Aim for a PDF size ceiling such as 100 KB or 500 KB with bounded local raster compression and honest quality trade-offs.
Runs locallyNo uploadYou control saved copies
Target-maximum compression
The worker tests bounded DPI and JPEG-quality candidates, keeps the page count and checks that the output PDF opens. The original is returned unchanged if it already fits or no smaller copy is produced. Any rasterized copy still needs your visual review, including small text, form values and comments.
● Local raster worker · no upload · cancellable
Compress a PDF toward a maximum KB size
The raster worker tests bounded page resolutions and JPEG qualities, keeping the clearest candidate at or below your requested byte ceiling when safely possible.
For a reliable result
Choose a PDF and enter the destination's maximum KB value.
Select a maximum DPI and confirm the image-only trade-off.
Inspect small text and the reported target status before downloading.
What changes
The original bytes are returned unchanged if they already fit the ceiling or no smaller candidate is produced. Compressed copies rebuild pages as images; selectable text, forms, links, signatures and accessibility structure can be removed.
Check the download
The target is a ceiling, not exact byte equality. If it is not met, a smaller candidate may be offered above the ceiling; otherwise the original is preserved. Review the target status and compare every required page with the original before using a rasterized copy.
Know before processing
Highly detailed or long PDFs may not remain readable at very small targets. Reduce pages or choose a larger ceiling when possible.
Quick answers
Will every PDF reach the target KB size?
No. Detailed or long documents may not fit an extreme ceiling safely. The result reports whether the target was achieved instead of presenting an oversized file as a success.
Why might text stop being selectable?
Compression rebuilds pages as images while testing resolution and quality. This can remove selectable text, links, forms, signatures and accessibility structure.
Reduce a PDF toward a byte ceiling and review the image-only trade-off
This compressor works toward a file-size ceiling. If the original already fits, it returns the original bytes after checking its page count. Otherwise, it renders pages into images and tries bounded resolutions and JPG qualities. If no smaller copy is produced, it preserves the original instead of offering a larger file.
Rendering replaces text, form controls, links and accessibility structure with flat page images. Keep the source and decide whether that trade-off suits the destination. A small readable attachment and an editable or accessible document may require different workflows.
Choose the target with the document's purpose in mind
Choose an approach for your document
Your situation
Recommended approach
The original already meets the upload ceiling
Use the original or verify the pass-through result. Raster compression is unnecessary when the file already meets the requirement.
A strict ceiling matters more than interactivity
Enter the permitted maximum and review the image-only acknowledgement. Check small text and the target status in the actual result.
Small type needs the clearest available attempt
Start with Up to 150 DPI. It is a maximum candidate setting, not a guarantee: lower resolutions may be tried to pursue the size target.
The target cannot fit a long document legibly
Split or extract relevant pages if the recipient permits separate files, or request a larger ceiling. Do not remove required content merely to make a number smaller.
Try a 200 KB ceiling on a deliberately larger sample
The compression-example.pdf sample contains one A4 page with a synthetic raster and a text title. Its source is above 200 KB so this exercise uses raster compression rather than the already-fits path.
Choose compression-example.pdf. Set Maximum output to 200 KB and Maximum raster quality to Up to 150 DPI.
Read and select the acknowledgement about possible loss of selectable text, forms, links and accessibility structure.
Select Compress PDF to target. Read Target status, Raster quality and Output size instead of assuming that a completed run met the ceiling.
Open the downloaded PDF and compare its title and image with the source. Confirm that it still has one page, inspect fine detail and try selecting the original text title.
Expected result. The larger sample takes the raster path, producing a one-page image-based PDF. The result states whether it met 200 KB, which here means 204,800 bytes. The title may remain visible while no longer being selectable text.
Check the downloaded result
Read Target status as well as file size. A smaller candidate or a preserved original can still be above the requested ceiling; Not met is not a successful target match.
Compare every necessary page and important detail, including form values, stamps and fine print. Page-count validation alone cannot establish visual fidelity.
Test text selection, search and interactions that the recipient needs. If those features are required, the rasterized result may be unsuitable even when it meets the size limit.
Compare output bytes with both the ceiling and the original. If rasterization does not produce a smaller file, the download contains the unchanged original and is labelled Download original PDF.
Troubleshooting
The result says Not met
None of the bounded resolution and JPG-quality candidates met the requested ceiling within the tool's constraints.
Try this: A smaller candidate or the unchanged original may be available, but neither is target-compliant. Increase the ceiling if allowed or split the document into files with fewer pages when the destination accepts separate files.
The result says Original PDF preserved
The tested candidates did not reduce the original file size. Efficient digital PDFs can grow when their text and artwork are rebuilt as page images.
Try this: Download the unchanged original if it suits your destination. The requested target was not met; consider a higher permitted ceiling or fewer pages per file rather than accepting a larger rasterized copy.
Search, copy or form fields no longer work
Raster compression replaces the original document objects with page images.
Try this: Return to the original if those features are essential. A later OCR step can add approximate searchable text, but cannot restore the original forms, links or accessibility structure.
The output is blurry or shows JPG artifacts
A tight ceiling can require reduced resolution and lossy encoding. The page-pixel safety cap can also lower effective resolution on large pages.
Try this: Try a larger ceiling or retain the source. Inspect small lettering at its intended display or print size rather than judging only a thumbnail.
The result says original already fit
The input was at or below the requested byte ceiling, so its bytes were returned without rebuilding the pages.
Try this: Check the file size and use that copy if it meets your requirement. This result has not demonstrated raster compression or removal of document features.
Limits and compatibility
Target and quality
The target is a maximum, not exact byte equality. The minimum is 50 KB; other limits depend on the device. The tool does not guarantee readability at arbitrary targets or choose a universally best compression method.
Local processing
Compression happens in the browser without overwriting the source. Password-protected files must be unlocked with a known password first. Cancellation does not create a completed result.
Your files and form values are processed on this device. Downloaded copies remain wherever
you save them. See how to check local processing
and our privacy policy for site data and optional services.
Common questions
Will adding OCR afterward preserve the same small size?
Not necessarily. Adding a text layer changes the file bytes. Check the final size again, and remember that compressing the OCR output through the raster path removes that text layer again.
Is this a lossless PDF optimizer?
No. Already-fitting files and attempts that produce no smaller file return the original, but compressed copies rebuild pages as lossy images. Use a different workflow if original document objects must be retained.