How to Compress a PDF Without Uploading It

Shrinking a PDF is usually sold as a one-click job. The questions that actually matter are simpler than the marketing: what does compression remove from your file, how much smaller can your document realistically get, and is compressing it a good idea at all? This guide answers all three, and shows how to do the whole job inside your browser with the file never leaving your device.

Why PDFs get large in the first place

A PDF is a container, not a picture and not a text file. One document can hold fonts, vector line art, embedded photographs, page descriptions, form fields, bookmarks, attachments and metadata — plus whatever objects earlier software left behind when the file was saved, exported or merged. When a PDF is unexpectedly huge, the cause is almost always one of a small number of things:

Images. This is the cause in the overwhelming majority of cases. A single scanned A4 page stored at 300 dpi is roughly 2480 by 3508 pixels — about 8.7 million pixels of photographic data. Ten of those pages is nearly 90 million pixels, and that is where the megabytes live.
Embedded fonts. A PDF can carry its own copy of every font it uses, sometimes more than one copy if the file was assembled from several sources. Fonts are usually small, but a font-heavy manual in a dozen scripts adds up.
Redundant objects. Files that were merged, re-saved, annotated or exported repeatedly tend to accumulate objects nothing references any more — old versions of a page, abandoned thumbnails, duplicate resources. They inflate the file and are invisible to the reader.
High-resolution or lossless artwork. Brochures and product sheets exported from design software can embed print-resolution images, or store them losslessly, when a screen or email copy would need a fraction of that data.
Uncompressed image streams. Not every exporter packs its images well. Some embed bitmaps with little or no compression, and those files can be ten times larger than they need to be.

The practical consequence is worth stating plainly before we go further: if your PDF is mostly text and vector lines, there is very little to remove. If it is mostly photographs or scanned pages, there is a great deal. Everything below follows from that one fact.

What compression actually does to a PDF

"Compressing a PDF" is an umbrella phrase covering two very different families of technique. They behave differently, and it helps to know which one a tool is doing before you run it.

Lossless steps: reorganise, discard nothing you can see

Lossless operations rewrite the file's structure without changing how any page looks. They deduplicate byte-for-byte identical objects, drop objects that nothing references, subset fonts down to the characters actually used, and strip optional extras such as embedded thumbnails or redundant metadata. Nothing visible changes, and the file is byte-perfect in terms of appearance. The catch is scale: on a tidy, freshly exported PDF, lossless cleaning typically saves somewhere between nothing and a few percent. It is a useful tidy-up, not a rescue plan.

Lossy steps: resample and re-encode the images

The dramatic reductions come from the images. Each embedded image is decoded to raw pixels, scaled down if it exceeds a maximum pixel dimension, then re-encoded — usually as JPEG — at a chosen quality level. JPEG is a lossy format: it discards fine detail that the human eye is unlikely to miss, and the higher the compression the more it discards. Done with a sensible ceiling, the page still looks sharp on screen and prints fine; done aggressively, small text smears, thin rules break up, and pale scans turn blotchy.

That is precisely the trade the tool on this site offers. It finds the JPEG images inside your PDF, re-encodes them at a size and quality you select, and writes them back into the document. It leaves your text, links and page geometry alone, because they were never the problem. It never enlarges an image, so a small graphic is left as it is rather than being blown up and re-compressed for no reason.

The tool also tells you the truth. After each run it reports the exact before and after size, and if it cannot help — because the file has no large images to re-encode — it says so instead of claiming a saving it did not achieve.

Scanned PDFs and text PDFs are not the same problem

People are disappointed by PDF compressors for one reason above all others: they expect a text document to shrink like a scan. It will not, and the reason is structural.

A scanned PDF — or a phone photo saved as PDF — is a stack of photographs. Each page is one large image, typically 200 to 600 dpi, and the file has almost no text objects at all; if it was not run through optical character recognition, there is nothing to select or search. Because the pages are images, re-encoding them is enormously effective. Reducing a 2 by 3 inch photo from 3400 pixels wide to 1600 pixels wide removes roughly 78% of the pixels before quality even enters the picture, and JPEG compression does more on top. A 15 MB phone scan dropping to 1 to 3 MB is normal, not exceptional.

