, and no setting will change that. The tool tells you which case you are in rather than leaving you guessing.
- How does it decide the resolution?
- By measuring, not guessing. The first pass renders at 150 dpi and weighs the result. JPEG size tracks pixel count, and pixel count is proportional to resolution squared, so the resolution needed to reach 100 KB follows arithmetically from that single measurement. The second pass usually lands. Up to five run, and the smallest result is kept.
- What happens when it cannot reach the target?
- It reports the smallest size it managed and the resolution it used, then stops. That is deliberate: silently returning a 180 KB file to someone who asked for 100 KB wastes their submission. If the document is long, the realistic answers are removing pages or splitting it — both of which reduce size without destroying legibility.
- Will the text still be selectable afterwards?
- No. Reaching a specific size in a browser means re-rendering each page as a photograph, so the output contains pictures of text rather than text. Search, selection and copy-paste are gone. Keep the original, and if the text must remain text, remove pages instead — that shrinks the file with no loss at all.
- Why would a form demand 100 KB in the first place?
- Because attachment caps are set per system, not per document. Court e-filing portals, older government gateways and some visa systems still carry limits chosen when storage was expensive. The number is arbitrary from your side, which is exactly why a tool that targets a specific figure beats one offering vague quality presets.
About this tool
A hundred kilobytes is the hardest PDF target in common use, and the most useful thing this page can tell you is when it is achievable. The answer is governed by page count rather than by any setting: once a page has been rendered as an image it carries a floor cost of roughly 20 to 40 kilobytes while remaining legible, so a two-page form fits and a twenty-page report does not.
The tool works to the number directly. It renders the document, measures the encoded result, and uses that measurement to predict the resolution needed — because JPEG size follows pixel count, and pixel count follows resolution squared, one real measurement is worth more than a stack of guesses. Most documents converge in two or three passes rather than grinding through eight fixed presets.
When the target is out of reach it says so, reports the smallest size it achieved and the resolution that produced it, and stops. That honesty is the point of the page. A portal that wants 100 KB rejects 140 KB without explanation, so a tool that quietly hands back 140 KB has not helped you — knowing that the document has a floor, and that splitting it is the real fix, has.
The trade-off is stated up front: pages become images, so text stops being selectable and searchable. Always keep your original. When the goal is simply clearing a limit and the text matters, removing unneeded pages or splitting the document achieves it with no quality cost at all. Everything runs locally in your browser, which is also why very long documents are slow — your own hardware does the rendering, several times over.