How can a PDF be squeezed to 200 KB?
To compress a PDF to a size such as 100 KB or 200 KB, add the file, pick the target, and press Compress. The page draws each page onto a canvas, saves it as a JPEG, and builds a new PDF from those pictures. It then tries different resolutions and qualities until the file is as sharp as it can be while staying under your number. The work happens in your browser and the file is not uploaded; the site does count page views.
The search is quick because it does not rebuild the whole file for each guess. It saves a sample of pages (first, middle and last) at several settings, estimates the total, picks the sharpest setting that fits, builds the real file and measures it. Since a sample is only an estimate, the real file can land a few percent over. If so, the target is tightened and the file is built again, up to three times.
Be clear about the cost. The result is a PDF made of page pictures. You cannot select or search the text any more, links stop working, and form fields are flattened. Very small targets also blur: a 20-page document squeezed to 100 KB has about 5 KB for each page.
Light re-save or page images?
The Light method opens the PDF and saves it again with its objects packed into compressed streams. Nothing about the pages changes, so text stays selectable and links keep working. It helps when a file was written wastefully, for example by an old export, and does nothing for a PDF whose size comes from photographs. Try both runs both methods, shows each size, and hands you the one that meets your target; if both do, you get the light one because it keeps the text.
| Light | Page images | |
|---|---|---|
| Text selectable | Yes | No |
| Links and forms | Kept | Lost |
| Reaches 100 KB | Only if mostly text | Usually, with blur |
| Typical saving | Small | Large, you choose |
Source: how the two methods work; the saving depends on the file.
What each target costs in sharpness
A scanned page is mostly white with some dark marks, which JPEG handles well. The table counts how many bytes each page gets at common targets. These are our arithmetic, not a measurement of your file.
| Target | 5 pages | 20 pages | 100 pages |
|---|---|---|---|
| 100 KB | 20 KB | 5 KB | 1 KB |
| 200 KB | 40 KB | 10 KB | 2 KB |
| 500 KB | 100 KB | 25 KB | 5 KB |
| 1 MB | 200 KB | 50 KB | 10 KB |
Source: calculated, ignoring about 350 bytes of PDF structure per page.
Around 20 to 50 KB a page, a typed A4 page at 100 to 150 DPI is readable. Under 10 KB a page the search drops to 72 DPI or lower and the small print goes soft. At 2 KB a page, only large headings survive, and the tool says so in the result. If the pages are the problem, remove some first with the merge and split tool.
Why a PDF is large in the first place
A PDF made from typed text is small, often tens of kilobytes a page, because the letters are stored as shapes from a font. A scan or a phone photo is a picture of the page, and a picture of an A4 page at 300 DPI is about 8.7 million pixels. Photos placed in a Word document keep their full camera size, so a ten-page file with a few photographs can pass 20 MB. That is why the light method rarely helps with scans, and why turning the pages into JPEGs at a lower resolution cuts the size so much.
Upload limits on forms
Job, scholarship and exam portals often cap a document at a size such as 100 KB, 200 KB or 500 KB. The wording and the limit change from one cycle to the next, so read the current instructions on the page you are filling in. A few rules of thumb: set the target a little under the limit, keep the page count down, and keep the original, since a page-image copy is a one-way conversion. If the form wants one image instead of a document, the PDF to JPG tool makes one picture per page, and the compress image tool brings a picture down to a KB limit. For photos and signatures sized for exam forms, see the exam photo and signature resizer.
Getting a small PDF without blur
- Start from the best source. Export from Word or a design tool instead of scanning a printout, and the text stays tiny and sharp with no compression at all.
- Scan in black and white or grayscale at 150 DPI. Colour scans cost far more bytes for a page of black text.
- Drop pages you do not need before compressing. Fewer pages means more bytes for each one.
- Go up a size if the preview is soft. 200 KB instead of 100 KB roughly doubles what each page gets.
Page images mode cannot be undone: the text is gone from the new file. Save the compressed PDF under a new name and keep the first one for anything you need to search or copy from.
Common questions
Is my PDF uploaded?
No. It is read, drawn and rebuilt inside your browser, and the result exists only in its memory until you download it. The site counts page views.
Why is the file not exactly 200 KB?
The search aims just under the target. A PDF cannot be padded to an exact size, and no form asks for one.
Why did Light barely change the size?
The size is made of pictures that were already compressed. Re-saving cannot shrink a photograph. Use Page images, or reduce the images in the source document.
Will the compressed PDF open everywhere?
Yes. It is an ordinary PDF containing one JPEG per page, which every reader and upload portal handles.
Does it work on password-protected files?
Not yet. Remove the password in the program that made the file, then add it here.
Sources
- @cantoo/pdf-lib, the open-source library that writes the PDF (MIT licence)
- PDF.js by Mozilla, which draws the pages (Apache 2.0 licence)
- MDN Web Docs: HTMLCanvasElement.toBlob()
Last reviewed November 9, 2026.