Compress PDF to Target Size

Name the size the file has to fit under. It works down to that figure by itself, and tells you plainly if the figure cannot be reached.

Hybrid processingPDFWorkingNo account needed

What this tool does

Compress PDF to Target Size takes one number and a unit. It starts with the same lossless rebuild the plain compressor uses and checks the result against your figure; if the file already fits, that is what comes back. The unit is read literally: a kilobyte is 1024 bytes here, so a 500 KB target means 512,000 bytes and a 2 MB target means 2,097,152. A target below 1 KB is raised to 1 KB, since no PDF is smaller than that.

If it does not fit yet, the job recompresses the images inside the PDF, trying the gentlest setting first and only going further while the file is still over. It steps down through six settings, from about 200 dpi at high JPEG quality to about 60 dpi at low quality, jumping ahead when the file is far from the target. Once it has found a setting that fits, it narrows in between that one and the one above it, so the result lands close to the size you asked for instead of far below it, and is the sharpest version that fits. Resolution is judged against the size each image is drawn at on the page, and text, fonts and vector graphics are never touched, so the text stays selectable the whole way.

Some targets cannot be reached that way, and the tool is straightforward about it. A document that is mostly text, with a few small pictures, has no resolution to give up, and a long photographic document can stay above a very small figure even at the lowest setting. Only when the image settings run out does it try its last resort, converting the pages to pictures, and it keeps that result only if it actually gets under your figure. For a page that is mostly text it uses black-and-white pictures in the fax encoding, which stay sharp at a fraction of the size of a colour photo of the same page; for pages with photographs it uses JPEG. It picks the highest resolution that fits and will not go below a level where text stops being readable. That last resort is never used on a document with cropped pages or a custom page unit, since it would change what each page shows. If it does fit that way, the result says the text is no longer selectable and that links, bookmarks and form fields were dropped, and, for black-and-white pages, that colours and shades of grey are gone. If nothing fits, you get the smallest file that keeps its text, with the size it reached and how close the picture version got.

The whole job runs on the processing service rather than in the tab, because the tools involved are native programs with no browser equivalent. Every run uploads, whatever the size. The upload travels over an encrypted connection, and both it and the returned file are deleted when the job expires, six hours by default. The ceilings are a 200 MB upload, 300 pages and three minutes of processing. A document that needs a password to open is refused before any of that, because its page count cannot be read without one.

How to use it

  1. Add the PDF and name the figure. One document or several, each sent as its own job. Type the number and pick KB or MB; the default is 500 KB, which is what most upload forms are asking for. That is the only choice there is.
  2. Let it work down to it. There is no resolution to pick and no box to tick. The job tries the gentlest image settings first and only goes further while the file is still too big, so you get the sharpest version that fits.
  3. Read the result, then download. The result shows the size before and after and the percentage saved, with a line saying how it got there, or, if it fell short, a plain explanation of the smallest size it could reach and what would help.

Processing & privacy

Hybrid processing. Smaller jobs can run locally; heavier jobs may use secure server processing.

Common use cases

Meeting a portal's stated cap

Visa applications, exam boards and tax filings name a number outright: the scan must be under 200 KB, or a PDF under 2 MB. Type that figure instead of guessing at a quality slider and re-uploading.

Squeezing past an email attachment limit

A scanned contract that the mail server bounces will usually clear once its page images are resampled, and it stays searchable if it had text. If a very small limit needs more than that, the result says so and offers the smallest file it reached.

Shrinking a scan that refused to compress

A scan is a stack of photographs, so a structural rebuild reclaims almost nothing from it. Resolution is the only thing that moves the number, and this is the tool that adjusts it for you.

Building a reference archive

Where a copy only has to be legible on screen, pages resampled towards the lower settings are a small fraction of the original size and still read well.

Questions

Is my PDF uploaded?
Yes, on every run and at every file size. The work is done by native programs on the processing service, so the document is sent there over an encrypted connection, compressed, and deleted along with its output when the job expires, six hours by default.
Will the text still be selectable afterwards?
Almost always, yes. Images are what get resampled, not text. The exception is the last resort: if even the lowest image setting cannot reach your figure and converting the pages to pictures can, that is what you get, and the result says the text is no longer selectable, that links, bookmarks and form fields were dropped, and, for black-and-white pages, that colours and greys are gone. If that is not acceptable, raise the target or keep your original.
Why did it stop above the size I asked for?
Because the figure cannot be reached without ruining the file. The result gives the smallest size it could get while keeping the text selectable. A document with few images has little resolution to give up, and a long photographic one can stay above a small figure even at the lowest setting. Try a larger target, or split the file with Split PDF and send fewer pages per part.
How sharp will the result be?
As sharp as your target allows. A generous target keeps the images near 200 dpi, and tighter ones step down to about 60 dpi, which is soft but readable for typed pages. A scanned page squeezed to a few tens of kilobytes will look noticeably rougher than the original, and that is the trade a tight limit forces.
How are KB and MB counted?
In binary units: 1 KB is 1024 bytes and 1 MB is 1024 KB. If a form quotes its limit in decimal megabytes, aim slightly under to be safe. Targets below 1 KB are raised to 1 KB, since no PDF is smaller than that.
Can I run it on a protected PDF?
Not if it asks for a password before it will open. Reading the page count is the first step and it fails without that password, so the job stops there. Strip it with Unlock PDF and compress the result. A file that opens freely but forbids printing or copying keeps those restrictions and is tidied losslessly, but its images are not recompressed.
What if my PDF is already under the target?
It comes back as it was, or with a small lossless tidy if that makes it smaller, and the result says the file was already under your figure. The tool never makes a file larger, and it never compresses harder than it has to.
Why did my text document turn into black-and-white pictures?
Because its size is mostly the text itself, and embedded fonts in particular, not pictures. A page of text saved with a large font, such as one for a script with thousands of characters, can weigh tens of kilobytes even with no images at all, and nothing about the images can shrink that. When your target is below that, a black-and-white picture of the page at 150 dpi is often smaller than the fonts it replaces. The result says so plainly, so you can choose to raise the target and keep the selectable text instead.

Related tools

  • Merge PDF — Combine several PDFs into one document in the order you choose.
  • Split PDF — Split one PDF into separate documents by page range.
  • Compress PDF — Reduce PDF file size while preserving the quality that matters.
  • PDF to JPG — Export every page of a PDF as a JPG image.
  • JPG to PDF — Turn one or many JPG photos into a single PDF.
  • Protect PDF — Add a password and set permissions on a PDF.