Compress PDF
Shrink a PDF down to a smaller size, entirely in your browser — the file never reaches a server. There are two paths. Lossless structure compression repacks the file internally and strips redundant metadata, leaving text, links, bookmarks and form fields untouched, with typical savings of 5-20%. Page rasterization re-renders every page to JPEG and rebuilds the PDF, cutting size by 50-90% but turning text into pixels you can no longer copy or search. Common situations: an email attachment hits the size cap, an upload form caps you at 5-10 MB, you are sending scans to someone who only reads, or you want lighter PDFs embedded in a site or doc.
- Two compression modes — keep the text layer, or rasterize every page
- Structure mode saves 5-20% and keeps text, links, bookmarks and form fields intact
- Rasterize mode saves 50-90% with three quality presets plus manual DPI and JPEG sliders
- Original file size and page count appear right after upload
- Before-and-after comparison when finished — change parameters and recompress if needed
- Everything runs in the browser. No upload, no server logs, works offline
- Encrypted PDFs fail fast at load time, never half-write a broken output
- Original PDF stays untouched between runs, so you can iterate parameters freely
Drop a PDF here, or click to pick a file
Choose 1 PDF file. The whole process runs locally in your browser.
Repack internals and strip metadata. Keeps text, links and bookmarks. Typical savings 5-20%.
Best for text PDFs, contracts, manuals, and files with outlines or links. It keeps the page intact and preserves copy, search, and navigation.
Pages
—
Original
—
Compressed
—
Saved
—
Upload a PDF and pick a mode — the result will appear here.
Overview
Two compression paths covering the different needs of text-heavy and image-heavy PDFs, with parameters you can tune and predictable results.
- 01
Lossless structure compression
Repacks the PDF internally using pdf-lib and strips redundant metadata. Text, fonts, links, outlines, bookmarks and form fields all stay intact. File size typically drops 5-20%. Best value for text-heavy PDFs.
- 02
Full-page rasterization
Renders every page to JPEG via pdf.js and rebuilds the PDF. The trade-off: text becomes a pixel layer you can no longer copy or search. The win: file size typically falls to 10-20% of the original, biggest gains on scans and image-heavy PDFs.
- 03
Three presets plus manual tuning
High, Standard and Smallest presets are one tap away. Want finer control? Drag the DPI and JPEG quality sliders directly. Presets and sliders share the same parameters, and editing one updates the other.
- 04
Before-and-after comparison
When finished, you see original size, compressed size, savings percentage and the current quality level. Decide at a glance whether to ship it or try different parameters.
- 05
Whole document at once
The current version compresses the entire PDF in one pass — no per-page selection. To compress only a few pages, use PDF Split first to extract them, then come back.
- 06
Page proportions preserved
A4, A3, US Letter, posters or any custom page size all render at their original aspect ratio. Nothing gets squashed or stretched.
- 07
Fail fast on encrypted PDFs
Password-protected files are rejected at load time with a clear error. No silent half-broken output, no minutes wasted before failing.
- 08
Runs fully locally
Upload, read, compress and download all happen in the browser. The file never reaches any server, the tool works offline, and tab memory is released when you close the page.
How to use
Upload, pick a mode, adjust parameters, download. Most files finish in seconds to a couple dozen seconds. No installs.
- 01
Drop a PDF into the upload area or click the button to pick a file. The tool reads file size and total page count right away.
- 02
Pick a mode. Text-heavy PDFs: start with Keep text layer — it preserves features and carries no risk. Scans and image-heavy PDFs: go straight to Rasterize pages for bigger savings.
- 03
In Rasterize mode, pick a preset. High for print or formal use, Standard for everyday sharing, Smallest for size-first, view-only forwarding.
- 04
Skip presets if you want — drag the DPI and JPEG quality sliders to fine-tune. Any change moves the preset to Custom.
- 05
Hit Compress. A few-dozen-MB PDF usually finishes in seconds. Hundreds of MB or hundreds of pages take longer, and progress updates while running.
- 06
Check the savings percentage and the before-and-after sizes. If the result looks rough, bump the preset up one notch and run again.
- 07
Click Download to save the compressed PDF. The original is never overwritten — try as many parameter combinations as you want.
Details
Built around the practical problem of PDFs being too large to send, embed or store. Parameters are tunable and the process is transparent.
