Start with the upload rule and the kind of PDF you have
To compress a PDF without uploading it, first check the portal, email, or form limit and then identify the document. A text-based report exported from Word is different from a photographed packet of forms. The right target is not the smallest possible file; it is a file that meets the limit while remaining usable. For example, a school portal may accept a small scanned transcript only if seals and small type remain readable.
Use PDF Compressor when you want the file processed in the browser. Its "Upload PDF" control selects a local file; the compression code reads its bytes locally and creates a downloadable PDF. It does load processing libraries from a CDN, so no document upload does not mean no network requests or reliable offline operation.
The default "Preserve Text (lossless cleanup)" mode saves the existing PDF with object streams and clears title, author, subject, keywords, producer, and creator fields. It does not downsample page images. Text, fonts, and vectors stay in document form, but the result is not a byte-for-byte copy and may save little or become larger. Do not treat this metadata cleanup as a complete privacy scrub.
"Rasterize to JPEG (maximum)" renders each page and places its JPEG into a new PDF. Text becomes pixels, including an existing searchable text layer. The new pages contain images rather than copied links, form fields, or accessibility structure. A visible signature image is not evidence that a digital signature remains valid; keep the original when signatures or document structure matter.
Choose a mode by the job, not by the percentage saved
Try preserve-text first for a report, invoice exported from an application, or a PDF whose links and selectable text matter. Download the result and compare its size. If the reduction is small, that is not necessarily a failure: there may be little redundant data to remove. Do not claim a specific compression percentage before testing your own file.
For a phone scan or an image-heavy packet, rasterizing is an option when image-only pages are acceptable. The current presets set two controls together:
- Low (high quality): JPEG Quality 0.90 and Max Width 2048 px.
- Medium (recommended): JPEG Quality 0.70 and Max Width 1500 px; this is the initial raster setting.
- High (smallest): JPEG Quality 0.50 and Max Width 1024 px.
These quality values are encoder settings, not predicted percentages saved. Start with Medium, then compare the downloaded result. Try Low from the original if small lettering deteriorates; try High only if the file is still too large and the resulting detail remains usable. None of the names promises a particular output size.
Advanced settings allow quality from 0.30 to 0.95 in 0.05 steps, plus width choices of Original size, 2048, 1500, 1024, or 800. Width is a rendering constraint, not a target PDF size or a promise that every page will have that many pixels. Rendering starts at twice the PDF.js scale-1 width and reduces that scale when needed, with a minimum scale of 0.5. Consequently, unusually wide pages can exceed the selected width. "Original size" removes the width constraint but still renders at scale 2 and re-encodes as JPEG; it does not retain the original scan resolution.
A practical workflow for forms, email, and portals
Keep the original before changing anything. Select the PDF, confirm the displayed filename and page count, choose a mode, and click Compress. Compare Original and Compressed sizes, then choose Download Compressed. The download uses the original basename with _compressed.pdf appended. Rename each candidate to record its mode or preset so repeated downloads are easy to distinguish.
For example, if a portal allows 5 MB, do not rely on a rounded display of 5.00 MB. Check the downloaded file's byte size against the portal's stated limit and leave room below it. The tool has no target-size input or automatic retry to reach 5 MB. Change settings and rerun from the original, not a previously rasterized output, to avoid another round of JPEG encoding.
If the document contains several unrelated attachments, reduce obvious large scans before merging. You can use PDF Merge after each file passes its own readability check. If a long document contains pages that are not required, remove them only when the recipient's instructions allow it. Compression cannot decide which pages are legally or administratively required.
Verify the downloaded PDF before you submit it
Open both the original and the downloaded file in a separate viewer before submitting:
- Compare page count, order, orientation, and page edges. Check every page for missing or clipped content.
- Inspect the smallest text, decimal points, stamps, and signatures at 100 percent zoom and at the intended reading size. Print a sample if paper readability is required.
- Search for a known phrase and copy a line in preserve-text output. Test any required links and form fields rather than assuming the mode name covers them.
- Scan required QR codes or barcodes from the output using the intended reader; visual resemblance alone is not a decoding test.
- Compare actual file sizes. If the result is larger or less usable, retain the original or try a different setting. Finally, verify acceptance in the destination form when ready to submit.
Browser-based processing reduces the need to send a file to a remote conversion service, but it is not a substitute for handling confidential documents carefully. The tool cannot guarantee that a portal will accept the result, retain interactive PDF fields, or preserve every accessibility feature. Keep the original until the recipient confirms receipt.
Frequently asked questions
Can I enter a password for an encrypted PDF?
There is no password field. The initial PDF load does not provide a password workflow, so an encrypted file may fail before compression is available. Obtain an authorized unencrypted copy rather than expecting this tool to unlock it.
Is 50 MB a hard file-size limit?
The tool FAQ recommends roughly 50 MB, but the component does not enforce that size cutoff. It reads the file into memory, and rasterizing also creates page canvases and a new PDF. That recommendation is not a success guarantee; page dimensions and available browser memory also matter. If processing fails, try a smaller source or a desktop workflow.
Why do the advanced values change after I select a preset?
Selecting a different preset overwrites both JPEG Quality and Max Width. Manual changes do not rename the selected preset, so inspect the actual advanced values before comparing runs, not just the highlighted level.
Can I compress only the large scanned pages?
The compressor applies one mode to the whole file and has no page-range selector. If your workflow permits it, use PDF Split to separate the scanned section, compress it, and merge the sections back in their original order. Check the reconstructed document again, including any required navigation or interactive features.