Optimize
Compress PDF Online – Reduce PDF Size in Your Browser
Three levels. Low keeps the text selectable; higher levels rasterise the pages to get the size down.
Runs on your deviceWorks with: Any PDF. Image-heavy files shrink most; a text-only PDF is already small.
Your files never leave this device — everything above runs locally.
Compress PDF files and reduce PDF size without installing anything
A PDF gets heavy for one of two reasons: it carries large pictures, or it was written wastefully. This page tackles both. Add a file, choose a compression level, press Compress PDF and you get a smaller copy back, along with a line telling you how many kilobytes were saved. If you need to shrink a PDF file for an email attachment, a college portal or a job application, this is the quick route.
Unlike a slider that says "quality 60" and leaves you guessing, the tool offers three named levels and tells you plainly what each one costs. The lightest level changes nothing you can see. The stronger two can turn pages into pictures, and the result card says so whenever that happens. Several PDFs can be added at once and are processed one after another.
How to make a PDF smaller, step by step
- Drop the PDF onto the box that says "Drop files here, or click to browse", or click it and choose the file. You can add more than one.
- Under Compression level, pick Low, Medium or High. Medium is selected by default.
- Press Compress PDF. The button shows a percentage while it works.
- Read the line under "Done". It shows the old size, the new size and the percentage saved, and it states whether the text is still selectable.
- Press Download. With several files you get a button labelled "Download all" that bundles them into one ZIP, and each file keeps its own name with "-compressed.pdf" added.
What Low, Medium and High actually do
Low is a lossless rewrite. The document is loaded and saved again with its internal objects packed together more tightly, which is the same information stored more economically. Nothing is re-drawn, so text stays selectable, links keep working and images keep their original pixels.
Medium and High work to a proportion of the original size. Medium aims for 60 per cent of it or less, and High aims for 35 per cent or less. The tool first checks whether the lossless rewrite already gets there. If it does, you get the lossless file and your text is untouched. If it does not, the pages are rendered and rebuilt as JPEG pictures at falling resolution until the goal is met. That is what produces the big savings on scans, and it is also why the text can no longer be selected afterwards.
| Level | Aim | Method | Text stays selectable |
|---|---|---|---|
| Low | Whatever a safe rewrite saves | Lossless re-save | Always |
| Medium | 60% of the original or less | Lossless if enough, otherwise pages become JPEG images | Only if the lossless rewrite was enough |
| High | 35% of the original or less | Lossless if enough, otherwise pages become JPEG images | Only if the lossless rewrite was enough |
How the compressor works behind the scenes
Everything is done with two open libraries running in your browser tab. pdf-lib reads the file and writes the lossless version. pdf.js, the renderer that Firefox uses to display PDFs, draws each page onto a canvas when the rasterising route is needed, and the canvas image is saved as a JPEG and placed on a fresh page of the same size.
The rasterising route is a ladder of six attempts. Each one lowers the render scale and the JPEG quality together, from 1.5 times at quality 0.8 down to 0.55 times at quality 0.3. Because a page is 72 units to the inch, those scales correspond roughly to 108 dpi at the top of the ladder and 40 dpi at the bottom. The tool stops at the first attempt that reaches your level's goal.
Every rebuilt page keeps its original width and height in points. Only the picture on it is smaller. That matters for portals that preview the first page, because the page does not suddenly look cropped or oversized.
A worked example with real numbers
Suppose you have a 4,800 KB PDF made from phone photographs of a marksheet, and you want it small enough to email. Because the pages are already JPEG pictures, the lossless rewrite barely helps: say it lands at 4,700 KB.
On Medium, the goal is 60 per cent of 4,800 KB, which is 2,880 KB. The lossless file is far above that, so the tool rasterises. Suppose the first attempt, at 108 dpi, comes out at 2,600 KB. That is under 2,880, so it stops there. The saving is 4,800 minus 2,600, which is 2,200 KB, and 2,200 divided by 4,800 is 45.8 per cent, shown as 46% smaller.
On High, the goal drops to 35 per cent of 4,800 KB, which is 1,680 KB. The 2,600 KB attempt is too big, so the ladder continues to lower resolutions until a rebuild lands under 1,680 KB or the ladder runs out. These sizes are illustrative, because a real scan depends on how much detail is on the page, but the arithmetic and the order of the steps are exactly what the tool does.
Reading the result line
The line under "Done" is the honest part of the page, so it is worth reading before you download. A typical result looks like "4800 KB → 2600 KB (46% smaller)", followed by either "lossless, so the text is still selectable" or "pages were rasterised to reach this level, so the text is no longer selectable".
If the file was already compact, you will see a message that it is already well optimised and that nothing safe could be squeezed out. That is a true answer, not a failure. A plain typed document of a few pages has little to remove, and no tool can make it dramatically smaller without turning it into pictures.
