M

Free PDF Compress — Reduce PDF File Size

Reduce PDF file size by stripping metadata, unused objects, and tightening object streams. 100% free and runs in your browser — your file never leaves your device.

Your file stays private. All processing happens in your browser. The file is never uploaded to any server.
What this does: strips embedded metadata, removes unused objects, and tightens object streams. Typical savings are 5-30% on office/business PDFs. PDFs that are already optimized — or that are mostly scanned images — may not shrink much. For deep image recompression of scan-heavy PDFs, a desktop tool like Adobe Acrobat usually does better.

What is this calculator for?

You exported a slide deck to PDF and it came out 24 MB — too big to attach to the client's email portal, which caps uploads at 10 MB. Or your accounting software produces statements that bloat to 8 MB per page because every page embeds the same logo, the same fonts, and a bunch of metadata. Or you scanned a packet and the resulting PDF is large enough that your CRM rejects it. Compressing a PDF is the third-most-common PDF chore after merging and splitting, and most people reach for whichever online tool comes up first on Google.

This tool runs entirely in your browser using pdf-lib. It does three things to shrink the file: strip embedded metadata (author, producer, creation timestamps, title fields, custom XMP data), remove unused PDF objects that editing software often leaves behind, and re-serialize the file using object streams — a more compact storage format introduced in PDF 1.5. For office-generated PDFs (Word exports, Google Docs PDFs, Excel reports, slide decks), typical savings land in the 5-30% range. For scan-heavy PDFs where most of the file size is embedded image data, savings are usually smaller — and the tool tells you so honestly rather than faking a compression number.

Important caveat that most marketing copy hides: this tool cannot recompress embedded images. The reason is technical, not a defect — JPEG and JPEG2000 image decoding/re-encoding inside a browser is slow and CPU-heavy, and pdf-lib doesn't ship that capability. Desktop tools like Adobe Acrobat and server-side tools like Ghostscript can downsample images from 300 DPI to 150 DPI or 72 DPI, which is the single biggest lever for shrinking scan-heavy PDFs. If your PDF is mostly scans or photos and you need aggressive compression, a desktop tool will outperform any browser-side approach.

How to use this calculator

Step 1: upload your PDF. Click the upload area and pick a file. The tool accepts any PDF up to 100 MB and reports the original size after upload. You'll see the same upload box go from "Click to choose a PDF file" to the filename and size of what you picked.

Step 2: read the disclosure. The orange info panel explains exactly what the tool does — strip metadata, remove unused objects, re-serialize with object streams — and what it doesn't do (recompress images or downsample). This sets expectations honestly. If the PDF is mostly scans, you'll see small savings; that's the truth, not a bug.

Step 3: click compress. The operation is usually fast — a few hundred milliseconds for office PDFs, a few seconds for very large or complex files. pdf-lib loads the source, applies the cleanup steps, and produces a re-saved version.

Step 4: review the results. Three numbers appear: original size, compressed size, percent saved. If the savings are under 5%, the tool surfaces a separate warning explaining that the source PDF is already well-optimized and recommending desktop tools for further reduction. There's no false claim of huge savings on already-optimized files; the goal is to be useful when it can be and honest when it can't.

Step 5: download the result. The green download button gives you the compressed file with -compressed appended to the original filename. Open it once to confirm it looks right (the visual content should be unchanged), then use the smaller version wherever you need it.

Understanding your results

The compression effort produces three categories of outcomes depending on what's in your source PDF.

Strong wins (15-40%): office PDFs created by tools that pack in heavy metadata or leave behind orphaned objects — particularly older versions of Word, OpenOffice, certain corporate template generators, and PDF editors that don't garbage-collect on save. These files have lots of "low-hanging fruit" — strippable XMP metadata, unused fonts, duplicate object references — and the re-save with object streams meaningfully tightens the encoding.

Modest wins (5-15%): well-formed PDFs from modern Office, Google Docs, LaTeX, or browser print-to-PDF. These files were already reasonably compact when they came out of the source software. Stripping metadata and re-serializing recovers a few percent, which can matter at the email-size threshold (going from 11 MB to 9.5 MB matters; going from 9 MB to 8 MB doesn't).