- Toggle between Keep text layer and Rasterize pages
- Three built-in quality presets: High, Standard, Smallest
- DPI slider, 50 to 200
- JPEG quality slider, 10 to 100
- Presets and sliders share the same parameters and stay in sync
- File size, total pages and first-page dimensions shown after upload
- Live progress while compressing — see the current page number
- Before-and-after sizes plus savings percentage on completion
- Original PDF stays untouched between runs
- Encrypted PDFs surface a clear error at load time
- Page aspect ratio preserved — A4, A3, US Letter, posters, custom sizes
- In Keep text mode, text, links, bookmarks and form fields are preserved
- Works in Chrome, Edge, Firefox, Safari, Brave, both desktop and mobile
- Runs entirely locally — no server logs, no upload, works offline
Use cases
Oversized PDFs usually come from scans, image-heavy reports and office multifunction printers. Here are situations that actually show up.
-
Email attachment bounces
Most mailboxes cap a single message at 10 or 25 MB. Image-rich reports rarely shrink enough with structure mode — Rasterize at Standard usually does it.
-
Sending to a colleague over chat
WeChat, Slack and Teams make large files painful to upload and download. Compress first, the recipient opens it much faster.
-
Free cloud storage filling up
A single scanned contract can take tens of megabytes, and free cloud quotas fill quickly. A batch run frees real space.
-
Upload forms with hard size caps
Government, HR and application portals often allow only 5-10 MB. If your original is larger, compression is the only path through.
-
Article and newsletter attachments
Shrink long reports before attaching them to articles or newsletters. Readers download faster, and fewer bail out waiting.
-
Resumes and portfolios
Recruiters do not need 600 DPI. Standard preset is usually plenty, the file gets cut in half or more, and reviewers scroll through smoothly.
-
Sending scans to read-only viewers
Scanned books or documents shared just for casual viewing can take Smallest preset — file size drops to roughly a tenth of the original.
-
Output from office scanners
Office multifunction printers often produce 1-2 MB per scanned page, hitting hundreds of MB on a thick stack. Rasterize pulls it back to an email-friendly size.
-
PDFs embedded on a site or in docs
Public download pages, Notion or Confluence attachments — smaller PDFs mean shorter wait times for visitors.
-
Older devices opening slowly
A few-year-old phone or low-spec laptop struggles with tens-of-MB PDFs. Compression brings page-turning back to smooth.
-
Archive cleanup
Hundreds of old PDFs sitting on a drive or in cloud storage. Pick the biggest few, batch compress, reclaim significant space.
-
Cross-border transfers
On flaky international links, smaller files cut wait time noticeably for both sender and receiver.
See also
Only want to compress some pages? First use PDF Split to pull them out. If the real goal is turning a PDF into images for an upload form, use PDF to Image directly. For more PDF and image tools, check File in the category page.
Best practices
Picking a mode and a quality level always involves a trade-off. These patterns hold up in practice.
- Text-heavy PDFs with few images: pick Keep text layer. You preserve search and copy, savings are modest but risk is minimal.
- Scanned contracts, image-heavy reports, office scanner output: go straight to Rasterize. Expect 1/5 to 1/10 of original size.
- Not sure what type of PDF? Try Keep text first. If almost nothing shrinks, switch to Rasterize.
- Walk quality down, not up. If Standard is good enough, stop there. You do not need to push every job to Smallest.
- 180 DPI / quality 80 works for formal sharing or later printing. 120 DPI / quality 50 is fine for on-screen reading. 96 DPI / quality 30 is size-first, with visible artifacts.
- PDFs you plan to print or run through OCR? Avoid Smallest. Print quality drops, OCR accuracy drops.
- A few-hundred-page high-res scan can consume gigabytes of browser memory in one shot. Desktop handles it better; on mobile, split first.
- If a run looks bad, change parameters and recompress from the original — do not stack compression on an already-compressed file.
- Sending to read-only viewers (managers, clients, family)? Smallest is usually enough.
- PDFs with signatures or interactive outlines: Rasterize strips those. Keep the original around.
- Cross-border email and slow uplinks: a smaller PDF saves a lot of waiting on both ends.
- For sensitive files, use a private window and clear the download history when you are done.
