PDF Compressor

Disabled {{ engineBadge }} {{ profileLabel }} {{ sourceSizeLabel }}

{{ rewriteDownloaded ? 'Downloaded' : 'Download local rewrite' }}

Source PDF:

Drop or browse one unencrypted PDF up to the browser work limit.

Compression profile:

Choose the size-quality target you would use in a production PDF compressor.

Target size:

Enter 0 for no fixed limit, or set an email/upload cap such as 10 MB.

Metadata cleanup:

Keep on for privacy-oriented local rewrites; turn off if document info fields must remain intact.

Output filename:

Example: board-pack-small.pdf. The .pdf extension is added if missing.

Browser work limit:

Raise only for a powerful desktop browser; production compression should use a backend worker.

Metric	Value	Detail	Copy
{{ row.metric }}	{{ row.value }}	{{ row.detail }}

Check	Status	Detail	Copy
{{ row.check }}	{{ row.status }}	{{ row.detail }}

Export to PDF Fullscreen

Embed:

Customize

Include current inputs

Size

Advanced

Width

Height

Aspect ratio

Max height

Collapsible embed

Allow fullscreen

Referrer policy

Sandbox tokens

PDF compression reduces a document's byte size while trying to keep the pages readable, printable, and structurally valid. The result depends on what makes the file heavy. A text-heavy report with many small objects behaves differently from a scanned packet, a slide deck full of JPEG images, or a form that has been saved many times.

The strongest size reductions usually come from image downsampling, image recompression, removing unneeded document data, and rewriting PDF structure more compactly. Those operations do not carry the same risk. Lowering image quality can soften photos or scans, while structural cleanup can reduce overhead without changing the visible page as much.

Exact file-size targets need caution. PDF optimizers often try several image and stream settings before a file lands under an email, upload, or archive limit. A smaller file can still be the wrong result if scanned text becomes hard to read, transparency is damaged, metadata must be retained, or an encrypted file needs password-aware handling.

Technical Details:

A PDF is a structured container of indirect objects, streams, page resources, images, metadata, cross-reference data, and a trailer. File size falls when redundant structure is rewritten, unused document information is removed, images are recompressed, or high-resolution images are downsampled to a resolution that still fits the intended viewing or printing use.

Object streams and cross-reference streams are PDF 1.5-era structures that can represent indirect objects and lookup data more compactly than older body and cross-reference sections. They are useful for structural savings, but they do not recompress the image pixels inside page resources. Image-heavy PDFs usually need a production optimizer that can decode, resample, recompress, and validate those embedded images.

Compression should be measured as a byte ratio, then checked against visual quality and document risk. A rewrite that is technically valid can still be larger than the original, especially when the source was already compact or when rewriting removes incremental-save overhead less efficiently than expected.

Formula Core:

The core size calculation compares the source bytes with the candidate output bytes. Positive savings mean the candidate is smaller; zero or negative savings mean the original is the better size choice unless metadata cleanup is more important than byte count.

Savings percent = (1 - \frac{B_{output}}{B_{source}}) \times 100

If a 20.00 MB source becomes a 17.40 MB local rewrite, the savings are 13.0 percent. If the same source is estimated at 9.80 MB after production image recompression, the estimated savings are 51.0 percent. The second result is a planning estimate until a production compressor creates and verifies the final PDF.

PDF compression signals and limits
Signal	Meaning	Compression implication
Object streams or xref streams	The PDF already uses compact structural storage for some indirect objects or lookup data.	Local structural savings may be limited because part of the compact rewrite has already happened.
Image filter markers such as DCT, JPX, CCITT, or JBIG2	The file contains image data that may dominate size, especially in scans and slide exports.	Real size reduction usually needs image recompression or downsampling outside a structural rewrite.
XMP or document information fields	The file may carry descriptive metadata such as title, author, creator, producer, or XMP metadata streams.	Neutral document information fields can reduce some exposure, but embedded metadata streams need deeper cleanup.
Encryption marker	The PDF advertises an encryption dictionary or password-aware handling requirement.	Compression should stop until a backend can handle passwords and encrypted objects correctly.
Missing header or uncertain EOF marker	The file may be corrupt, incomplete, or not a PDF despite its extension.	Repair and validation should come before compression.

Target-size estimates are a planning aid rather than a guarantee. The same target may be easy for a photo-heavy handout and unrealistic for a vector drawing, tagged form, or already-optimized report. The safest comparison keeps the source file, profile, target limit, and visual review path stable across repeated runs.

Everyday Use & Decision Guide:

Start with a single unencrypted PDF and the Balanced sharing profile. That profile is the sensible first pass for ordinary handouts because it estimates moderate image recompression and keeps readability ahead of the smallest possible byte count. Use Smallest readable for strict email or upload caps, and use Print-safe quality when printed diagrams, photos, or scans matter more than aggressive reduction.

Set Target size only when there is a real cap, such as 10 MB. The estimate can say whether the selected production profile appears under that cap, but the final answer still needs a real optimizer and a visual check. Enter 0 when there is no fixed limit so the result does not imply a target that nobody needs.

Keep Metadata cleanup on when a local rewrite should replace common document information fields with neutral export values.
Turn Metadata cleanup off when title, author, creator, or producer fields must be preserved for document control.
Raise Browser work limit only on a capable desktop browser. Large PDFs can duplicate memory while being parsed and saved.
Check Compression Metrics for source size, profile, local rewrite size, production estimate, target limit, and enablement blocker.
Check Optimization Checks before trusting the result. Encryption, image-heavy sources, sampled scans, and browser-limit blocks should slow the handoff.

