Compress PDF to 1MB — Free, Fast & Private
One megabyte sounds generous next to the 100 KB and 200 KB limits on some forms, and for a two-page scan it is. The trouble starts with the bundles that 1 MB boxes are usually meant for: a full set of supporting documents scanned in one go, a property or loan file, a thesis appendix, a medical history, a stack of receipts for a claim. A phone scanning app or an office copier can produce those at tens of megabytes. This page opens the MiniPx PDF compressor with its target at 1 MB so you can bring a file like that down in your browser and see exactly what size it landed at.
Check the limit on your own form first; portals and offices differ, and some list a separate maximum for each document type. Smaller targets are on the chips above if you need them.
Which 1 MB?
A megabyte can mean 1,000,000 bytes or 1,048,576 bytes (1,024 × 1,024), and forms rarely say which. The tool aims at 1,000,000. A file under that is under both, so it passes whichever one the form uses.
One thing that can look odd: the size readout in the tool divides by 1,024, while the target is counted in thousands. So a result that meets the target reads as a number below 1,000 KB — 1,000,000 bytes is about 976.6 KB on that scale (1,000,000 ÷ 1,024). Your computer's file manager may use either convention. It is the same file, counted differently, and the target was met against 1,000,000 bytes.
A larger budget also changes the advice. At 1 MB, a short scan is rarely the problem, and a size target is not always the right tool: it re-encodes every page as an image even when a gentler setting would do. If your file is a handful of pages and you only need it lighter, the No limit option with a compression level may keep more detail. Use the target when the form enforces a number.
How the tool gets a long scan under 1 MB
Each page is rendered and re-encoded as a JPEG image, and each page keeps its original page size, and the document is rebuilt in the same order. With a target set, a byte search picks the JPEG quality page by page so the pages together fit the budget; a page that cannot fit on quality alone also has its image reduced in pixel dimensions, in steps, while the page box stays exactly the same. The finished PDF is measured against 1,000,000 bytes, and the number you see is that measurement.
| Pages | Roughly per page |
|---|---|
| 4 | 250 KB |
| 20 | 50 KB |
| 40 | 25 KB |
| 100 | 10 KB |
Those shares come from dividing 1,000,000 bytes by the page count, less the few kilobytes the PDF structure itself needs. A twenty-page bundle at 1 MB gets about the same per page as a four-page file at 200 KB. A hundred-page bundle gets about what ten pages get at 100 KB.
Time, memory and one measured run
On the built site we ran a real four-page scanned PDF of 8,538,448 bytes (8.1 MB as the tool shows it) on this target. It came out at 997,648 bytes, in 7 to 9 seconds in Chromium on a laptop. That is a single measurement of a single file, and a long bundle is a different job: the byte search may encode each page several times, so time grows with the number of pages, and every page has to be held in memory while the new PDF is written. On a phone, a very long scan can run out of memory before it finishes. If that happens, split the bundle into two or three parts — many forms accept several files anyway — and compress each part. Closing other tabs before a long run also leaves the browser more room to work.
Splitting is also how a long file gets a target at all: a KB target works on PDFs of up to 40 pages. And each part gets the whole 1 MB to itself, so three parts of a sixty-page bundle get about three times the per-page budget the single file would have had. If the form takes one file only, the order of pages is kept exactly as in your original, so you can check the result against the source page by page.
If 1 MB cannot be reached, a red notice says so, shows the smallest size the PDF could be made, and warns that a form enforcing 1 MB will refuse it. You can still take that file if it is smaller than your original. If the rebuilt PDF is not smaller at all, nothing is offered and your original stays the better file.
Know what this method changes
Every page comes back as a picture of itself. Selectable text does not survive, so the result cannot be searched or copied from. For a scanned bundle that costs nothing, because it was pictures to begin with. For a long PDF exported from software, such as a report or a statement downloaded from a bank, it is the wrong route; the tool checks the first pages and warns you if it finds live text. With a size target on, colour also comes back as standard sRGB. All of it — rendering, the byte search, re-encoding and the rebuild — runs in this browser tab, and the PDF is never uploaded.