Compress a PDF
to a target size

or drop one here
Compression level
Target size
Choose a PDF file first.

This page shrinks a PDF to a size you name, for example 200 KB for a form field or 2 MB for an attachment, and the file never leaves your computer while it happens. Four local techniques take turns beating your number, every result is opened again and checked before download, and there is no account, no queue and no watermark.

Shrinking a PDF in four moves

  1. Bring in the file

    Press Choose a PDF or drop the document onto the page. It opens inside this tab, on your own machine, so a confidential contract gets the same treatment as a flyer.

  2. Name your ceiling

    Pick a level (Highest, Average or Low), tap a preset chip such as 100 KB, 200 KB, 500 KB, 1 MB or 2 MB, or type the exact cap into the KB or MB box. The number works as a ceiling: the engine tries to land below it, never on it.

  3. Press Compress and let the methods race

    One press starts the work on your own processor. A structural rebuild, image re-encoding and font subsetting run first, and a full page redraw at 150 dpi steps in only when nothing else reached the target.

  4. Check the result, then keep the file

    The result shows the new size, the percentage saved and the original size, next to a side by side preview. Press Download PDF when the numbers satisfy you; the file arrives named after the original with -compressed added, for example report-compressed.pdf.

If you want the full tour of what each method drops and why some documents resist, how it works walks through all four. When the finished file still overshoots your cap, size breakdown shows which pages and pictures carry the weight.

One evening, three uploads, one 500 KB cap

Picture Maya, a freelance bookkeeper, finishing a client pack at 9 pm. The tax portal accepts one PDF per document and stops at 500 KB. Her scanned receipts sit at 4.1 MB, the signed engagement letter at 1.8 MB, and the spreadsheet export at 900 KB. Three files, one cap, and a midnight deadline.

She runs each through this page with 500 KB typed in. The receipts, mostly phone photos of paper, land at 460 KB through image re-encoding. The letter, two pages of text with one signature scan, drops to 300 KB on the structural pass alone. The spreadsheet export barely moves, because numbers and grid lines were never heavy to begin with, and she learns that without uploading anything anywhere.

Your evening probably looks similar: a portal with a number, a file above it, not much time. Type the number the recipient printed, keep a small margin below it (forms that count a kilobyte as 1,000 bytes are about 2.3% stricter than this tool, which counts 1,024), and let the preview confirm the pages still read well before you send.

What the levels and presets actually set

The three levels are shortcuts for a fraction of your file: Highest aims at a quarter of the current size, Average at one half and Low at three quarters. They exist for the common case where nobody gave you a number and you only want the file roughly smaller. Average is preselected, and for a first attempt it is the sensible choice.

The preset chips are the opposite: exact ceilings for recipients who did give you a number. Tapping 500 KB fills in that cap for a job portal; tapping 2 MB fills it for a roomy attachment. Anything else, a 300 KB exam field or a 750 KB tender box, goes into the text box with KB or MB beside it. The tool refuses caps at or above the current size, since a file that already fits needs no work.

During the run, the progress steps narrate which methods are working, and the Details panel below the result keeps the full report, including which method won. That name is worth reading: a structural win means the pages are pixel identical to the original, an image win means pictures were re-encoded to fit, and a redraw win means the pages were rebuilt at screen resolution with the text laid back on top, still selectable. The guide to reducing PDF size maps common recipient caps to the setting that usually satisfies them.

When the result misses, and what fixes it

The most common miss is a target the file cannot honestly reach: twenty scanned pages squeezed toward 200 KB will either land above the number or come back soft. Read how far short the run stopped, the page states it plainly on a miss, then cut pages before cutting quality. Delete PDF pages drops the cover sheets and blank versos scanners love to add, and Split PDF turns one impossible file into parts that each fit the cap with room to spare.

The second miss is a locked file. A PDF that asks for a password on open resists the rebuild tricks compression relies on, so those methods skip it and the size barely moves. Run it through Unlock PDF first, then compress the unlocked copy.