A born-digital PDF — a contract, invoice, report or ebook exported from a word processor — is a different animal. Its pages are described by text strings, font selections and vector drawing commands, which are already compact. A 40-page typed report can easily be 400 KB. There are simply no giant images inside it for the compressor to work on, so the honest result is a file that barely changes. Any service promising to turn a 400 KB text PDF into 60 KB is either degrading the content into pictures or not telling you what it did.

Mixed documents sit in between: a typed report with a few scanned receipts appended, or a scanned page with a photo pasted into it. Here the compressor will shrink the image-bearing pages and leave the rest untouched, so the saving is real but partial.

A quick rule of thumb: look at the page count and the file size. Under about 200 KB per page, your file is probably mostly text and will not shrink much. Over about 1 MB per page, it is almost certainly packed with images and will compress well.

What you lose when you compress, and when not to compress

Compression is not free. The smaller you push a file, the more detail you give up, and there are documents where that trade is a bad idea regardless of how good the result looks on screen.

What degrades first is fine detail: small print inside a scan, thin hairlines in a chart, faint pencil marks, subtle gradients, and the crisp edges of small logos. Photographs tolerate heavy compression gracefully; diagrams, tables and small annotated text do not. If a page is going to be printed at full size, projected, or read closely on a large monitor, prefer the Light setting or skip compression.

Put the file down and walk away — do not compress — in these cases:

Digitally signed documents. A signature covers the exact bytes of the file. Rewriting it to shrink the images invalidates the signature, and the recipient's reader will report the document as modified.
Documents under an archiving or legal standard such as PDF/A, or any file an auditor, court or regulator may examine. These have fixed requirements, and altering embedded images can break conformance.
Master and source copies. Keep an uncompressed original. Compress a working copy for email or upload, never the only copy you have.
Scans you still intend to run through OCR. Downsampling too far makes characters harder for recognition software to read, so if you need a searchable text layer, OCR first and compress afterwards.
Print-critical artwork — business cards, packaging, large-format posters. The printer needs the resolution you are thinking of throwing away.

Always keep the original. Compression is a one-way quality trade: once detail has been discarded it cannot be restored. Save your compressed copy under a new name and keep the untouched original somewhere safe, so a file that shrank nicely for email is never the last surviving version of a document.

Step by step: compressing a PDF without uploading it

The whole operation happens in one browser tab. Nothing is sent anywhere, there is no temporary copy on a server, and there is no queue in somebody else's storage waiting to be swept up.

Open the Compress PDF tool. It works in Chrome, Firefox, Safari and Edge on desktop and mobile — no account, extension or install.
Drag your PDF onto the drop area, or click Select file and pick it from your device. Files up to 100 MB are accepted; very large files simply need a lot of device memory to decode.
Choose a compression strength:
- Strong (smallest) — resizes images to a maximum of 1600 pixels on the long edge and re-encodes at quality 50. The biggest savings, best for scans you only need to read on screen.
- Balanced (recommended) — 2200 pixels at quality 68. The default for good reason: a large reduction with little visible difference on a normal screen or home print.
- Light (best quality) — up to 3000 pixels at quality 82. For files you still want to look crisp, or may print at size.
Click Compress PDF. The tool decodes the embedded JPEG images in your browser tab, re-encodes them at your chosen setting, and rebuilds the file. Images are never enlarged, so small graphics are left untouched.
Read the report. It shows the original size, the new size and the saving achieved — and states plainly when a file had nothing worth compressing, rather than pretending otherwise.
Download the smaller PDF. Your original file on disk is not modified.

Shrink a PDF now, without uploading it. The Compress PDF tool runs entirely in your browser. Your document is processed on your own device and never transmitted to a server.

Open Compress PDF

How much smaller can a PDF actually get?