Limitations
Compression is always a trade-off. These edges are worth knowing before you commit.
- Rasterize mode turns text into pixels — you can no longer copy or search. Keep the original PDF if you need that.
- Rasterize mode also drops outlines, bookmarks, links, form fields, annotations and digital signatures.
- On PDFs that have already been heavily optimized (Adobe Acrobat saving with optimization, for example), Keep text mode may save under 5%. To go further, switch to Rasterize.
- Encrypted PDFs, corrupted PDFs, strict PDF/A or PDF/X files, and DRM-protected files may fail to load or compress.
- Rasterize uses JPEG, which can introduce mild artifacts on pure-text pages. Pure-text PDFs are not the best match for that mode.
- Very large files (hundreds of MB, thousands of pages) may consume enough browser memory to fail, especially on mobile. Split first to be safe.
- The compressed output is a fresh PDF — original title, author and keywords are not carried over.
- Rendering uses the browser-bundled pdf.js, which differs slightly from Adobe Acrobat in anti-aliasing and font fallback. A few embedded fonts may shift slightly after compression.
- One PDF at a time. For batch jobs, upload them one by one.
FAQ
Common questions about compression modes, quality, privacy and how this differs from other tools.
What is the difference between Keep text layer and Rasterize pages?
Keep text layer just repacks the PDF internals and strips redundant metadata. Text, links and bookmarks are preserved. Savings are usually 5-20%. Rasterize pages re-renders every page to JPEG and rebuilds the PDF. Savings can reach 50-90%, but text becomes pixels — no more copying or searching.
Which should I try first?
Start with Keep text layer. For text-heavy PDFs, that one step is often enough and preserves all features. If it saves less than 5%, switch to Rasterize.
What exactly are the three presets?
High = 180 DPI with JPEG quality 80, suited to formal sharing or later printing. Standard = 120 DPI with quality 50, fine for on-screen reading. Smallest = 96 DPI with quality 30, prioritizes size; expect visible compression artifacts.
Is text still readable after Smallest?
On screen, yes. Zoom in or print and the quality drops noticeably. Smallest is recommended only for screen-only forwarding. For anything you plan to print, stay at Standard or above.
Can I compress an encrypted PDF?
No. Password-protected PDFs fail at the load step. Remove the password with a dedicated PDF decryption tool first, then come back to compress.
Is my PDF uploaded anywhere?
No. Everything runs locally in your browser. The file never reaches any server. The tool works offline. When you close the tab, the in-memory copy is released.
Does the tool overwrite my original file?
No. The output is a new PDF; the original stays untouched on your disk. Try as many parameter combinations as you like.
Why does my PDF barely shrink?
Two common reasons. The file was already heavily optimized, so Keep text has nothing to remove. Or the file is mostly plain text and was small to begin with. In both cases, Rasterize can still reduce size — at the cost of turning text into pixels.
How large a PDF can I compress?
No hard limit — browser memory decides. Tens of MB usually works fine on desktop. For hundreds of MB or thousands of pages, split the PDF first and compress in batches.
Does it work on mobile?
Yes. iOS Safari and Android Chrome both work. Mobile memory is tighter, though — 50+ MB scans or Rasterize at high DPI can run into trouble.
Can I OCR the rasterized output?
Yes. OCR works on image content regardless of whether the source PDF had a text layer. Lower DPI does mean less accurate recognition, so use High preset if OCR comes next.
Can I compress an already-compressed PDF again?
Technically yes, but not recommended. Rasterize is lossy; repeated passes degrade quality fast without much further size reduction. Always start from the original.
Why does the result look worse after compression?
Rasterize redraws each page and re-encodes with JPEG, which is inherently lossy. Keep text layer does not have this problem but only saves 5-20%. If quality matters, pick High preset or stick to Keep text.
How is this different from SmallPDF, iLovePDF or Adobe Acrobat?
Different processing location: this tool runs locally in the browser, files do not leave the device. Online services upload to their servers and send results back. Feature-wise this tool only does compression, no OCR or PDF-to-Word, but you can chain it with the same site for PDF split, merge, organize, watermark and PDF-to-image.
Related tools
Compression is usually the last step before sending. After this, the PDF is ready for email, cloud storage or upload forms. Need to split, merge, watermark or convert to images instead? They are all in the same category.