Minimal wins (under 5%): scan-heavy PDFs where most of the file size is JPEG image data, image-PDFs from camera apps, PDFs that have already been compressed by a previous tool, and PDFs from professional production workflows that already use object streams. For these files, the tool surfaces a transparent message: it's already optimized, browser-side cleanup can't shrink it further, use a desktop tool with image downsampling if you need aggressive compression. This honest reporting is the main editorial difference between this tool and most "miracle" online compressors that always claim to save 50%+ — the math doesn't actually support that on every file.

The compressed PDF behaves identically to the original for reading. Text is still searchable and selectable. Images render at the same resolution. Forms still display correctly (interactive behavior is preserved). Fonts render with the same hinting. The visible content is unchanged — only the invisible plumbing has been tidied. If you want to verify byte-for-byte that the visible content matches, open both files side-by-side and zoom in; you should not be able to tell them apart.

A worked example

Aisha works at a small architecture firm. The firm uses an older drafting-export workflow that produces PDFs with a lot of embedded metadata — every drawing has the project name, lead architect's email, drafting software version, internal review timestamps, and a custom XMP block for the firm's document management system. A typical floor plan PDF clocks in around 18 MB.

The city's permit portal caps uploads at 10 MB per file. Submitting any drawing requires shrinking it first. Aisha used to rely on Smallpdf, but the firm's IT policy now blocks third-party file uploads — design documents and project metadata can't leave the firm's network even briefly.

She uploads a floor plan to the browser compress tool. Original: 18.3 MB. Compressed: 12.7 MB — about 30% saved. Still over the 10 MB cap, so she opens the original file in the drafting software, lowers the image-export DPI from 300 to 200 (a one-time setting change), re-exports, then runs that 11 MB output through the compress tool again. Final size: 8.4 MB. Submitted to the permit portal in under five minutes total, no files leaving the firm's network.

Variation: Marcus is a freelance copywriter sending a 28-page proposal to a prospective client. The PDF was exported from Google Docs and weighs in at 4.2 MB — the client's spam filter quietly drops attachments over 4 MB. Marcus runs the file through compress; the metadata cleanup and stream re-encode bring it down to 3.4 MB. He attaches the compressed version to the email and it lands in the client's inbox the first try. Total time: 15 seconds.

Variation that doesn't work: Helen has a 60-page scanned contract PDF that weighs 95 MB. She runs it through compress hoping to email it. The tool processes the file in about 8 seconds and reports: original 95 MB, compressed 94 MB, saved 1.1%. The honest-results message appears explaining that the bulk of the file size is embedded scanned-image data and that browser-side cleanup can't recompress those images. Helen opens the original PDF in Adobe Acrobat, uses File → Save As Other → Reduced Size PDF, picks the "Acrobat 10 and later" compatibility setting, and the same file comes out at 18 MB after Acrobat downsamples the embedded images. That's the right tool for that specific job — and the compress tool was right not to pretend otherwise.

Related resources

To combine PDFs after compressing them individually, see PDF Merge. To extract specific pages from a compressed PDF (the savings on a smaller subset are often much bigger), the PDF Split tool handles range and per-page modes. To build a PDF from photos (which then benefits from compress), the Image to PDF tool accepts JPG and PNG. For aggressive image-aware compression, Ghostscript (open-source, command-line) and Adobe Acrobat Pro's Reduce File Size feature are the standard tools.

Frequently Asked Questions

How does this compression actually work?

Three steps: (1) strip embedded metadata — author, producer software, creation timestamps, title fields. (2) Remove unused PDF objects left behind by the editor that made the file. (3) Re-serialize using PDF object streams, a more compact storage format introduced in PDF 1.5. Typical savings for office-generated PDFs (Word, Google Docs, exported reports) are 5-30%.

Why didn't my PDF shrink much?

