Compress PDF

🔒 Runs in your browser — files are never uploaded

Shrink a PDF until it fits the size limit you are up against. Pick a target and the tool works out how much compression is needed. Nothing is uploaded.

How it works
  • Select a PDF. The tool reads it locally and shows its page count and current size.
  • Choose the maximum size you need — the presets cover the common form and email limits, or enter your own in KB.
  • Click Compress PDF. It tests how much compression the target requires, then rebuilds the file and downloads it.

How this actually works — and what it costs you

Most tools are vague about this, and it matters. Compression here works by rendering each page as an image and re-encoding it as a JPEG, then rebuilding those images into a new PDF. It is the only approach a browser can take without sending your file to a server.

The consequence is straightforward: text stops being text. A page that had selectable, searchable, copyable words becomes a picture of those words. Visually it looks the same. But you cannot search it, copy from it, or have a screen reader read it.

Whether that matters depends entirely on what happens to the file next:

  • Uploading to a form or portal? It makes no difference. A person or a scanner looks at the page either way.
  • Emailing a document someone will read? Still fine.
  • A contract, a reference document, anything you will search later? Keep the original. Compress a copy.

The tool checks your file when you select it and warns you if it contains real text, so you are never caught out by this.

What compresses well, and what doesn't

Scans compress brilliantly. A scanned document is already a photograph, so re-encoding it loses almost nothing you can see while cutting the file by 80–95%. Since scans are also what people most often need to shrink for an upload, this is the common case and it works well.

Text documents compress well but change character. A text-heavy PDF exported from Word is small in the first place because text is stored as instructions rather than pixels. Turning it into images can occasionally make it bigger at high quality settings — the tool detects this and compresses harder — but you lose the text layer.

Photo-heavy documents are the sweet spot. A PDF built from phone photos is nearly all image data and shrinks enormously with no meaningful quality loss.

Picking a realistic target

TargetRealistic for
20–50 KBA single mostly-white page. Anything longer will be visibly degraded — resize or remove pages instead.
100 KB1–3 page scans. The standard exam and government form limit. Achievable and usually still legible. More on hitting 100 KB →
200–500 KBMost upload portals. Comfortable for documents up to about ten pages with good quality retained. More on hitting 500 KB →
1–2 MBEmail attachments, visa and university applications. Roomy — longer documents keep near-original quality. More on hitting 1 MB →
The single best trick: remove pages first. File size scales almost linearly with page count. A 20-page PDF squeezed into 200 KB gives every page 10 KB, which looks terrible. If the form only needs 4 of those pages, remove the rest first and each remaining page gets 50 KB — five times the detail at the same total size.

What this tool does not do

Being direct about the limits, since plenty of sites are not. Desktop software and server-side services can do things a browser cannot: rewriting a PDF's internal object structure, subsetting embedded fonts, and re-encoding embedded images while keeping the text layer intact. That last one is what you would ideally want, and it needs to parse and rebuild the PDF's internals rather than re-render it.

So if you need a compressed PDF that stays searchable, this is not the right tool and no browser-based tool will be. What you get here instead is: no upload, no account, no watermark, no page limit, and a file that will pass the size check you are trying to pass.

Frequently asked questions

How much can a PDF be compressed?
Scanned documents commonly shrink by 80 to 95 percent with little visible change. Text documents exported from Word are already small and gain less. The tool tells you the before and after sizes so you can judge the result.
How do I compress a PDF to exactly 100 KB?
Choose the 100 KB preset, or enter 100 in the custom box. The tool tests how much compression that requires and applies it. For multi-page documents, remove any pages you don't need first — that helps far more than compressing harder.
Why can't it reach my target size?
Usually because the document has too many pages for that budget. Every page needs some minimum amount of data to stay legible. Remove unnecessary pages, or accept a slightly larger target — the tool tells you the smallest it managed.
Does compressing reduce the quality?
Yes — this is lossy compression, and that is unavoidable when making a file smaller. Gentle targets are visually indistinguishable; very small targets are visibly softer. Always open the result and check it is readable before submitting it.
Can I undo the compression?
No. The discarded detail is gone permanently. Your original file is never modified — a new file prefixed compressed- is downloaded — so keep the original.
Is my PDF uploaded to a server?
No. The whole thing runs in your browser — pages are rendered locally, re-encoded, and rebuilt into a new PDF on your device. Nothing is transmitted, which matters because the documents people compress are usually ID scans, certificates and bank statements.
Will the text still be selectable afterwards?
No. Compression works by turning each page into an image, so text becomes pixels. That is fine for submitting a form, where a human or a scanner reads the document either way, but keep your original if you need a searchable copy.
Does it work on scanned documents?
That is exactly what it is best at. A scan is already a picture, so re-encoding it loses almost nothing visible while cutting the file dramatically. Text-based PDFs compress well too, but lose their selectable text.