By Usama Farooq · Published 5 Apr 2026 · Updated 16 Nov 2026
Best Ways to Compress PDF Without Losing Quality
PDFs are great for sharing documents consistently, but they can get huge—especially when they include scans, screenshots, or high-resolution images. Large PDFs are painful to email, slow to download on mobile, and often rejected by upload portals that enforce size limits. The good news is you can usually reduce PDF size while keeping excellent readability.
This guide covers the best ways to compress PDFs without losing quality, including practical “no regret” optimizations, what to do for scanned PDFs, how to choose a target size, and how to avoid common mistakes. We also link to free browser-based tools so you can do everything quickly without uploading sensitive documents to a third party.
TL;DR
- Start with lossless optimization — it costs nothing and often trims 5–20%.
- Scanned PDFs are big because of images; reducing resolution to 100–150 DPI is the highest-impact fix.
- Splitting a document changes zero quality and solves most portal size limits instantly.
- Only recompress aggressively when you know the document will be read on screen, never printed.
- Always verify at 100% zoom and keep the original file as a backup.
Why PDFs get large: the three size drivers
Before compressing, it helps to know what you are actually shrinking. PDF size almost always comes from one of three drivers, and each one responds to a different technique.
1. Images — the dominant driver
Scans, photos, and screenshots are by far the most common reason a PDF balloons. A scanned page is a photograph of the page, not text: a color scan of one letter-size page at 300 DPI contains roughly 25 million pixels, and if it is stored uncompressed or weakly compressed it can weigh 10–25 MB by itself. A ten-page scan can easily exceed 100 MB, while the same ten pages as typed text would be under 100 KB. Presentation decks with full-bleed photos and phone screenshots have the same problem — high pixel counts saved at full fidelity.
Resolution (DPI) and compression format are the two dials here. Doubling the DPI quadruples the pixel count, so a scanner set to 600 DPI produces files roughly four times heavier than at 300 DPI, with no visible gain on screen. Format matters too: a photographic scan stored as PNG can be several times larger than the same scan stored as JPEG at a reasonable quality setting, because PNG is lossless and JPEG is designed for photos.
2. Embedded fonts — the quiet multiplier
PDFs embed fonts so the document looks identical on any device, which is the right default — but it adds weight. A full font file can be 100–500 KB per style, and documents assembled from multiple sources sometimes carry several full fonts, or even duplicate copies of the same font from different sections. Proper “subsetting” (embedding only the characters the document actually uses) fixes this without changing appearance, but not every PDF producer does it well.
3. Embedded objects and structure — the hidden overhead
Form fields, annotations, comments, hidden layers, OCR text layers, and leftover objects from editing all add weight. Some PDF producers also write inefficient file structure — repeated resources, unoptimized streams — that can be rebuilt more compactly. This overhead is usually smaller than image weight, but it is the first thing a lossless optimizer removes, which is why the same PDF can shrink noticeably just from a clean re-save.
What each compression method sacrifices
“Without losing quality” means different things depending on the method. Here is an honest breakdown so you can pick the right one for your document:
| Method | What it changes | What you lose |
|---|---|---|
| Lossless re-save | Rebuilds file structure, removes redundant objects | Nothing visible |
| Image recompression (JPEG) | Re-encodes images at lower quality | Fine detail, smooth gradients can band |
| Downsampling (lower DPI) | Fewer pixels per image | Sharpness at high zoom and in print |
| Font subsetting | Embeds only used characters | Nothing, unless someone edits with new text |
| Removing annotations/layers | Strips comments, hidden layers, OCR text | Searchability, comments, editability |
| Splitting | Divides into multiple files | Nothing — but you manage several files |
The key insight: downsampling is the method most likely to hurt readability, while lossless re-save and splitting never touch content. Aggressive JPEG recompression sits in between — it usually looks fine at 100% zoom but degrades when someone zooms into fine print or prints the page.
How to choose quality settings
The right settings depend entirely on what happens to the PDF after you send it. A contract that will be printed needs far more care than slides viewed on a phone.
| Document purpose | Scan resolution | JPEG quality | Safe compression style |
|---|---|---|---|
| Screen reading / email | 100–150 DPI | 60–70% | Moderate lossy + lossless |
| Upload portals (forms, applications) | 150 DPI | 65–75% | Moderate lossy, verify text crispness |
| May be printed | 200–300 DPI | 80–90% | Lossless first; light lossy only |
| Archive / legal | 300 DPI | Highest / lossless | Lossless only |
A practical rule of thumb: if text remains crisp at 100% zoom, the settings are fine for screen use. If letters blur, raise the DPI or quality before sending — a 2 MB file that is unreadable is worse than a 6 MB file that is legible.
