A 30 MB attachment that bounces off an email limit. A portfolio that takes forever to load. Large PDFs are a common headache — here's what actually makes them big and how to fix it.
A PDF is a container, and its size depends almost entirely on what's packed inside. The usual culprits:
Since images dominate most large PDFs, reducing their resolution or switching to a more efficient compression produces the biggest savings. A 300 DPI image resampled to 150 DPI is one-quarter the pixel count — and for on-screen viewing, 150 DPI is more than sufficient. JPEG compression at quality 75-85 is usually indistinguishable from the original for photographs.
This is lossy compression: some image detail is permanently discarded. For documents destined for print at high quality, you may want to keep higher resolution. For email, web, or on-screen review, the trade-off is almost always worth it.
Instead of embedding an entire font (every glyph in every weight), font subsetting includes only the characters the document actually uses. A font with 3,000 glyphs might only need 80 for a given document. Most modern PDF export tools do this by default, but older or misconfigured software sometimes embeds full fonts. Re-exporting the document from the original application with subsetting enabled can recover significant space.
PDF editing tools that add and remove pages often leave orphaned objects — images, fonts, or content streams that no page references any more, but that still sit in the file. A cleanup pass that removes unreferenced objects and rebuilds the cross-reference table can trim surprising amounts of dead weight without touching visible content at all.
If the PDF's text and vector drawing instructions are stored uncompressed, applying Flate compression to those streams can save 50-70% of their size. This is completely lossless — the rendered page is bit-for-bit identical. Most tools do this automatically during optimization.
Embedded thumbnails, XML metadata, document history, and attached files can be removed if you don't need them. This is usually a minor saving compared to image compression, but it helps — and for privacy-sensitive documents, stripping metadata is good practice regardless. meldpdf has a dedicated remove PDF metadata tool for this.
If you still have the original document (a Word file, InDesign layout, or presentation), re-exporting to PDF with optimized settings is often the most effective single step. You get clean font subsetting, proper image compression, and no accumulated cruft from editing. Most applications have a "reduce file size" or "optimized" export preset.
Lossless techniques — removing unused objects, compressing streams, stripping metadata, subsetting fonts — should always be applied. They reduce size without any quality trade-off.
Lossy techniques — image resampling and JPEG recompression — trade some image quality for much larger size savings. Use them when:
Avoid lossy compression when the PDF contains line art, technical drawings, or text rendered as images — these degrade more visibly than photographs.
meldpdf's compress PDF tool runs entirely in your browser. Pick your file, choose a compression level, and download the result — your document is never uploaded to a server. It handles image resampling, stream compression, and object cleanup in one step.
It depends on the method. Lossless compression (removing duplicate objects, compressing streams) preserves quality perfectly. Image resampling reduces quality slightly but often imperceptibly at reasonable settings.
It varies enormously. A scan-heavy PDF with uncompressed images might shrink 80-90%. A text-only PDF that's already optimized might barely budge. The biggest gains come from image-heavy files.
Standard compression should not. Removing unused objects or re-exporting from scratch can occasionally strip interactive elements, so check the result if your PDF has fillable forms.
Last updated: 3 September 2026.