Compress PDF
Compress a PDF locally by rendering selected pages to JPEG images. Control DPI and quality, compare sizes and understand the flattening tradeoff.
A smaller copy, with the tradeoff visible.
IN YOUR BROWSER / PRIVATE INPUT100% in your browser, files never leave your device.
Choose your source files and settings. Originals stay unchanged.
How to use the
Compress PDF
- 01
Select the pages and DPI
Choose up to 60 pages for rendering. Lower DPI reduces pixel count, but small print and diagrams may become less clear.
- 02
Choose JPEG quality
Set a quality level appropriate for the content. Keep enough detail for fine text and linework.
- 03
Compare actual bytes
Download only after checking clarity and file size. A vector-text document can grow when converted to page images.
Choose a visual rewrite or a raster copy
Metadata & structure only rewrites the whole document using PDF object streams and removes standard Info metadata and XMP. It preserves visible text and embedded image data; it does not optimize every PDF object or promise savings.
Raster mode renders only the selected pages as JPEG images. Text will no longer be selectable or searchable, and links, forms and annotations are flattened. Unselected pages are excluded. Keep the source and compare actual bytes before accepting either mode.
This compressor rebuilds page appearance as images
Each selected PDF page is rendered at the chosen DPI, encoded as JPEG and placed into a new PDF page. This is a raster-copy workflow, not a general optimizer for every internal PDF object. New page metadata is not copied from the source.
Searchable text, editable form behavior, links, annotations and document structure are flattened or omitted. Check the supported rendered appearance of form fields and annotations. Keep the original if those features or accessibility structures must remain available.
Use a scanned report as a realistic candidate
Scanned report
DPI: 150
JPEG quality: 75
Selection: all pages within the rendering limitA newly rendered PDF copy.
Compare the measured size and fine text.
Savings depend on the original encoding.When another approach is more appropriate
Keep vector text when it matters
For a small text-based PDF, raster conversion may be a poor tradeoff. Retain the original rather than accepting a larger, less accessible copy.
Extract only the necessary pages
Removing an unrelated appendix can save more than reducing the quality of every page.
Proofread tiny text and charts
Fine grid lines and superscripts are good places to spot excessive JPEG artifacts or insufficient DPI.
Compress PDF:
questions, answered.
Does compression keep searchable text?
No. This tool rebuilds pages as JPEG images in a new PDF. The resulting page text is not inherently searchable.
Why did my PDF become larger?
Compact vector text can require more bytes when rendered as a full-page image. Compare the actual result and keep the original when appropriate.
Does it remove the original metadata?
The new raster PDF is created without copying standard document metadata. This is not a comprehensive sanitization of what is visibly rendered on each page.
Can I compress a 200-page file in one render job?
Rendering is limited to 60 selected pages per job. Extract smaller sections when necessary rather than exceeding device-memory limits.
Is there a PDF compression mode that preserves selectable text?
Yes. Metadata & structure only rewrites the entire PDF without rasterizing its pages. Savings may be small or negative because embedded images are not recompressed. It is not secure redaction or exhaustive metadata sanitization.
The calculations and checks above describe this tool's implementation. Use these primary references for the underlying formats and publishing guidance.
pdf-lib document operations ↗ (opens in a new tab)PDF.js rendering ↗ (opens in a new tab)A smaller copy, with the tradeoff visible.
Make a new copy, inspect the output and keep the original until you have verified the result.
Use the Compress PDF↑