Quick answer
To compress a PDF, choose the file, keep the “Keep text” method, set the compression level (70% by default) and press “Compress PDF”. The tool recompresses the images inside the file while text stays selectable and searchable. For scans that must be even smaller, pick “Strong”: every page becomes a JPEG image and the text is no longer selectable. Everything runs in your browser; nothing is uploaded.
What does compressing a PDF do?
Most of the weight of a heavy PDF sits in its images: photos, product shots, pages scanned on a copier or a phone. Compressing a PDF means re-encoding those images at a quality that still looks right, shrinking images that are far larger than the page needs, and saving the file in a leaner structure. Text, fonts and vector lines stay untouched, so the document is still crisp when you zoom in.
How to compress a PDF
- 1
Press “Choose file” or drop one PDF (up to 200 MB) onto the box. The tool reads its page count and size.
- 2
Pick a method: “Keep text” (recompress the images inside, text stays selectable) or “Strong” (redraw every page as an image, best for scans).
- 3
Drag the compression level: left for a smaller file, right for sharper images. 70% is a balanced choice for on-screen documents.
- 4
Press “Compress PDF”. Compare the original and compressed sizes, and use “View compressed PDF” to check the pages before downloading.
- 5
Press “Download compressed PDF”. If the result is not smaller, the tool says so and suggests keeping the original.
How the two methods differ
| Keep text | Strong | |
|---|---|---|
| What gets compressed | Only the embedded images (JPEG, raw images) | The whole page: each page becomes a JPEG |
| Text can be selected / searched / copied | Yes, unchanged | No |
| Best for | Reports with photos, catalogues, slides saved as PDF | Scans, photographed paperwork, image-only files |
| Text-only files (no images) | Small gain, mostly from leaner saving | May end up larger than the original |
| Links, bookmarks, form fields | Kept | Lost |
Compression level and image size after compression
With “Keep text”, images whose longest side exceeds the limit below are scaled down; with “Strong”, pages are redrawn at the matching resolution.
| Level (slider) | Keep text: longest image side | Strong: page resolution | Good for |
|---|---|---|---|
| Below 50% | 1,400 px | 96 DPI | Messaging apps, email size limits |
| 50–69% | 1,800 px | 96–120 DPI | Reading on a phone, online submissions |
| 70–84% | 2,400 px | 120 DPI | Reading on a computer, office printing |
| 85% and up | 3,000 px | 150 DPI | Sharp prints, detailed photos |
Swipe sideways to see the whole table →
An image is only replaced when the new version is at least 10% smaller; already well-compressed images, tiny images, CMYK (print) images and transparency masks are kept as they are so colours stay correct.
Real-world examples
A job application with a photo and scanned certificates
Use “Keep text” at 70%: the certificate scans are recompressed, while the CV text can still be searched and copied by a recruiter.
An upload portal with a size limit
A copier scan of several tens of MB: choose “Strong” at 55–65%, check with “View compressed PDF” that the text is still legible, then upload.
A product catalogue exported from Canva or PowerPoint
Product photos are often far larger than the page: “Keep text” at 70–80% brings them down to a sensible size and links keep working.
Emailing a contract
A text-only contract written in Word is already light; the tool will report that compression does not help much. Keeping the original is best.
Common problems and fixes
- It says the compressed copy is not smaller.
- The file is mostly text, or its images are already well compressed. Try “Strong” for scans; for text documents, keep the original.
- It says the file is password protected.
- A PDF with an open password cannot be read for compression. Open it in a PDF reader, enter the password, choose Print → Save as PDF and compress the new copy.
- After “Strong” I cannot search the text.
- That is expected: every page is now an image. Use “Keep text” if you need searchable text, or run each page through Image to Text (OCR).
- Images look blurry and small print in them is hard to read.
- Move the level to the right (85% or more) to keep images larger and sharper, then compress the original again.
- The browser runs out of memory on a very long file.
- Close other tabs, or use Split PDF to cut the file into parts, compress each part, then Merge PDF.
Why are some PDFs so large?
A page of text in a PDF weighs only a few KB because the text is stored as characters plus a font. A 12-megapixel phone photo, on the other hand, can weigh 3–5 MB, and when it is placed in Word or Canva and exported to PDF it is often kept at full size even though it only fills a few centimetres of the page.
Scans are heavier still: every page is one big image, and many copiers save at 300–600 DPI with light compression. The most effective fix is to bring images down to the size you actually need and re-encode them as JPEG at a sensible quality.
What the tool does with your file
With “Keep text”, the open-source pdf-lib library reads the file structure, finds each embedded image, decodes and re-encodes it with the browser, replaces only the images that actually get smaller, and saves the file with compressed cross-references (object streams). This runs in a background thread so the page stays responsive.
With “Strong”, each page is drawn by pdf.js (Mozilla's library, also Firefox's PDF viewer) into a JPEG image and assembled into a new PDF with the same page sizes. Every step happens on your device; the file is not sent anywhere.
- Digitally signed files: compressing changes the content, so the signature is no longer valid. Compress before signing.
- Form fields, links and bookmarks are kept with “Keep text” and lost with “Strong”.
Glossary
- Embedded image
- An image stored inside the PDF file, such as a photo, a logo or a scanned page.
- JPEG (DCTDecode)
- Lossy image compression that works very well for photos; PDF calls this filter DCTDecode.
- DPI
- Dots per inch (2.54 cm). An A4 page at 120 DPI is about 990 pixels wide.
- Object stream
- A way of packing many small PDF objects into one compressed stream so the file is leaner (PDF 1.5 and later).
- Rasterizing
- Turning a whole page (text, shapes and images) into a grid of pixels, like a screenshot of the page.
Frequently asked questions
Is my PDF uploaded to a server?
No. The file is read and compressed in your browser, so it is safe for contracts and personal papers.
Does compressing a PDF remove the text?
Not with “Keep text”: only the images inside are recompressed, and the text can still be selected, searched and copied. “Strong” turns each page into an image, so the text is no longer selectable.
How much smaller will my file get?
It depends on the file. Files with large photos or scans usually shrink a lot; text-only files barely change because they are already small. The tool always shows the size before and after so you can decide.
How do I get a PDF under 1 MB or 2 MB?
Try “Keep text” at 50–60% first. If that is not enough, use “Strong” at 40–55%. Very long files can be split with Split PDF.
Can I compress a password-protected PDF?
Not if it needs a password to open. A PDF that only restricts printing or editing but opens normally can be compressed.
Is there a page or size limit?
One file at a time, up to 200 MB. There is no hard page limit, but very large files need a lot of browser memory.
Ratings and comments
Loading…