The email that finally made me build a compressor said, in full: "Attachment too large — message not delivered." I'd spent an evening assembling a 38-page proposal with embedded product photos, exported it at 51 MB, and tried to send it to a client whose mail server capped attachments at 20 MB. It bounced twice before I even noticed.
I did what everyone does — googled "compress PDF free," picked a result, and uploaded a document full of unreleased pricing to a server in a country I couldn't name. It came back at 6 MB and looked fine. And I sat there realizing I'd just handed a confidential proposal to a stranger's temp folder to save four minutes.
That's the itch behind the Compress PDF tool on toolz.dev, and behind this guide. Compression isn't magic and it isn't lossless-by-default — but if you understand what's actually making your file big, you can cut it by half or more without anyone being able to tell. Here's how it works, which setting to reach for, and the specific mistakes that turn a clean document into a blurry one.
TL;DR: Open toolz.dev/tools/compress-pdf, drop your PDF in, pick a compression level, and download. It runs 100% in your browser — the file never uploads. For everyday sharing, medium compression cuts 40–60% with no visible loss because text compresses losslessly; only images take the hit. Before compressing, trim dead weight with Delete PDF Pages; if you're combining files, merge first and compress the result once.
Why Do PDF Files Get So Large in the First Place?
You can't compress intelligently until you know what's taking up the space. In my experience, four things account for nearly every oversized PDF, and they respond to compression very differently.
Images are the number-one offender, by a wide margin. A single high-resolution photo can add 5–15 MB, and a document built from many of them — a catalog, a portfolio, a real-estate listing — can balloon into the hundreds of megabytes. The two dials that matter here are resolution and format. A 600 DPI scan holds four times the pixel data of a 300 DPI scan of the same page, and most of those pixels are invisible on a screen. An image stored uncompressed inside the PDF is dramatically larger than the same image stored as JPEG.
Embedded fonts are the sneaky one. A PDF can carry entire font families so it renders identically on any device, and each face can run from 100 KB to several megabytes. A document that embeds eight fonts in full can carry 10–20 MB of nothing but glyph data — before a single word of actual content.
Redundant objects pile up in files that have been merged or edited many times: duplicate copies of the same logo, orphaned data structures, unused fonts, metadata bloat. None of it shows on the page; all of it counts toward the size.
And layers and annotations — comments, highlights, form fields, design layers — add data that most recipients never need.
Does Compressing a PDF Actually Ruin the Quality?
This is the fear that keeps people uploading to sketchy sites at maximum quality "just to be safe," so it's worth answering plainly: it depends entirely on what gets compressed, and the two halves of your document behave in opposite ways.
Text and vector graphics compress losslessly. Text in a PDF is stored as font glyphs and character positions — instructions, not pixels — and the algorithms that shrink them (Flate/Deflate, the same math inside a ZIP file; font subsetting; object deduplication) throw away nothing. Your paragraphs stay razor-sharp at any zoom, at any compression level. This is the part people don't believe until they see it.
Images are the part that can degrade, because that's where lossy compression earns its file-size savings. JPEG recompression discards high-frequency color detail the eye barely registers; downsampling drops a 600 DPI image to 150 DPI when it's only ever viewed on a screen that can't show the difference. At sensible settings, none of this is visible. Push it too far and you get blockiness in smooth areas and mush in fine textures.
So the honest summary: your text is never at risk, and your images are only at risk if you over-compress them. A good compressor does several things at once — recompresses images, downsamples the ones that exceed the resolution you actually need, subsets fonts to only the characters used, strips duplicate objects, and applies lossless compression to everything structural.
How Do You Compress a PDF Online, Step by Step?
The mechanics are the easy part. Here's the flow on the Compress PDF tool, with the checks that matter.
Step 1 — Open the tool. Go to toolz.dev/tools/compress-pdf. No signup, no install, works in any modern browser.
Step 2 — Add your PDF. Drag it onto the drop zone or click to browse. The tool reads the original size so you have a baseline to compare against.
Step 3 — Pick a compression level. This is the only real decision. The three presets map cleanly onto intent:
| Level | Typical reduction | Looks best for | Watch out for |
|---|---|---|---|
| Low (high quality) | 20–40% | Documents headed for professional print | Barely shrinks image-light files |
| Medium (balanced) | 40–60% | Email, web viewing, everyday sharing | The right default for almost everyone |
| High (maximum) | 60–80%+ | Archiving, strict upload caps | Visible artifacts if you zoom hard |
Step 4 — Compress. Click the button. Processing happens locally in the tab — no upload, no queue, no waiting on a network transfer. Larger files take a few seconds because your own machine is doing the work.
Step 5 — Check the result, honestly. The tool shows original versus compressed size and the reduction percentage. Open the output and look at it at the size you'll actually use it — 100% on screen, or printed at real size. Confirm the text is crisp (it will be), the images are acceptable, and charts and colors survived.
Step 6 — Download, or step back down. Happy? Download. Too soft? Re-run at a lower level. Because nothing uploaded, iterating costs you seconds, not another privacy gamble.
How Much Can You Actually Compress a PDF?
"How small will it get" has no single answer, because it depends on what's inside. Here's what I've come to expect by document type, which helps you predict whether compression will even solve your problem.
