A 15 MB scan lands in your inbox and you need to forward it — but the attachment bounces because of a size limit. Desktop PDF software and upload-based online compressors are the usual fixes, but both have trade-offs. There is a third option: compress the PDF entirely in your browser.
Why PDF file sizes grow
PDFs are containers. A single file can hold vector text, embedded fonts, high-resolution photos, and scanned page images. Size spikes usually come from:
- Scanned pages stored as full-page bitmaps instead of text
- High-resolution photos embedded without recompression
- Duplicate or unused objects left over from editing or export
- Unoptimized structure where streams are not packed efficiently
Text-only PDFs are often small. Scan-heavy PDFs behave more like a stack of images — and images dominate file size.
Browser-based compression without uploads
Modern browser tools can read a PDF into memory, process it locally, and give you a smaller download — no account, no server upload, no waiting in a queue.
That matters when the document is sensitive: contracts, medical records, financial statements, or internal reports. Keeping processing on your device removes a whole class of privacy risk.
Three compression strategies
Not every PDF needs the same treatment. A practical tool offers presets tuned to different goals:
Light — structure cleanup
Light rebuilds the PDF without turning pages into images. It is the best first choice for text-heavy documents: reports, invoices, forms, and exported Word or Google Docs PDFs.
You may see modest size reduction — often from cleaner object streams — without changing how text renders. Text stays sharp because it is not rasterized.
Balanced — re-render with moderate JPEG quality
Balanced renders each page through PDF.js, converts the result to JPEG, and embeds it in a new PDF. This is the default for mixed documents and scans where you want noticeable size savings while keeping pages readable.
Expect longer processing on multi-page files. Each page is rendered individually, so a 40-page scan takes longer than a 4-page letter.
Maximum — smallest file, softest detail
Maximum uses stronger JPEG compression. File size drops further, but fine detail in photos and scans may soften. Use it when the priority is fitting an email limit or attachment cap, not preserving every pixel.
Avoid Maximum on PDFs where crisp text in diagrams or small type is critical — rasterization can make tiny labels harder to read.
Step-by-step: compress a PDF in the browser
- Open PDF Tools and switch to the Compress tab
- Drop your PDF or click to browse
- Pick Light, Balanced, or Maximum
- Wait for page-by-page progress on large files
- Download the compressed PDF and compare file size
Tip: try Light first on text documents. If the result is still too large, run Balanced. Reserve Maximum for scans and photo-heavy files where size matters most.
How it works under the hood
Browser PDF compression typically combines two libraries:
- PDF.js (Mozilla) renders each page to a canvas when raster presets are used
- pdf-lib builds a new PDF and embeds JPEG images or copies pages for Light mode
The pipeline runs entirely in your tab:
- Load — the file is read as an ArrayBuffer in memory
- Process — pages are copied or re-rendered according to the preset
- Save — a new PDF binary is generated with object streams enabled
- Download — you get the result as a local file
No network round-trip. No copy on someone else’s server.
When browser compression is enough
| Situation | Recommended approach |
|---|---|
| Email attachment over the size limit | Browser tool — Balanced or Maximum |
| Text report or invoice | Browser tool — Light first |
| One-off scan before sharing | Browser tool — Balanced |
| Batch compressing hundreds of files | Command-line tool (Ghostscript, qpdf) |
| OCR, forms, redaction, or true text editing | Desktop PDF editor |
| Password-protected PDF | Unlock or decrypt first — most browser tools cannot process encrypted files without the password |
Common mistakes to avoid
- Compressing already-compressed scans twice — running Maximum on a file that was already heavily compressed may barely shrink it while adding artifacts
- Using Maximum on slide decks with small text — rasterization can blur labels and footnotes
- Expecting Light to halve a photo album PDF — Light does not recompress embedded images aggressively; use Balanced or Maximum instead
- Forgetting page count — 100-page scans take time locally too; that is normal, not a bug
Related workflows in the same tool
PDF Tools on ToolSnap also covers merge, split, extract, PNG export, and visual compose — all with the same local-first privacy model. If you need page images instead of a smaller PDF, use the PNG tab to export selected pages as images or a ZIP archive.
For most everyday tasks — shrinking a scan before email, reducing a report attachment, or preparing a document for upload — browser-based compression is fast, private, and good enough.