PDF![]()
Local PDF processing with limitations stated
PDF Compressor
Shrink a scanned or photo-heavy PDF in your browser — pages are rasterized to JPEG, so text stops being selectable.
- Add your PDF
- Choose compression
- Download
Choose compression level
Pick a level, then compress.
Important: every page becomes a JPEG image. Selectable text, links, forms, bookmarks, and annotations are not retained as interactive PDF features.
Compression complete
Your PDF is ready
The finished file stays available in this tab until you compress another PDF or leave the page.
Counted with the browser's own PerformanceObserver from the moment this page finished loading, analytics included. It reports what it observes rather than proving a negative: resource timing does not cover WebSocket connections, work a browser extension does on your behalf, or requests made before load. The network panel does.
Check it yourself in DevTools — the method, in five steps.
How to Compress a PDF File Online — 4 Easy Steps
- Upload PDFClick or drag one PDF file into the box. Up to 100 MiB and 1,000 pages are supported.
- Choose LevelPick Low, Recommended, or High depending on whether you want the best quality or the smallest file. Each level sets a DPI and JPEG quality.
- CompressPages are rendered one at a time in your browser. The result is kept only when it is actually smaller than the original.
- Set Filename and DownloadType your own file name before saving, or keep the suggested one. No watermark is added.
Understand the output
What This Tool Produces
- Original and output file sizes
- Measured reduction only when the result is smaller
- Page dimensions retained in PDF points
- A new image-based PDF
Controls and Method
Compression level
Each level sets a DPI and a JPEG quality together. The named levels describe the trade-off being made, not a guaranteed size reduction.
DPI
Controls raster pixel dimensions. Doubling linear resolution creates roughly four times as many pixels before JPEG compression.
JPEG quality
A codec control, not a percentage of retained quality. 75 does not mean “75% identical.”
Size comparison
Reduction is calculated from actual input and output bytes. Some already-efficient PDFs cannot be made smaller by this method.
What Compression Changes
Every page becomes a full-page JPEG. Selectable text, vector paths, live links, form fields, comments, bookmarks, and searchable text are not retained as interactive PDF objects. Their visible appearance may be present in the rendered page image.
- Text may look less sharp, especially at 72 DPI or low quality.
- Scanned pages and photographs are the strongest candidates for meaningful reduction.
- Mostly text/vector PDFs or files with already-compressed images may shrink little or even grow.
- Mixed page dimensions are retained; the tool does not force A4 or Letter.
Choosing a Level
| Level | Renders at | Likely use | Trade-off |
|---|---|---|---|
| Low | 150 DPI · quality 90 | Documents where image detail matters | Largest output and heaviest memory use |
| Recommended | 96 DPI · quality 75 | General screen sharing | Balanced detail and size |
| High | 72 DPI · quality 50 | On-screen preview or a strict upload cap | Smallest raster dimensions; fine text may soften |
When Not to Use This Compressor
Keep the original when you need selectable or searchable text, accessibility tagging, exact vector quality, live forms, bookmarks, annotations, archival properties, or a valid digital signature. A structural PDF optimizer would be a different tool.
Rasterizing is often the wrong first move. If only a few pages matter, splitting out the pages you need drops the file size without touching quality at all, because pages you removed cost nothing. If the weight comes from photographs you still control, compressing those images first and then rebuilding the PDF from them keeps any real text in the document selectable, which this compressor cannot do once every page is a picture. And when the PDF was exported from a Word file, converting the original DOCX again usually beats compressing the bloated export.
Practical Uses
- Reduce a photo-heavy scan for an upload limit.
- Create a lighter visual copy of receipts for email.
- Downsample a portfolio preview where selectable text is not required.
- Make a sharing copy while retaining the original master.
Important Limitations
- The method is lossy and rasterizes every page.
- The processed result is rejected when it is not smaller than the original.
- Output size depends on page count, dimensions, detail, source compression, DPI, and JPEG quality.
- A signed PDF’s signature will not remain valid in the rebuilt image PDF.
- Large or high-DPI documents can exceed a mobile browser’s memory even when they pass the file-size limit.
Frequently Asked Questions
Why did my PDF barely shrink?
It may already use efficient compression, contain mostly text/vector data, or be rendered at settings that produce more image data than the source.
Does compression reduce image quality?
Yes. Pages are JPEG-encoded, so lower DPI or quality can remove detail and introduce artifacts.
Will text remain selectable?
No. This implementation rasterizes pages; text becomes part of a page image.
Why is no download offered for some files?
If the processed PDF is not smaller, the tool reports that result instead of presenting a larger file as a successful compression.
Is a quality value of 75 equal to 75% quality?
No. JPEG quality scales are encoder controls and do not represent a fixed percentage of visual fidelity.