PDF Size Estimator
Work out roughly how small your PDF will get before you spend time on it. Tell it the current size and what's inside, and you get two numbers: what our browser tool can do without touching quality, and what a full image recompression could do if you're willing to trade some. Most compression advice online quotes one number for every file, which is why it's usually wrong.
Your file
On Windows, right click the file and choose Properties. On Mac, select it and press Command plus I.
If you're not sure, open it and zoom in on some text. Blurry or speckled edges mean it's a scan.
Estimate
Will it fit?
How PDF compression works
A PDF is a container. Inside it sit streams of data: the text and its fonts, any images, and a pile of bookkeeping about page structure, embedded colour profiles, and metadata. Compression means shrinking those streams, and there are two very different ways to do it.
The first is structural optimisation, and it's lossless. The tool removes objects nothing points at any more, deduplicates resources that appear several times, drops stale metadata, and rewrites the cross reference table. Nothing visible changes, which is why it's safe to run on anything. It's also why the gains are modest: on a file that's mostly one enormous scan, there's very little bookkeeping to throw away. Our browser compressor does this kind only.
The second is image recompression, and it's lossy. The tool pulls out each image, re-encodes it at a lower quality or downsamples it to fewer pixels, and puts it back. This is where the big reductions come from on scans and photo-heavy files, and the loss is permanent, so you keep the original. It's also computationally heavy, which is why most tools doing it properly run on a server rather than in a browser tab.
The compression methods themselves are part of the format rather than a vendor feature. PDF is defined by the ISO 32000 family of standards, which specifies the stream filters a file may use, including Flate for general data and DCT for JPEG images. What isn't standardised is how aggressively any given tool applies them, which is exactly why two compressors handed the same file return different sizes. The PDF Association publishes technical documentation on how these pieces fit together.
Sources: Compression behaviour for structural optimisation reflects the documented performance of our own browser-based compressor, which reports 10 to 40 percent on text-heavy files and 5 to 20 percent on image-heavy ones. Format definitions, including stream filters and image encodings, come from the ISO 32000 family, published as ISO 32000-1:2008 and continued as ISO 32000-2 for PDF 2.0, with public technical documentation maintained by the PDF Association. Attachment limits are the published caps for each service at the time of writing and are worth rechecking, since providers change them.
Questions people ask about compressing PDFs
How much can a PDF actually be compressed?
It depends almost entirely on what is inside it. A text-heavy PDF carrying unused objects and bloated metadata can lose 10 to 40 percent through structural cleanup alone, with no visible change. A scan is already a compressed image inside a container, so cleanup barely touches it and you need to recompress the images themselves to get anywhere. Anyone quoting a single percentage without asking what is in your file is guessing.
Does compressing a PDF reduce its quality?
Not necessarily, and the distinction matters. Structural optimisation removes unused objects, duplicate resources, and stale metadata, which is lossless, so the pages look identical afterwards. Image recompression is different, since it re-encodes pictures at lower quality to save space and that loss is permanent. Our browser tool does the first kind only, which is why it is safe but modest.
Why is my PDF too large for email?
Gmail caps attachments at 25MB, Outlook at 20MB for many accounts, and a surprising number of government and university portals cap at 5MB. Scanned documents blow past these fastest, because each page is a full photograph. A ten page colour scan at 600 DPI can easily exceed 40MB while the same document produced as text would be under 1MB.
What makes a scanned PDF so large?
Every page is an image rather than text. Scanning resolution drives the size roughly with the square, so doubling from 300 to 600 DPI produces about four times the data for the same page. Colour adds more again over greyscale. If you control the scanner, dropping to 200 or 300 DPI and greyscale for text documents cuts more than any compressor will.
Is PDF compression standardised?
The compression methods are. PDF is defined by ISO 32000, and the format specifies the stream filters a file may use, including Flate for general data and DCT for JPEG images. What is not standardised is how aggressively any given tool applies them, which is why two compressors given the same file can return outputs of very different sizes. The PDF Association publishes documentation on the format.