Compress PDF Files Online for Free

PDF Compress: Shrink a PDF by Cleaning Up Its Internal Structure

A PDF file often carries more weight than its visible content needs — duplicate objects that were never cleaned up during editing, images embedded at a higher resolution than the page displays, fonts included in full when only a handful of glyphs are actually used. This tool takes an uploaded PDF, hands it to a PDF engine called mupdf running on the server, and asks it to re-save the document with unused objects garbage-collected and its streams, images, and fonts recompressed, then returns the resulting file along with the original and new sizes so you can see exactly what changed.

Unlike several of the other PDF tools on this site, this one can't run entirely inside your browser — the recompression work genuinely happens on the server, using the same mupdf engine under the hood whether you're using this page's interface or calling the metered API directly. That's a meaningful distinction worth being upfront about, and the privacy section further down covers exactly what that means for the file you upload.

How to Compress a PDF with This Tool

1. Choose the PDF you want to shrink and start the conversion. Because this is a server-side tool, the file is uploaded as part of the request rather than processed on your device.
2. On the server, mupdf opens the document, discards objects nothing in the file actually references anymore (a "garbage collection" pass), and re-saves the remaining streams, embedded images, and fonts with recompression applied.
3. The optimized PDF comes back as a download, and the response also carries the original file size and the compressed file size, so the interface can show you a plain before-and-after comparison — for example, "2.4 MB → 1.1 MB (54% smaller)" — rather than asking you to trust that something happened.
4. If you're calling this from a script instead of the page, the metered API returns the compressed PDF as the response body with the original and compressed byte counts in response headers, so an automated pipeline can log or react to the actual size change.

There's no slider for "how aggressively to compress" here — it's a single automatic optimization pass, not a menu of quality levels. That keeps the tool simple, but it also means the result you get is determined entirely by what mupdf finds to clean up in your specific file, which is the honest core of the section below.

Why People Need to Compress a PDF

• Staying under an email attachment or upload-form size limit. Plenty of mail providers and web forms cap attachments at a few megabytes, and a report full of high-resolution screenshots or scanned pages can easily blow past that.
• Speeding up how quickly a document opens or loads. A leaner PDF opens faster in a viewer and downloads faster over a slow connection, which matters for documents shared with clients or the public.
• Reducing storage footprint across a large document archive. Compressing hundreds of stored PDFs, even by a modest percentage each, adds up meaningfully across a whole archive or document management system.
• Cleaning up a PDF that's been edited or re-saved many times. Documents that have passed through multiple rounds of editing in different tools tend to accumulate leftover objects that a garbage-collection pass can remove even without touching the visible content at all.
• Preparing a file for faster upload into another system. Whether it's a CMS, a client portal, or a government filing system, a smaller file simply moves through the pipeline faster.

What Compression Actually Does to a PDF — and What It Doesn't Promise

It's worth being direct here rather than making a vague promise: this tool performs a real, structural optimization — garbage-collecting unused objects and recompressing streams, images, and fonts — not a "magic" universal compressor that guarantees any particular percentage of reduction. How much (if any) size actually comes off depends entirely on what was in the source file to begin with. A PDF full of large, uncompressed bitmap images or one carrying bloated leftover objects from years of edits in different software will typically shrink a lot, because there's genuinely wasteful content for mupdf to clean up. A PDF that was already produced efficiently — say, one exported cleanly from modern software with already-compressed images and no leftover cruft — may come back close to the same size, or in rare cases marginally larger, simply because there was nothing wasteful left to remove.

Source PDF characteristics	Typical outcome
Large uncompressed or high-resolution embedded images	Often shrinks substantially
Many rounds of edits/re-saves in different software	Frequently has leftover objects that clean up well
Already exported cleanly from modern, efficient software	May shrink only slightly, or not at all
Mostly vector text with few or no images	Smaller potential gains — text streams are already compact

None of this makes the tool less useful — it just means the honest expectation is "this will optimize whatever inefficiency actually exists in your file," not "this guarantees your PDF gets dramatically smaller no matter what." The before-and-after size comparison shown after each run reflects exactly what happened to your specific document, not a marketing estimate.