When you have a number to hit instead of a level
Levels are proportions, not promises. If a portal says the file must be under 200 KB, you cannot know in advance whether Medium or High will land there. For that job, use the targeted compressor, where you type the limit and the tool steps down until it fits. The usual order is to try Low or Medium here first, and move to the targeted tool only when a hard limit is involved.
Common uses for a smaller PDF
- Email attachments that bounce because a mailbox caps messages at around 20 to 25 MB.
- College and university submissions, where a portal accepts one PDF up to a few megabytes.
- Job applications that combine a CV, certificates and ID proofs, often after you scan photos into a PDF from a phone.
- Government and bank forms that ask for address proof or income documents under a size limit.
- Sharing over WhatsApp, where a 40-page scan sent at full size is slow on a mobile connection.
- Archiving old scanned files so a folder of records takes a fraction of the disk space.
Tips for getting the smallest file with the least damage
Start at the gentlest level and only go further if you need to. Low costs nothing, and Medium only rasterises when it must, so you often keep selectable text for free.
If a document mixes typed pages with one large scanned appendix, it is worth separating them first. Split the file to pull out the heavy pages, compress those on their own, and then join the parts together again. The typed pages keep their text and the scan gets the heavy treatment.
Photographs saved at 12 megapixels inside a Word document are the most common cause of a bloated PDF. Reducing the pictures before they go into the document saves more than any later compression. For a loose image that needs to hit a size, reduce the file size of the picture itself.
Always open the compressed file and zoom to the smallest text on the page before you send it. A signature, a stamp or a small table is where over-compression shows first.
Mistakes that waste time
- Expecting a text-only PDF to shrink by 70 per cent. If the file is mostly typed words, there is little to remove and the tool will say so.
- Choosing High by habit. On a scan it can push resolution down to a level where small print becomes hard to read, and a file that is hard to read is a worse problem than a file that is a little large.
- Compressing a file you need to search or copy from. After rasterising, the words are a picture. Keep the original, or choose Low.
- Compressing the same file again and again. Each rasterising pass starts from the already degraded picture, so quality only falls.
- Overwriting your original. The output is a new file called compressed.pdf, but keep the source anyway in case you need a sharper copy later.
Limits and honest limitations
Each PDF can be up to 100 MB and 500 pages, because the work happens in your browser tab and the ceiling is your device's memory. A file beyond either figure is refused with a clear message before anything is processed.
A password-protected file is refused, with a pointer to removing the password. A damaged file is refused with advice to repair it first.
The rasterising route does not reach every goal. If even the lowest rung of the ladder cannot meet the level's aim, the tool still returns the smallest version it produced, provided that version is smaller than the lossless one. It does not warn you that the aim was missed, so compare the sizes on the result line against what you wanted, and look at the pages. For a hard limit, the targeted tool reports a miss explicitly.
Form fields, bookmarks and annotations survive a lossless rewrite in most cases but are flattened into the picture by the rasterising route. Treat a rasterised copy as a finished, read-only version for sending, not as a working file.
Privacy: your file stays on your device
Compression runs entirely in your browser. The PDF is read from your disk into the page's memory, processed there, and offered back as a download. It is not uploaded to a Jobsixo server, so an identity document or a marksheet never leaves your computer or phone. There is no account and no watermark on the result.
Related tools
Other ways to lighten a PDF are to delete pages you do not need or to turn each page into JPG pictures when a portal wants images. When the size is fine but the file will not open, repair comes before compression.
Frequently asked questions
How can I reduce the size of a PDF file for free?
Add the PDF, choose a compression level and press Compress PDF. The tool is free, needs no account and adds no watermark. Start with Low or Medium, and read the result line to see how much was saved.
Will compressing a PDF make it blurry?
Low does not change a single pixel, so it cannot blur anything. Medium and High can lower page resolution when the lossless rewrite is not enough. On screen this is usually fine, but zoom in on small print to check before you send the file.
Why did my PDF barely get any smaller?
Most likely it was already efficient, or it contains mostly typed text, which has very little to remove. The result line will say the file is already well optimised. Files full of photographs and scans shrink the most.
Can I still select and copy text after compression?
Yes after Low, and after Medium or High when the lossless rewrite alone reached the goal. If the pages had to be rasterised, the text is part of a picture and cannot be selected. The result line tells you which case you are in.
Is there a size limit for the PDF I can compress?
Each file can be up to 100 MB and 500 pages. Larger files are refused before processing, because the work is done in your browser's memory.
Is my PDF uploaded to a server?
No. Compression runs inside your browser tab and the file is never sent to a Jobsixo server.
Can I compress several PDFs at once?
Yes. Add them all, pick one level, and each is processed in turn. The results come back as separate files named after the originals, with a button to download them together as a ZIP.
How do I compress a PDF to an exact size such as 100 KB?
Use Compress to size instead. It takes the number you were given and keeps lowering quality until the file fits, or tells you plainly that the limit cannot be reached.
Can the compressed file end up bigger than the original?
It should not for a normal file, but a PDF that was already tightly packed can come out a few bytes larger after the rewrite. If the result line shows no saving, keep your original.