Most of the bytes in a typical PDF are embedded images, embedded fonts, or pre-encoded text streams — and pdf-lib can't recompress those in a browser. If your PDF is mainly scanned pages or photos, you've already got near-optimal compression. For deeper compression of scan-heavy PDFs, use Adobe Acrobat's 'Reduce File Size' feature or a server-side tool like Ghostscript — both can re-encode images at a lower DPI.

Is the output PDF still searchable and editable?

Yes. The text remains text (still searchable, still selectable, still copy-pasteable). Fonts are preserved. Forms and annotations are preserved. Only the metadata fields and unused objects are removed — none of that affects readability or functionality.

Is my PDF uploaded for compression?

No. Everything happens in the browser. Most online PDF compressors (Smallpdf, iLovePDF, Adobe's online tools) upload your file to their servers, process it there, then send you back a download. Mubboo doesn't — there's no server-side step, which means sensitive documents (financial statements, contracts, medical records) stay on your machine.

What's the size limit?

100 MB per file. Above that, browsers tend to run out of memory while loading the full PDF. For very large PDFs, split them first with the PDF Split tool, compress the parts, and merge back together with PDF Merge.

When should I use a desktop tool instead?

When your PDF is image-heavy (scans, photo books, portfolio documents) and you need aggressive compression — say 50%+ reduction. Desktop tools like Adobe Acrobat Pro can downsample embedded images from 300 DPI to 150 DPI or 72 DPI, which is the lever this browser tool can't pull. For ordinary text-and-chart PDFs, the browser-side approach usually does the job.

Will this break my PDF?

Extremely unlikely. pdf-lib produces standards-compliant PDF 1.7 output that opens correctly in every major reader (Adobe Acrobat, Preview on macOS, Foxit, Chrome's built-in viewer, Edge's viewer). The cleanup steps remove only invisible elements — metadata fields, unused objects, structural overhead — and leave all visible content alone. The one edge case to know about: some specialized PDF/A archival features rely on extensive metadata; if you need a PDF/A-compliant file (e.g., for long-term archival in regulated industries), don't strip metadata. For ordinary office use, stripping is fine and often desirable for privacy reasons.

Does this strip my author name and personal info?

Yes — that's a feature, not a bug. The metadata fields removed include author, producer (the software that made the PDF), creator, subject, title, keywords, and any custom XMP data. This is genuinely useful for privacy: a PDF created on your laptop will often include your operating system username, your Office license name, and the file path where it was saved. Stripping that information before sharing is good hygiene, especially for any document going to people outside your organization. If you specifically need to keep your name on the document, add it back in Adobe Acrobat (File → Properties → Description) after compressing.

Can I batch-compress many PDFs at once?

Not in this tool — it processes one file at a time by design. For batch jobs, the practical workaround is to run the source files through PDF Merge first (combine them into one large PDF), compress that combined output, then split the compressed result back into individual files using PDF Split. The result is roughly equivalent to compressing each file individually, with fewer clicks. For genuinely large batches (dozens of files), desktop tools or command-line utilities (Ghostscript, qpdf) are faster — they can run unattended over a folder.

Why does my PDF still look the same?

Because the visible content is unchanged — that's intentional. Compression in this tool means "smaller file with the same readable content," not "lower-quality version." Text still renders at the same crispness, images still display at the same resolution, fonts still hint correctly. If you wanted to also reduce image quality (true lossy compression), you'd need a tool that can re-encode the embedded images at lower DPI, which is the territory of Adobe Acrobat Pro and similar desktop applications. This tool deliberately stays in the lossless cleanup-only space, where you trade nothing visual for the byte savings.

Is there a downside to running it on every PDF I share?

Two minor ones. First, the metadata stripping removes any internal tracking fields your organization might rely on (some document management systems read XMP custom properties to route files automatically). If you depend on those fields, don't compress. Second, very small PDFs (under 100 KB) sometimes get slightly larger after compression because the per-file object-stream overhead exceeds the savings on a near-empty document. The tool will report this honestly — "saved 0%" or even a tiny negative — and you can simply use the original. For PDFs over about 200 KB, the savings are almost always positive.

Sources