Lossy vs lossless compression, which image format to use, how far you can safely compress, and what actually makes PDFs bloated.
Compression reduces file size by removing data. Lossy compression discards image information a viewer is unlikely to notice; lossless compression reorganizes data so nothing is lost but the file still shrinks. Knowing which one you're applying and how far to push it is the difference between a fast-loading page and a gallery of blurry artifacts. Image Compressor → | PDF Compressor → Why File Size Matters More Than People Think Images are typically the largest thing on a web page — often more bytes than the HTML, CSS, and JavaScript combined. That has direct consequences: Load time. Page speed affects both rankings and how many visitors stay long enough to read anything. Mobile data. A visitor on a metered connection pays for every unoptimized megabyte. Email limits. Most providers cap attachments around 20–25 MB, which a handful of phone photos exceeds easily. Upload limits. Job portals, government forms, and CMS uploads frequently cap files at a few megabytes. Choosing the Right Image Format Format Compression Best for Avoid for JPEG Lossy Photographs, complex images with gradients Text, sharp lines, transparency PNG Lossless Screenshots, logos, anything with transparency or text Large photographs WebP Both Almost everything on the web typically 25–35% smaller than JPEG at equal quality Contexts requiring very old browser support SVG Vector Logos, icons, illustrations scales to any size without loss Photographs AVIF Both Smallest files of the modern formats Where broad tool support matters The most common mistake: saving photographs as PNG. A PNG photo can be several times larger than the equivalent JPEG with no visible advantage, because lossless compression works poorly on the complex, noisy data that makes up a photograph. The reverse mistake: saving screenshots as JPEG. Lossy compression on sharp text edges produces visible fringing artifacts. Screenshots belong in PNG. Converting between formats is straightforward with an image converter . How Far to Compress For JPEG and WebP, quality settings behave predictably: 90–100 - near-original quality, minimal size reduction. Rarely worth it for web use. 75–85 - the practical sweet spot. Substantial size reduction with no visible difference in normal viewing. 60–75 - visible on close inspection, acceptable for thumbnails and background images. Below 60 - visible artifacts, blocking, and colour banding. Start at 80 and inspect the result at full size. If you can't see a difference, try lower. Compare details specifically faces, fine text, smooth gradients like skies, and areas of solid colour, where banding appears first. Resize Before You Compress This is the step people skip, and it matters more than the quality slider. A phone camera produces images around 4000 pixels wide. A blog post displays them at maybe 800 pixels. Compressing the 4000px file still leaves you serving four times more pixels than anyone sees. Resize to display dimensions first, then compress. A image resizer handles this in one step. Rough targets for web use: Full-width hero image: 1920px wide In-article image: 800–1200px wide Thumbnail: 300–400px wide Resizing first often cuts file size by 80% or more before any compression is applied. Compressing PDFs PDF bloat almost always comes from one source: embedded images at full camera resolution. A scanned document or a slide deck full of screenshots carries every one of those images at whatever resolution it was inserted at. A PDF compressor reduces size by downsampling embedded images and re-encoding them. What that means in practice: Text-only PDFs barely shrink text is already tiny in file terms, so there's little to remove Image-heavy PDFs shrink dramatically often 70–90% Scanned documents shrink most every page is a photograph Before compressing, check whether you need every page. Splitting out the relevant pages with a PDF splitter often solves the size problem entirely, without touching quality. For scanned documents specifically, aim for around 150 DPI for screen reading and 300 DPI if the document will be printed. Anything above 300 DPI adds size without visible benefit for most documents. A Note on Privacy Many online compressors upload your files to a server, process them there, and return the result which means your documents sit on someone else's infrastructure for some period. For photos of a passport, a signed contract, or medical records, that's a meaningful concern. Browser-based tools that process files locally avoid this entirely: the file never leaves your device. The ToolBox image and PDF tools work this way nothing is uploaded. Frequently Asked Questions What's the difference between lossy and lossless compression? Lossy compression permanently discards data that's unlikely to be noticed, achieving much smaller files used by JPEG. Lossless compression reorganizes data so the original can be perfectly reconstructed, giving smaller reductions used by PNG. Lossy suits photographs; lossless suits screenshots, logos, and text. What quality setting should I use for JPEG? Around 80 is the prac