Method 1: Lossless PDF optimization (best first step)
Start with a lossless “optimize/re-save” operation. This rewrites the PDF in a more efficient structure without changing the visible content. It is fast and safe, and it is a good default for documents that are already readable and should not degrade. Realistic savings range from 5% to 20% — occasionally more for PDFs produced by inefficient generators.
Try ToolsOfWeb's browser-based PDF Compress tool. It runs client-side (no upload required) and shows before/after size so you can keep the better version. Results vary — some PDFs shrink significantly, others stay similar, and a few can grow slightly depending on the original producer.
How to do it, step by step
- Save a copy of the original first. Even lossless processing can surprise you on complex files, and you always want a fallback.
- Open the tool and drop in your PDF. Because it runs in your browser, the file never leaves your device — important for contracts, invoices, and ID scans.
- Run the optimization and compare sizes. Check the before/after numbers the tool reports. If the saving is tiny (under a few percent), the file was already efficient — move on to the image-focused methods below.
- Spot-check pages at 100% zoom. Even though the operation is lossless, verify that forms, images, and fonts render identically before you send it.
Method 2: Reduce image weight (best for scanned PDFs)
If your PDF is mostly scanned pages, the right compression strategy is almost always image optimization. Scanners and phone apps can embed very large images per page. The trick is to keep pages readable while lowering image resolution and using efficient formats.
Pick the right resolution
Resolution is the single biggest lever. A scanner default of 600 DPI produces images four times heavier than 300 DPI, and for on-screen reading even 150 DPI is usually plenty. If your scan was captured at 600 DPI and you will only ever view it on screen, downsampling to 150 DPI cuts image data by roughly 16x with almost no visible difference in a PDF viewer.
Pick the right format
- Photographic scans (color, grayscale): JPEG is the efficient choice. Quality around 65–75% looks clean on screen at a fraction of the size.
- Sharp diagrams and text-heavy graphics: PNG keeps edges pixel-perfect, which matters for small text in screenshots. Use it only where JPEG artifacts would blur edges.
- Black-and-white document scans: If the page is purely text, scanning or converting to true black-and-white (1-bit) instead of grayscale can shrink each page to tens of kilobytes.
Compress before the PDF is built
If the PDF was created from images, the highest-impact optimization happens before the PDF exists at all: compress the original images first, then re-export the PDF. ToolsOfWeb includes an Image Compressor and Image Resize tool for exactly this. Re-exporting from a photo library at 150 DPI JPEGs instead of full-resolution originals can turn a 90 MB scan bundle into an 8 MB PDF that looks identical on screen.
Method 3: Split large PDFs into smaller files
Sometimes the best “compression” is simply splitting. If an upload portal limits file size, you can split a 40-page scan into 2–4 smaller PDFs without changing quality at all. This also makes documents easier to navigate and faster to open on low-end devices.
Splitting is the only method on this list with literally zero quality cost — it just divides pages between files. It is also the fastest fix when you are up against a deadline: a 22 MB report on a 10 MB-limit portal becomes two ~11 MB halves in under a minute, with no settings to tune and nothing to verify visually.
Use PDF Split to extract a page range into a smaller PDF, or download separate pages when needed. If you later need to re-combine documents, use PDF Merge.
Method 4: Remove unnecessary content (only when safe)
Some PDFs include extra content that increases size without improving the final result: hidden layers, unused objects, or embedded resources. A clean re-save can help, but you should be careful with documents that include form fields, annotations, or digital signatures.
The OCR text layer deserves special mention. Scanned PDFs often carry an invisible text layer (so you can search and copy text), and it adds modest weight. Removing it saves a little space but kills searchability — almost never worth it for a document people will read seriously. Likewise, flattening form fields or comments can break workflows downstream. When in doubt, leave the content alone and lean on the image methods instead.
If the PDF includes a signature or must remain legally intact, avoid aggressive processing. Prefer the lossless optimization approach and verify the output carefully.
Before-and-after scenarios
To calibrate expectations, here are realistic outcomes for common document types. Your numbers will vary with the original producer, but these ranges reflect what typically happens:
| Document | Before | After | What was done |
|---|---|---|---|
| 20-page contract scanned at 600 DPI | 96 MB | 6–9 MB | Downsampled to 150 DPI, JPEG ~70% |
| Slide deck with full-bleed photos | 38 MB | 9–12 MB | Lossless re-save + image recompression at 75% |
| Text-only report from Word | 2.4 MB | 1.9–2.2 MB | Lossless re-save (font subsetting) |
| Receipt scans for expenses (phone app) | 14 MB | 3–4 MB | Converted grayscale, JPEG 65%, kept OCR layer |
| Job application packet (portal limit 10 MB) | 22 MB | 2 × 11 MB → 2 × ~7 MB | Split in half, then light image recompression |
Notice the pattern: the more image data a PDF carries, the more dramatic the savings. Text PDFs were never big, so do not expect miracles there — a 20% cut on a 2 MB file saves 400 KB, which matters only if you are counting every byte.