A quick word on what "garbage collection" and "recompression" mean concretely, since both terms get used loosely elsewhere. Garbage collection here refers to walking the PDF's internal object graph and discarding anything nothing else in the file points to anymore — a common byproduct of editing a PDF repeatedly, where old versions of an image, an unused font definition, or a since-deleted annotation can linger in the file's structure even though nothing visible references them. Recompression means re-encoding the streams that remain — the data behind embedded images, fonts, and content — using tighter compression settings than whatever was applied when the file was originally produced. Both operations change only how the document is stored internally; neither touches the page content a viewer actually renders, which is why compressing a PDF here doesn't alter what it looks like when opened.

Running the tool a second time on an already-compressed output is unlikely to shrink the file any further. Once mupdf has garbage-collected the unused objects and recompressed the remaining streams, there's typically nothing left over for a repeat pass to find — a second run on the same file is more likely to come back close to identical in size than to produce another meaningful reduction, which is a useful sanity check if you're ever unsure whether a file has already been through this process.

Compression vs. Other Ways to Shrink a PDF

A few other approaches exist for reducing a PDF's size, each with different trade-offs against this tool's structural, no-quality-loss-to-text approach:

Approach	How it reduces size	Trade-off
This tool's structural optimization	Garbage-collects unused objects, recompresses streams/images/fonts	Gain depends on source; already-lean files see little change
Manually re-exporting from the original source document	Regenerating the PDF from scratch, often with better image settings	Requires having the original editable file, which isn't always available
Down-scaling embedded images before re-inserting them	Aggressively reduces image resolution/quality directly	Can visibly degrade image quality; needs image-editing steps
Splitting a large document into smaller files	Reduces the size of each individual file, not the total	Doesn't shrink the content itself, just divides it up

Where Does My PDF Go When I Compress It?

Because compression genuinely requires the mupdf engine running on a server, the PDF you upload here is transmitted to that server to be processed — this is not one of the client-side tools on this site that keeps everything on your device. Once mupdf has produced the optimized file and it's been returned to you, the server does not keep a copy: the uploaded file exists only for the duration of that one request and is discarded afterward, not retained, archived, or inspected beyond what's needed to run the optimization. If keeping a file entirely local is a hard requirement for a particular document, that's a real constraint of this tool worth planning around — but for the vast majority of PDFs people compress (reports, forms, shared documents that aren't especially sensitive), a brief, one-time transfer to perform a server-side optimization is a reasonable trade for getting the same recompression quality a desktop application would produce.

Why didn't my PDF get any smaller after compression?

This happens when the source file was already efficiently structured — no leftover unused objects, images already reasonably compressed, no bloat from repeated edits. The tool performs a real optimization pass rather than a fixed-percentage trick, so a file with nothing wasteful in it can legitimately come back the same size or, occasionally, a hair larger.

Is there a way to choose a stronger or lighter compression level?

No — this tool runs a single automatic optimization pass rather than offering adjustable compression levels. It's the same garbage-collection-plus-recompression treatment applied consistently to every file, not a slider you can push further.

Will compressing my PDF make the text blurry or lower quality?

No. The optimization targets the file's internal structure — unused objects, image and font stream compression — not the text itself, which is stored as vector drawing instructions rather than pixels and isn't degraded by this process.

Is my file uploaded to a server, or does this run in my browser?

It's uploaded to a server. Unlike several client-side tools elsewhere on this site, PDF compression requires the mupdf engine, which runs server-side; your file is processed there and discarded afterward rather than staying on your device throughout.

Can I compress a PDF that's already password-protected?

Compress this file first if it isn't locked, or unlock it before compressing — the underlying engine works with the document's structure directly, so an encrypted PDF should be decrypted with the password tool first if you need both operations done.

Does compressing a PDF change how many pages it has, or reflow the content?

No. Page count, page order, and the visual layout of every page are left exactly as they were — compression only touches the file's internal storage (unused objects, stream and image compression), never the content itself. Open the compressed file and it should look pixel-for-pixel identical to the original, just stored more efficiently on disk.

Is there a file size limit for this tool?

Yes — uploads are capped at 25MB per file, which comfortably covers the vast majority of everyday PDFs, including reports with dozens of embedded images. A document near or above that limit is exactly the kind of file most likely to benefit from compression in the first place, so it's worth trying even for a file that's close to the cap.

Related Tools

If your file needs a password removed or added rather than shrunk, the PDF password protect and unlock tool handles that separately. To combine several documents before compressing the result, start with the PDF merge tool. If a document only needs a handful of its pages kept, the PDF split tool can pull those out first, often shrinking the effective size more than compression alone. And once your file is as small as it's going to get, the PDF to JPG converter is useful if you need individual pages as standalone images instead.