Search DevTools

Jump to any tool or page

Image Compressor

Compress JPG, JPEG, WebP, and SVG images without losing quality. Supports multiple images, live progress, and drag-and-drop, all in your browser.

Settings

Images

Choose or drop some images and they will appear here.

Data Transformation

About Image Compressor

Reduce image file size in the browser by re-encoding at a chosen quality or converting to a more efficient format such as WebP or AVIF. The trade-off that governs everything here is that lossy compression discards information permanently, so each save cycle removes detail that no later step can restore — compressing an already-compressed JPEG compounds damage rather than undoing it.

Related Data Transformation tools

JSON FormatterImage to JSONPII RedactorSheet to JSONJSON to CSV/XLSX

Frequently asked questions

What is generation loss, and how do I avoid it?
Every JPEG save quantises the image's frequency coefficients, discarding fine detail. Decoding and re-encoding repeats that quantisation on data already stripped once, so artefacts accumulate: blocky 8x8 edges, ringing around high-contrast lines, and colour blotching in smooth gradients. The damage is cumulative and irreversible. Always re-encode from the original source rather than from a previously compressed export, and if a file must pass through several editing steps, keep it in PNG or another lossless format until the final save.
When is PNG the wrong choice, and when is it the right one?
PNG is lossless and compresses by finding repeated patterns, which works brilliantly on flat colour, sharp edges, and text — screenshots, logos, diagrams. It works poorly on photographs, where every pixel differs slightly from its neighbour and there is little to deduplicate, often producing files several times larger than a visually indistinguishable JPEG. The reverse also holds: saving a screenshot as JPEG puts ringing artefacts around every letter. Match the format to the image content, not to a default.
What does chroma subsampling do to my images?
JPEG separates brightness from colour and stores colour at reduced resolution, exploiting the eye's greater sensitivity to luminance detail. At 4:2:0 the colour planes are halved in both dimensions, cutting size substantially with little visible cost on photographs. It is destructive on saturated edges though — red text on a dark background smears noticeably, and thin coloured lines bleed. For screenshots or graphics with sharp colour boundaries, 4:4:4 subsampling or a lossless format avoids the effect entirely.
Should I use WebP or AVIF instead of JPEG?
Both typically produce smaller files at comparable quality — WebP by roughly a quarter to a third, AVIF often further, particularly at low bitrates where its handling of gradients is markedly better. WebP support is now effectively universal; AVIF is broadly supported but encoding is considerably slower, sometimes by an order of magnitude, which matters for bulk conversion. AVIF also tends to smooth fine texture at aggressive settings, so film grain and detailed foliage can look waxy rather than merely softened.
Why did resizing or re-encoding strip my image's metadata and colour?
Re-encoding through a canvas produces new pixel data with no EXIF block, so orientation, GPS, camera settings, and copyright are all lost. Orientation is the visible one: a photo relying on an EXIF rotation flag appears sideways once that flag is gone, unless the pixels were rotated first. ICC colour profiles disappear the same way, so a wide-gamut image re-encoded without its profile is reinterpreted as sRGB and shifts towards duller, flatter colour. Losing metadata is often desirable for privacy, but it should be a decision.