The selected PDF is read in the browser. The local rewrite, when available, is also created in the browser as a downloadable file. That local path can test structural savings and neutral document information fields, but it does not downsample or recompress embedded images.

A common misread is treating the production estimate as a finished compressed PDF. It is a readiness signal. If the file is image-heavy or the target is tight, use the estimate to choose a production profile, then verify the actual compressed file's size and page appearance before replacing the source.

Step-by-Step Guide:

Use the page as a short compression review before a production handoff.

Drop or browse one PDF in Source PDF. If the file is too large for the current Browser work limit, use a backend optimizer instead of forcing the browser to process it.
Choose a Compression profile. Use Balanced sharing first unless the destination is screen-only or print quality is the main concern.
Enter a Target size in MB when an attachment, portal, or storage rule requires one. Leave it at 0 for open-ended reduction.
Decide whether Metadata cleanup should stay on. Keep it on for a privacy-oriented local rewrite; turn it off when document information fields are part of the record.
Review the warning messages. Stop on encrypted PDFs, image-heavy PDFs that need backend recompression, and scans that were sampled to protect browser memory.
Use Create local rewrite only after analysis is ready. Download the local rewrite when it is smaller or when neutral document information fields are more important than byte count.
Open PDF Size Profile to compare the source, local rewrite or local estimate, backend profile estimate, and target limit when a target was entered.
Use JSON or the tables when another person needs the selected profile, estimates, warnings, and blocker recorded for follow-up work.

The completion signal for this page is a clear review state, not a production-ready compressor. A finished run should leave you with the original file, any local rewrite worth keeping, and a documented backend compression plan when image recompression is required.

Interpreting Results:

Compression Plan means the source was readable enough to estimate profile savings. Read it with the disabled badge in mind: the page can analyze and attempt a local structural rewrite, while production image recompression is still outside the enabled surface.

PDF Compressor result statuses
Result cue	How to read it	What to do next
Local Rewrite Ready	The browser created a rewritten PDF and it is smaller than the source.	Open the downloaded file and check page appearance before using it.
Local Rewrite Larger	The rewrite created a file that is not smaller than the original.	Keep the original unless metadata cleanup is the stronger requirement.
Image compression: Review	Image markers were found, such as JPEG, JPEG2000, CCITT, JBIG2, or image subtype signals.	Use a backend optimizer that can recompress or downsample images and then inspect the pages.
Encryption: Blocked	An encryption marker was found or the PDF needs password-aware handling.	Use a password-capable backend path; do not rely on browser rewriting.
Production readiness: Blocked	The current page is not the final production compressor.	Use the profile estimate and warnings as handoff evidence for backend compression work.

Production Estimate is most useful for comparing profiles, not for declaring success. A target that appears reachable should still be verified against the actual optimized PDF. A target that appears missed usually means the source needs stronger settings, image edits, page removal, or a different source export.

PDF Size Profile turns the same comparison into bars. The source bar should be the baseline, the local bar should reflect the local rewrite or local estimate, the backend profile bar should show the planned production result, and the target bar appears only when a fixed MB limit was entered.

Worked Examples:

Scanned packet with an email cap

A 22 MB scanned packet must fit under 10 MB. Smallest readable may estimate a backend profile below the target because image downsampling drives the savings. If Optimization Checks reports image markers, the next step is production recompression followed by a page-by-page readability check of small text, signatures, and stamps.

Text-heavy report already using object streams

A 4.8 MB report shows object streams and xref streams in the scan. The local rewrite may save very little because the source already uses compact structural storage. Print-safe quality may show a modest production estimate, but replacing the source is not worthwhile unless the actual output is smaller and looks identical.

Metadata-sensitive board pack

A board pack needs neutral document information fields before sharing. Keep Metadata cleanup on and create a local rewrite. If the rewrite is slightly larger, the decision becomes a privacy and record-control question rather than a size win. Open the file and confirm the pages still render correctly before handoff.

Encrypted client PDF

An encrypted client file triggers a blocked encryption check. The correct path is to use a password-aware backend optimizer or obtain an unencrypted source from the document owner. A browser structural rewrite should not be used as a workaround for encrypted objects.

FAQ:

Does the selected PDF leave my browser?

The selected file is read in the browser for analysis and local rewrite attempts. The page does not send the PDF to a compression service from this workflow.

Why is the page marked disabled?

The current surface can analyze a PDF and attempt a local structural rewrite. It remains disabled because production-grade compression needs backend image recompression, stream recompression, repair handling, and encrypted-file handling.

Why did the local rewrite get larger?

Some source PDFs are already compact. Rewriting object streams and document information fields can add overhead, so the original can remain the smaller file.

Can the target size be guaranteed?

No. The target check is an estimate until a production optimizer creates the output. Exact caps often need iterative image settings and a final byte-size check.

Which profile should I start with?

Use Balanced sharing for general handoff, Smallest readable for strict screen or attachment caps, and Print-safe quality when print readability matters most.

Glossary:

Object stream: A compact PDF structure that can store certain indirect objects inside a stream.
Cross-reference stream: A compact lookup structure that helps readers find objects in newer PDF files.
Image downsampling: Reducing image resolution so scans or photos use fewer pixels and fewer bytes.
Image recompression: Encoding embedded images again with a chosen quality or compression method.
XMP metadata: Structured document metadata that can be embedded in a PDF separately from common document information fields.
Linearized PDF: A PDF arranged for faster first-page display while the rest of the file continues loading.

References:

Reduce PDF file size with advanced options, Adobe Acrobat Help, November 5, 2025.
Parsing PDF's dialects, PDF Association.
Glossary of PDF terms, PDF Association.
Object and Cross-Reference Streams, QPDF documentation.