Any honest answer is a range, because the result depends on what is inside your specific file. These are the outcomes you can reasonably expect from the Balanced and Strong settings:

Phone-photo scans (10–20 MB): the most dramatic case. Strong setting commonly brings a 15 MB file down to 1–3 MB, an 80–90% reduction, because each page is a multi-megapixel photo far larger than any screen needs.
Flatbed scans of documents and books (2–6 MB): typically a 40–75% reduction at Balanced, with readable text preserved. This is the setting most people want for emailing forms and receipts.
Image-heavy brochures and product sheets: frequently 50–80% at Strong, more if the export embedded print-resolution images at full size.
Text-based contracts, invoices and reports: often no meaningful change at all, because there are no large images inside to re-encode. Expect a few percent at best.
Already-optimised files: if somebody has compressed the PDF before, a second pass finds little. Running it twice does not help; the tool will report that it could not improve the file.

Two cautions come with those numbers. First, do not chase the smallest possible file if the document still has to be read carefully — pick the lightest setting that meets your size target and stop there. Second, check the result before you delete or replace anything: open the compressed PDF, scroll to the densest page, and zoom in. If small print is still legible and the diagrams are intact, you are done. If not, go back and choose a gentler setting.

Compressing a PDF on your phone

The tool is an ordinary web page, so it works in mobile Safari and Chrome with nothing to install — which matters, because the files people most often need to shrink are the ones they just photographed or received on a phone.

iPhone and iPad: open securepdfnow.com/compress in Safari, tap Select file, then Browse and pick the PDF from On My iPhone, iCloud Drive or Files. Choose a strength, run it, and the compressed file lands in Files → Downloads.
Android: open the page in Chrome, tap Select file, and choose the PDF from Downloads or Drive.
Keep files modest: a phone has less memory than a laptop. Files under about 30 MB are comfortable on a phone; a 90 MB scan may make the tab sluggish, so compress very large files on a computer if you can.
Mind the battery and heat: decoding and re-encoding hundreds of images is real work. On a long scan, keep the screen from sleeping until the progress bar finishes and the download is saved.

Because the work happens locally, compressing on a phone is also the safest option when the document is sensitive. A photo of a passport or a payslip can be shrunk for an upload without first being handed to a compression website — which is exactly the kind of file you would rather not send through a third party in the first place. If you want to understand how to tell a local tool from a server-based one, we walk through the checks in why uploading a PDF is a privacy risk.

Frequently asked questions

Is my PDF uploaded to a server?

No. The Compress PDF tool decodes and re-encodes the images inside your file using JavaScript running in your own browser tab. Your document is never transmitted anywhere, so there is no copy on our side to store, delete or lose — there never was one.

Why did my PDF barely shrink?

Almost always because it is a text-based PDF rather than a scan. Compression works on embedded images, and a document made of text and vector lines has few large images to re-encode. A 400 KB typed report has little to give. If your file is a scan and still will not shrink, it may already have been compressed before.

What do the three settings actually do?

Strong downsizes images to a maximum of 1600 pixels on the long edge and re-encodes at quality 50. Balanced uses 2200 pixels at quality 68. Light keeps up to 3000 pixels at quality 82. Images are never enlarged, so a small image is left alone. Balanced is the right choice for most documents.

Will compressing break my document?

Your text stays selectable and links and page layout are preserved, because only the embedded images are re-encoded. Quality can suffer at the Strong setting — small print and fine lines soften first — so keep the original and check the result on the densest page before you rely on it. Do not compress a digitally signed document, as rewriting the file invalidates the signature.

Is there a file size limit?

100 MB per file. Very large files need a lot of device memory to decode and re-encode, so we block anything above that rather than let your browser tab crash. On a phone, files under about 30 MB are comfortable.

Can I compress a password-protected PDF?

Not while it is locked — an encrypted PDF cannot be decoded until it is opened. Remove the password first with the Unlock PDF tool, then compress it. If you need the protection back afterwards, re-apply it with Protect PDF — but remember that adding encryption rewrites the file, so compress first and protect last.

Once the file is smaller you can combine it with other documents using Merge PDF, or build a fresh PDF from images with Image to PDF — every tool on this site runs locally on your device.

See all free PDF tools