Text-heavy documents — reports, papers, manuals — compress the least, usually 10–30%. There's little to squeeze because text is already compact; the savings come from font subsetting and structural cleanup. If a text-only document is stubbornly large, the weight is almost certainly embedded fonts, not content, and subsetting is your lever.
Image-heavy documents — catalogs, portfolios, brochures — compress the most, often 50–80%, because images dominate the size and respond dramatically to recompression and downsampling. If the document is only ever viewed on screen, dropping images to 150 DPI can halve the file with no perceptible difference.
Scanned documents — contracts, forms, signed letters — land in the middle at 30–60%, and this is the category where I've been burned. A scanned page is essentially one big image, so its size hinges on scan resolution. Which brings me to the mistake worth flagging loudly.
The scan mistake I made: I once took a 600 DPI signed contract that was too big for a portal, ran it through aggressive high compression, and shipped it. The signature came out with visible blocky artifacts — legible, but obviously degraded, on a legal document. The better move, if you still have the paper, is to rescan at 300 DPI rather than over-compressing a 600 DPI scan. A lower-resolution scan produces a smaller and cleaner file than crushing a high-resolution one, because you're not stacking JPEG artifacts on top of scan noise.
What's the Smart Order When You're Doing More Than Just Compressing?
Compression rarely happens in isolation — it's usually the last step in getting a document ready to send. A few sequencing rules save real rework.
Compress after merging, not before. If you're combining several PDFs, merge them first and compress the single result. One pass across the whole document is more efficient than compressing each source file and then stitching already-compressed pieces together — and it avoids double-compressing anything.
Trim before you compress. Blank scanner pages, internal-only cover sheets, and appendices nobody asked for are pure weight. Pull them with Delete PDF Pages first so you're not spending compression effort on pages that shouldn't be there at all.
Know your target number before you start. Check the actual limit you need to hit — a 20 MB mail cap, a 10 MB portal, a 5 MB form field — and pick the level that clears it on the first try instead of guessing and re-running.
And for the awkward case where different sections need different quality — say, a text-heavy body plus a photo-heavy appendix — you can split the file, compress each part at its own appropriate level, and merge them back. It's more effort, so I only reach for it when a single global level genuinely can't satisfy both halves.
Frequently Asked Questions
How do I compress a PDF without losing quality?
Use the Compress PDF tool at the low or medium level. Both apply lossless compression to text and document structure and use conservative image handling that stays invisible at normal viewing size. Your text is never at risk regardless of the setting — only embedded images are touched by lossy compression, and at medium level the difference is imperceptible on screen.
How much smaller can I make a PDF?
It depends on what's inside. Text-heavy documents typically shrink only 10–30% because text is already compact. Image-heavy documents can drop 50–80%. Scanned documents usually land at 30–60% depending on the scan resolution. The tool shows you the exact reduction after processing, so you never have to guess.
Is it safe to compress confidential PDFs online?
With most online compressors, no — your file uploads to a remote server you can't audit, which is a genuine problem for contracts, financials, or anything under NDA. The Compress PDF tool on toolz.dev processes everything in your browser. The document never leaves your machine, which is exactly why I built it that way after uploading a confidential proposal to a random site once.
Does compressing a PDF make the text blurry?
No. Text is stored as font glyphs and character positions and is compressed losslessly, so it stays sharp at any zoom level no matter which compression setting you choose. Only embedded images are affected by lossy compression. If a compressed PDF looks soft, it's the images that degraded, never the text.
What compression level is best for email?
Medium is right for almost every email. It typically cuts 40–60% while keeping quality that looks perfect on screen, which is where email attachments get viewed. If the file is still over the cap after medium, try high — the extra artifacts are usually invisible at normal viewing size and acceptable for on-screen reading.
Can I compress a password-protected PDF?
Not directly — the file has to be unlocked first. If you know the password, open the PDF, then use your system's "Print to PDF" option to produce an unlocked copy, and compress that copy with the Compress PDF tool. Without the password, there's nothing legitimate any compressor can do.
Can I undo compression if I went too far?
No. Lossy compression permanently discards image data, and you can't recover the original quality from the compressed file. This is the one reason to always keep your original, uncompressed PDF as a backup — if a compression pass comes out too aggressive, you re-run from the original, not from the degraded output.
Why is my mostly-text PDF still huge?
If a text document is large despite having few images, the weight is almost always embedded fonts. A PDF can carry entire font families in full, adding many megabytes before any content. Font subsetting — including only the characters actually used — is the fix, and a good compressor applies it automatically, which is why even text-heavy files usually shed at least 10–30%.
How do I compress a PDF to under 1 MB?
Start at medium and check the resulting size, then step up to high only if you're still over. If a 1 MB ceiling is still out of reach, the file is image-bound: split it with Split PDF and send the sections separately, or downscale the source images before exporting the PDF. Chasing 1 MB with maximum compression on a photo-heavy document produces visible artifacts, and at that point a link beats an attachment.
Does compressing a PDF break its searchable text, links, or bookmarks?
No. Compression works on the stored data — fonts, image streams, redundant objects — not on the document's logical structure. Selectable text stays selectable and searchable, hyperlinks keep working, and bookmarks survive. If text in your PDF wasn't searchable before compression, it was a scanned image to begin with, and compression neither creates nor destroys that.