The third is impatience with a giant scan: a 300 MB archive on a phone browser can stall the tab, because source, parsed document and candidates all live in tab memory at once. Move to a desktop, close other tabs, or split first. If the pages only need to be pictures, PDF to JPG pulls them out as images instead of sending a document at all.

Jobs this page is the wrong tool for

Sometimes the honest answer is a different route. A file headed for professional print should not be compressed here at all: the redraw floor of 150 dpi suits screens, not presses, so ask the sender for a print ready export or run print preflight before you touch anything. A password you do not have belongs with the sender or your IT desk, not with a guessing game. And an archive of hundreds of pages that must stay in one piece is calmer work for a desktop app with no tab memory ceiling; this page stays the right call for the everyday case of one file, one cap, one evening.

Short answers before you press Compress

Does Highest quality mean the best looking result?
No, and the name trips people up. Highest means the strongest shrinking (a quarter of the original size), Low the gentlest (three quarters). If looks matter most, start at Low or with a roomy exact cap, and check the preview.
Why does the tool refuse my target size?
Because the cap must sit below the current file size. A 900 KB file with a 2 MB target has nothing to do, so the page caps the box at just under the file size. Type a smaller number or pick a lower preset.
Will the compressed file keep working in every reader?
Yes. The output is a plain PDF with the same pages and the same document details (title, author and the rest), verified by reopening it and recounting pages and characters. Anything that kept less than 98% of the text is refused and never offered for download.
Can I compress a scan of handwriting?
You can, but budget generously. Handwriting and fine print soften first when image quality drops, so give each page plenty of kilobytes, zoom the preview to full size, and read the smallest line before downloading.
What if the file must stay exactly as sharp as the original?
Then aim high and let the lossless methods win: a structural rebuild plus font subsetting changes no pixels at all. The page on compression without visible change is built around exactly that goal.
Is there a daily limit or an account?
Neither. No counters, no sign in, no watermark. The only ceiling is your tab memory on truly huge files.

More PDFGym tools that run beside this one

Each tool below works the same way: inside your browser, checked by reopening before download, and free without signing in.

Merge PDFs →Bundle invoices or chapters into a single file
Organize Pages →Reorder, rotate, split, and delete pages visually
Rotate & Crop →Straighten crooked scans and trim ugly margins
Images to PDF →Turn JPG and PNG photos into one compact PDF
HEIC to PDF →iPhone photos straight to PDF, no converter apps
TIFF to PDF →Stack multi-page scanner TIFFs into one PDF
WebP to PDF →Web images reborn as a portable PDF
BMP to PDF →Ancient bitmaps converted into modern PDFs
Privacy Scan →Find and strip hidden metadata before you share
Protect & Unlock →Lock a file with a password - or free a locked one
Repair PDF →Rebuild corrupted files that refuse to open
PDF Size Breakdown →See which pages, images, and fonts eat the bytes
Print Preflight →DPI, CMYK, and bleeds checked before the press
PDF to Images →Crisp JPG or PNG pages for slides and the web
PDF to TIFF →One multi-page TIFF for print and archive workflows
Extract Images →Pull out the original full-resolution photos
PDF to Markdown →Headings, lists, and tables as clean Markdown
Markdown to PDF →Notes and docs rendered into a clean PDF
PDF to CSV →Tables lifted out into clean CSV rows
PDF to Excel →Spreadsheet pages rebuilt as real Excel files
Sign PDF →Draw, type, or upload your signature in the tab
Stamp PDF →Page numbers, stamps, and light watermarks
Compare PDFs →Spot text and visual differences side by side
Flatten PDF →Bake forms and comments in, permanently
Fill PDF Form →Complete interactive forms without printing
Create PDF Form →Build fillable forms with fields and checkboxes
N-up & Booklet →Two, four, or more pages on every printed sheet
Comic Archive →CBZ comics converted for any e-reader
Redact PDF →Burn confidential text out for good

Handy targets and deeper reading

Caps people meet most

Very small caps

Roomy caps

Tricky originals

Learn the craft