Decision guide: which method first?
Work through this list in order. Stop as soon as the file is small enough — every extra method risks quality you did not need to sacrifice.
- Always start with lossless optimization. It is free, fast, and never degrades anything.
- If a portal limit is the problem, split. Two files at full quality beat one degraded file.
- If the file is scan-heavy, optimize images. This is where 50–90% reductions come from.
- If you still have the source files, rebuild.Re-exporting at sensible resolution beats patching a bloated PDF.
- Only strip content (layers, OCR, annotations) when you are certain nobody needs it downstream.
How to choose a “good” compressed size
There is no universal target, but here are common goals:
- Email-friendly: under 10–20MB (varies by provider).
- Upload portals: often 2–5MB or 10MB limits.
- Fast mobile viewing: smaller is better, but keep text readable.
After compression, always check a few pages at 100% zoom. If the PDF contains scanned text, make sure letters remain crisp and not overly blurred. For important documents, verify the first page, a middle page, and the last page before sending or submitting.
Common mistakes to avoid
- Over-compressing scans until text becomes blurry. Small file, unreadable content — the classic failure. JPEG quality below ~60% or DPI below ~100 on typed text starts to smear letter edges.
- Compressing the same file multiple times. Each lossy pass degrades quality further while saving less. Always compress from the highest-quality original, never from an already-compressed copy.
- Compressing signed or legal PDFs carelessly. Verify signatures, form fields, and annotations after any processing — or send the original.
- Forgetting to keep the original file. Storage is cheap; regret is not. Archive the original before any lossy step.
- Optimizing the wrong driver. Running a lossless optimizer five times on a scan-heavy PDF will not help; the images are the problem. Diagnose first (images vs. fonts vs. structure), then treat the right one.
- Checking quality only as thumbnails. A degraded PDF can look fine in a small preview and fall apart at 100% zoom. Always check at full size.
Quick workflow (recommended)
- Back up the original file, then run a lossless optimize using PDF Compress.
- If you still need a smaller file, split the PDF into ranges using PDF Split. This is the safest way past a portal limit.
- If the PDF is scan-based and extremely large, downsample the images to 100–150 DPI at ~65–70% JPEG quality, or compress/resize the original images with Image Compressor and re-export the PDF.
- Verify readability at 100% zoom — first page, a middle page, and the last page — before uploading or emailing.
Conclusion
The best PDF compression strategy depends on what is inside the file. Start with lossless optimization, then split large documents when portals have size limits, and focus on image optimization for scanned PDFs. With a privacy-first, browser-based workflow, you can reduce file size quickly without uploading sensitive documents to unknown servers.
Related guide: How to Split PDF Online Free (Step-by-Step Guide).
FAQs
Yes, with lossless optimization: the PDF is rewritten in a more efficient structure without changing any visible content. Typical savings are modest (often 5–20%), but it costs you nothing in quality. Bigger savings almost always involve some image quality trade-off.
A scanned page is a photograph of the page, not text. A single color scan at 300 DPI can contain 25 million pixels; a text page is a few kilobytes of characters. That is why image-heavy PDFs shrink dramatically when you reduce resolution or use efficient compression.
Around 150 DPI is usually plenty for on-screen reading, and even 100 DPI stays legible for typed text. 300 DPI is the standard when the document may be printed. Anything above 300 DPI in a shared PDF is usually wasted file size.
Be cautious. Any processing that rewrites the PDF can invalidate a digital signature, and some methods flatten form fields or annotations. If the document is legally binding, prefer the lossless approach and verify the signature afterward — or send the original as-is.
Because tools target different parts of the file: some only rebuild the PDF structure, others recompress images or downsample them. The original producer matters too — a PDF built inefficiently shrinks a lot, while an already-optimized one barely changes.
If you still have the source files (Word document, original photos), rebuilding is usually better: export with sensible image resolution and let the producer optimize the structure. Compressing an existing PDF is the practical option when the source is gone.
Compress your PDF now
Run a free, private, lossless optimization with PDF Compress — the file never leaves your browser. If it is still too large, split it with PDF Split or shrink its images with Image Compressor.
