Skip to main content

Free PDF Compressor

Compress PDF files to reduce their size. Choose a quality level — higher compression means smaller files but lower image fidelity.

Smallpdf's free plan allows 2 tasks a day — here PDFs are compressed on your device, with no account. · checked Sep 2026

Drop files here, or click to browse

PDF · processed on your device

or press Ctrl+V to paste from your clipboard

You can drop multiple files at once

Browse files

Your files never leave your device. All processing happens locally in your browser.

Scanned or image-heavy PDFs shrink the most with the image options. Text-only PDFs are usually already small — if nothing helps, we will tell you the file is already optimized rather than hand you a bigger one.

Common uses

  • Compress PDF for EmailShrink PDF files to send via email — stay under attachment limits and ensure recipients can download them quickly.
  • Compress PDF for UploadReduce PDF size before uploading to online platforms — meet file size limits and speed up transfers.
  • Compress PDF for ArchivingOptimise PDFs for long-term storage — save disk space, fit more documents, and keep archives organised.
  • Compress PDF for SharingMake PDFs easy to share — smaller files download faster and work better on messaging apps, cloud drives, and links.
  • Compress Legal PDF DocumentsReduce legal PDF file sizes without sacrificing readability — perfect for contracts, filings, and court submissions.

How It Works

Most "free PDF compressor" sites silently upload your PDF to their servers. Here, compression happens entirely in your browser using pdf-lib — a popular JavaScript library that reads and rewrites PDF files locally.

  1. 1
    You drop a PDF
    The file is read into your browser's memory as a byte array. Nothing leaves your device.
  2. 2
    pdf-lib parses the PDF structure
    PDFs are essentially containers of fonts, images, and instructions. pdf-lib reads the structure and identifies which embedded images can be re-encoded at lower quality to reduce size.
  3. 3
    Images are re-encoded at your chosen quality
    You choose the quality level. Embedded JPEG images are re-encoded with that quality setting. The lower the quality, the smaller the output PDF — but text and vector content remain unchanged.
  4. 4
    A new PDF is written and offered for download
    pdf-lib reassembles the PDF with the compressed images and writes the result to a new in-memory file. Click Download to save it locally.

Safe for contracts, tax returns, medical records, anything sensitive. Your PDF and its contents never reach our server — there is no server in this workflow.

Guide

Worked example

Press "Try a sample". It adds sample-scan.pdf, and the left side shows a thumbnail of page 1, the file size and "2 pages". A file up to 8 MB and 40 pages starts on its own; for a bigger one, press "Compress" on each card. The cards are "Optimize (keep text)", "Compress as images" and "Maximum compression". A finished card shows its size, a "↓" line ending "% smaller" and "Download", and the smallest gets a "Best" badge. Now click "1 MB" under "Compress to a target size" and press "Compress to target". The tool first shrinks only the images, and may say "Reached by shrinking only the images, so the text is still selectable." If that cannot fit, it renders pages as images and says "Text-keeping mode could not reach this size, so the pages were rendered as images (text is not selectable)."

