Compress PDF
Shrink PDF file size in your browser. Image-heavy PDFs usually drop 30–70%. Nothing uploaded.
Select a PDF
or drop one PDF here
Compresses image-heavy PDFs by 30–70%. Text-only PDFs are usually already optimal.
Your file is ready
Processed entirely in your browser — the file never left your device.
How to compress a PDF
Three steps. Runs in your browser.
Pick a PDF
Drop or select one PDF. It is read locally — no upload.
Pick a quality
Low / Medium / High. Medium is the right default for most files.
Compress & download
Images get resampled in-browser, then the smaller PDF downloads.
What is "compress PDF"?
Compressing a PDF means reducing the byte size of the file without changing what it contains. For most documents that gets done by two things: resampling images down to the resolution you actually need and re-saving the file with object-stream compression (a feature of the PDF spec that many generators leave off). The combination keeps the look of the document while trimming the bloat email gateways and cloud-storage quotas care about.
The reason it matters is practical. A 30 MB scanned contract will bounce off most corporate email filters; the same contract at 4 MB sails through. A 600-page report you can read for a flight stays comfortably under a typical cloud-storage tier; the uncompressed original might not. Compress PDF is the quietly useful step that makes everything else (sending, sharing, archiving) easier.
How compress PDF works in your browser
When you drop a PDF on this page, your browser reads it into an ArrayBuffer locally — no upload. The tool walks the document objects with pdf-lib, finds embedded image XObjects, decodes the ones above your target resolution with the canvas API, resamples them down, and re-encodes as JPEG at a quality level that matches your preset. The new image streams replace the originals while every page reference inside the PDF keeps working.
Whether or not there are images, the final save uses pdf-lib's
useObjectStreams: true mode plus XMP metadata stripping.
That alone usually trims 5 to 15 percent from text-heavy PDFs. The
whole pipeline runs inside your tab. Nothing is uploaded, no account
is required, and when you close the tab everything is gone from
memory.
Common use cases
- Emailing scanned contracts. Bring a 30 MB scan down to a 3–5 MB attachment that gets past corporate mail filters.
- Sharing photo-heavy decks. Marketing decks with full-page photos shrink dramatically without visible loss on screen.
- Uploading to government portals. Many tax / immigration / school portals cap uploads at 5–10 MB per file. Compress hits the limit cleanly.
- Storing archives. Quarterly archives of statements and invoices shrink by half or more, freeing cloud storage.
- Preparing files for slow connections. If recipients open files on mobile data, smaller PDFs are noticeably faster to load.
Privacy & security
Most online "compress PDF" sites upload your file to their servers. That model works, but it means an extra third party touches every contract, scan, and receipt you compress. imisspdf runs the entire pipeline in your browser using pdf-lib. There is no account, no retention window, and no upload step that could leak. For sensitive documents — contracts, medical records, tax returns — the in-browser model removes a class of risk you cannot audit on someone else's server. See our iLovePDF privacy review for the standard upload model.
Frequently asked questions
For image-heavy PDFs — scans, photo decks, brochures — 30 to 70 percent is typical at the medium setting. Text-only PDFs are usually already optimised by the program that produced them; expect 5 to 15 percent from metadata removal and object-stream re-encoding. If the file does not shrink, that means it was already efficient and re-compressing would only hurt quality.
Object-stream re-saving and metadata stripping are lossless. Image downsampling is lossy by definition — the chosen quality preset controls how aggressive it is. High keeps roughly print resolution (300 DPI). Medium keeps screen-quality detail (150 DPI). Low is meant for files you only need to read on a phone (72 DPI).
Yes. The whole compression runs inside your browser with pdf-lib and the canvas API. The PDF is loaded into your tab's memory, images are re-encoded locally, and the smaller PDF is handed back as a download. Nothing is uploaded, no account is required, no daily limit applies.
If the PDF only contains text (most exported Word / Google Docs / LaTeX files), there are no images to downsample and the original was probably already saved with object streams. In that case Compress PDF can only strip metadata, which often does nothing. You are not doing anything wrong — the file is just already optimal.
Yes — and that is where you usually see the biggest wins. A typical 300-DPI scan compressed at the medium preset (150 DPI) drops to roughly a third of the original size with no visible loss when viewed on screen. For archival quality, use high; for email-only delivery, use low.
Tips for best results
- Pick Medium first. It is the right default for almost every file. Only switch to Low if you need extreme size, or High if the document will be printed.
- Compress after merging. Running compress on the merged file rather than each part gives better metadata savings.
- Compress before signing. Compressing after applying a digital signature will invalidate the signature; do it the other way around.
- For scans, use Medium. 150 DPI is enough for legibility on screen and prints fine. 300 DPI scans almost always have extra resolution that nobody will see.
- If the file does not shrink, that is fine. Some PDFs are already optimal. The tool will not silently re-encode if there is nothing to gain.
Related PDF tools
- Merge PDF — combine PDFs first, then compress the result.
- Split PDF — split a huge file into smaller compressed parts.
- PDF to JPG — extract images out of a PDF when you only need the pictures.
- Flatten PDF — bake form fields and annotations into static content before compressing.