Compress PDF — Structural Optimization
Re-save a PDF with object streams and report the real size change. Your files are never uploaded to our servers. ToolkitPoint processes them on your own computer, so confidential documents stay with you.
Processed in your browser. Your files are not uploaded.
How to use Compress PDF — Structural Optimization
- Select the file in "Select PDF to Compress"; the card shows its page count and size.
- Choose a "Compression Preset": Lossless Structural Optimization, Web & Ebook, Screen & Email or Grayscale Document.
- Keep "Strip non-essential document metadata & XML history" ticked to blank Title, Author, Subject and Keywords, or untick it to leave them.
- Keep "Enable PDF 1.5 Object Streams (Cross-reference compression)" ticked for the smallest structure; untick it for older readers.
- Press "Compress PDF Now" and compare the original and new sizes on the Compression Report before pressing "Download Compressed PDF".
Worked example
Take a 2.4 MB report exported from a word processor, with Title and Author set. Choose Screen & Email, leave both checkboxes ticked and press Compress PDF Now. The report card shows the original and new size side by side with a percentage badge such as -12% Saved. If the size does not drop, it says Already Optimal. The reopened file has an empty Title and Author.
When to use this tool
- Squeezing an export from a design or office program under an email attachment limit.
- Removing author and keyword metadata before sending a document to a client.
- Checking whether a scanned PDF can shrink at all, expecting little change.
Important limitations
The Web, Screen and Grayscale presets re-encode embedded JPEG images only; images in other formats, text and vector graphics are left as they are, and a file with no such images may not shrink. Structural optimization alone cannot shrink already-compressed JPEGs. Encrypted or password-protected PDFs must be unlocked before compression.
Frequently asked questions
Do the Web, Screen and Grayscale presets change image quality?
Yes, for embedded JPEG images: Web re-encodes them at 70% quality up to 1600 px, Screen at 50% up to 1000 px, and Grayscale removes their colour. Text, fonts and vector graphics are not touched. Images stored in other formats are left as they are and counted in the report. The option names describe these settings rather than a measured DPI.
Why did my file get no smaller?
A PDF made mostly of scanned pages or JPEG photos is already dominated by image data, which a structural re-save cannot shrink. The badge then reads Already Optimal and metadata is still cleaned.
What does the metadata option actually clear?
Title, Author, Subject and Keywords are set to empty, and Producer and Creator are replaced with the tool's own names. With the extra clean-up menu on its default, page thumbnails and the XMP metadata stream are also removed. Dates are not specifically removed.
Will text stay selectable?
Yes, because pages are not turned into pictures. Fonts, vector lines and text are kept as they are, which is why the savings are modest compared with image-based compressors. For measured results across five document types, see the PDF compression benchmark in the Research section.
Which PDFs are rejected?
A file that needs a password to open fails to load with an error message. Remove the protection with the Unlock PDF tool first, then compress.
Can I aim for a file size limit?
Yes. Enter a target in KB and the tool steps through stronger JPEG settings until the file fits. If even the strongest setting cannot reach it, the report says so and shows the smallest size reached with an images versus fonts versus other breakdown.