Which settings to use, and the limits

  • Pick the mode by what the PDF contains. "Optimize (keep text)" clears the title, author, subject and keywords, rebuilds the file structure and shrinks oversized images. Text and links stay as they were. "Compress as images" draws every page at 150 DPI and saves it as a JPEG at quality 72; "Maximum compression" does the same at 96 DPI and quality 60. The new file holds only those page pictures on a white background, so text can no longer be selected, and the page size is kept. Badges say "searchable" or "image". Scans and photo-heavy PDFs shrink most; text-only PDFs are usually small already.
  • Choose an Image quality for Optimize. "Image quality" has "High", "Medium" (the default) and "Low". They cap the long edge of each image at 2000, 1600 and 1000 px and save it again as a JPEG at quality 80, 72 and 50. Only images of at least 24 KB are touched, and a new image is kept only if it is at least 10% smaller. Plain JPEG and 8-bit grey or RGB images are handled; CMYK and indexed-colour images are left alone. Picking another level runs Optimize again if it had already finished.
  • Use a target size for upload limits. Click 100 KB, 200 KB, 500 KB, 1 MB, 2 MB, 5 MB or 10 MB, or type megabytes (from 0.01) in the custom box, then press "Compress to target". The KB buttons are exact: 100 KB means 102,400 bytes, although the box shows 0.098. The tool tries Optimize at "High", then "Medium", then "Low", and keeps the first that fits. Only then does it turn pages into images, testing 150, 96 and 72 DPI and a JPEG quality between 30% and 85%. It judges this from the first, middle and last pages, so the real file can land a little over.
  • Expect a wait with big or many-page files. Up to 8 MB and 40 pages, all three cards run as soon as the file loads. Above either number the page says "This is a large PDF, so each option runs only when you click it (image options can take a little while)." The whole file is read into memory, each run works on its own copy, and the image options draw every page on a canvas, so a very large scan can run out of memory on a small device. If a run fails, try "Optimize (keep text)" first, which does not redraw whole pages.
  • Drop several PDFs to compress them with one setting. Drop two or more PDFs, or press "Add more files" while one is open, and the page shows "Compression for every file" with four choices: the three modes and "Compress to a target size". The button reads like "Compress 3 PDFs", and the files run one at a time. Each row notes the saving, such as "↓ 42% smaller". "Download all (3)" gives a ZIP named compressed-pdfs.zip. A file that would not get smaller stays as it is, with the note "not smaller, original kept", and the original goes into the ZIP.
  • Check the download before you send it. Each card has its own "Download", and the "Best" card also gets a large one that shows its size. Optimize, and a target reached by shrinking images, save as name_optimized.pdf; the image modes save as name_compressed.pdf. Open the file and check small print, and any text you need to copy or search. A "Do more with this file" row can pass the result to "Sign PDF" or "Password protect" inside this browser. The file is read, compressed and saved in your browser, and nothing is uploaded.

Common problems and how to fix them

A card says "Not smaller", with "This mode didn't make the file smaller."
A result counts only if it is at least 1% smaller than the original, and then it is offered for download. A text-only PDF has little to shrink, and turning a page of plain text into a picture can even make it bigger. Try "Optimize (keep text)", a lower "Image quality", or a target size. If all three cards say so, the page says "This PDF is already well-optimized — none of the options make it meaningfully smaller. Your original is the best version." and offers "Download original".
It says "Could not get under 1 MB — this is the closest this PDF can get."
The lowest setting the target mode uses is 72 DPI at JPEG quality 30%, and the file was still above your size. The size is also judged from three sample pages, so the final file can land a little over. The closest result is still offered. Ask for a slightly bigger target, or split the PDF into parts, or remove pages you do not need, and try again.
It says "report.pdf is password-protected, so this tool can't open it. Remove the password first."
The drop zone looks for an encryption entry at the start and end of the file. That includes a PDF that opens without a password but has editing or printing restrictions. Saving such a file again would make it unreadable, so it is refused. A link such as "Open this file in Unlock PDF →" passes the same file on inside the browser; drop the unlocked copy here afterwards.
It says "Could not read this PDF — it may be encrypted or corrupted."
The PDF reader that counts the pages and draws the thumbnail of page 1 could not open the file. Common causes are a download that stopped part-way, a damaged file, or a file that is not really a PDF. Save or export it again from the program that made it, or download it again, and drop the new copy. Encryption is only one possible cause, even though the message mentions it.
It says "Compression failed. Please try again or try another file."
Something went wrong while one option was working on the file, and the page does not say what. It is shown on that card only, so another mode may still work: "Optimize (keep text)" does not redraw whole pages, while the image modes do. A very large scan can also run out of memory. Try again, try another mode, and for a huge file use a computer with more memory.

How it compares

Every mode shrinks a PDF in one of two ways: it saves the large images again at lower quality, or it turns each page into a picture at a fixed 150 or 96 DPI (down to 72 DPI in target mode). It does not remove fonts, flatten forms, run OCR, or let you set quality for each image. After the image modes, text cannot be selected or searched, and links, bookmarks and form fields are not kept. A desktop PDF program offers finer control over each image and font. In exchange you see all three results side by side, you can aim at a file size, nothing needs installing, and the PDF is handled in your browser.

Related Tools

Frequently Asked Questions