Compress PDF File Size — Reduce 90% Without Losing Quality (2026)

Compress PDF file size — why your PDF is so big in the first place

To compress PDF file size effectively, you first have to know what is making the file large. A PDF file is a container holding three classes of data: text + vector graphics (almost free in size), embedded raster images (typically 80–95% of the total file size), and metadata (fonts, colour profiles, attached files, form fields). When you compress PDF file size, the tool attacks class #2 — embedded images — because that is where 90% of the bytes live.

The 6 levers that compress PDF file size

Every PDF compressor — browser-based, server-based or desktop — uses some combination of these 6 techniques. Knowing them lets you predict how much your specific PDF can shrink:

Lever 1 — JPEG re-encoding (biggest single lever)

Embedded images are re-encoded at a lower JPEG quality setting. Quality 0.85 = mild, 0.65 = moderate, 0.5 = aggressive, 0.3 = extreme. Each step roughly halves the image data. Typical impact: 30–80% size reduction on image-heavy PDFs.

Lever 2 — Image DPI down-sampling

Most embedded scans are stored at 300 DPI (print-quality). Down-sampling to 150 DPI (screen-quality) halves the pixel count and roughly quarters the image data. Typical impact: 40–60% reduction on scanned PDFs.

Lever 3 — Greyscale conversion

Colour scans use 3 bytes per pixel; greyscale uses 1. Converting colour scans to greyscale instantly cuts image data by 60%. Works perfectly for ID cards, forms and signed documents that don't need colour. Typical impact: 50–60% reduction.

Lever 4 — Font subsetting

Most PDFs embed entire font files even when only 30 characters are used. Subsetting embeds only the characters that actually appear. Typical impact: 5–15% reduction on text-heavy PDFs.

Lever 5 — Stripping unused metadata

PDFs often carry hidden data: thumbnails, attached files, XFA form schema, multiple revisions, embedded colour profiles. Removing these is risk-free. Typical impact: 10–25% reduction.

Lever 6 — Object stream compression

Modern PDF spec allows multiple objects to be packed into a single compressed stream. Most modern PDFs already use this; older PDFs (pre-2010) often don't. Typical impact: 5–10% reduction on legacy PDFs.

Real benchmarks: how much does each PDF type compress?

We ran 30 PDFs through FreePDFCompress at the 1MB target. Average results by category:

How to compress PDF without losing quality (the safe approach)

If quality preservation is more important than hitting the smallest size, follow this exact sequence:

1. Set your target one step LARGER than you think you need. If the portal allows 5 MB, aim for 4 MB — gives the compressor headroom to keep image quality.
2. For text-heavy PDFs: any target works — text is vector, never loses quality. Pick the smallest cap your portal allows.
3. For mixed PDFs: use a moderate target (50–70% of original). Anything more aggressive starts visibly compressing photos.
4. For scanned PDFs: convert to greyscale FIRST, then compress. 90% smaller, zero quality loss for text-only scans.
5. Verify before uploading: open the compressed PDF and zoom into the smallest text. If readable, you are done.

The fastest way to compress PDF file size right now

If you just want to compress one PDF in 15 seconds: open FreePDFCompress, drop your file, pick the target, click compress. The tool applies all 6 levers automatically with adaptive JPEG quality — you don't need to think about which lever to use. (For agencies needing per-element control, Adobe Acrobat Pro is the desktop alternative — paid, slower, more granular.)

For specific size targets, use the dedicated tool pages: 1 MB, 2 MB, 500 KB, or 200 KB.