Reduce image file size and resize photos without installing anything. Compression happens entirely in your browser, so your images are never uploaded to a server.
Two things shrink a file. Quality decides how much colour detail JPEG or WebP throws away at the same resolution; max size cuts the pixel count itself. For a photo destined for the web, bringing the long edge down to 1600–2000px saves far more than dropping quality alone.
Quality decides how much colour information JPEG or WebP discards. The pixel count stays the same and the file gets smaller by throwing away detail the eye is least likely to miss.
Dimensions cut the pixel count itself. Halving the long edge quarters the pixels, and the file size falls roughly in proportion.
For anything destined for a web page, resizing first is far more effective. Taking a 4000px photo down to 1600px at 80% quality beats leaving it at 4000px and dropping quality to 60% — smaller file, better-looking result on screen.
JPEG and WebP are lossy: they genuinely discard data, and what is discarded never comes back.
PNG is lossless. It only compresses repetition, so it barely shrinks photographs where every pixel differs, and works well on screenshots, icons and diagrams where large areas share a colour.
Never re-compress repeatedly. Running an already-JPEG photo through JPEG again stacks new artefacts on old ones. Always start from the original if you still have it.
For photographs, WebP is 25–35% smaller than JPEG at equivalent quality and every current browser supports it. Choose JPEG when maximum compatibility matters more.
Anything needing transparency has to be PNG or WebP. JPEG has no alpha channel, so transparent areas fill with white.
Screenshots keep their text edges crisp in PNG. JPEG produces ringing artefacts around lettering.
Logos and flat graphics belong in SVG if you can manage it, since vectors scale without loss and usually weigh least. That is outside this tool's scope.
In-article images want 1200–1600px on the long edge, under 200KB. Above that, mobile load time becomes noticeable.
Social preview and OG images use 1200×630px, the standard aspect, under 300KB.
Email attachments run into receiving servers that typically cap at 10–25MB. For multiple photos, aim under 1MB each.
Online form uploads have strict, small limits, and 2MB is common. Pass the limit first and worry about quality second, which means reducing dimensions before quality.
Compression runs through the browser's Canvas API on your own device. The image never crosses the network.
That distinction is real when the image is an ID document, a contract or a medical record. Services that upload may promise deletion after processing, but the file still passes through someone else's disk in the meantime.
The trade-off is memory: very large batches can slow the tab or stall it. Process in smaller groups when that happens.
Phone photos carry EXIF data — capture time, camera model, and often GPS coordinates. Posting an unedited photo taken at home publishes those coordinates.
Because this tool redraws the image through Canvas, EXIF does not survive into the output. Compression incidentally strips the location data.
If you are organising a photo library and need capture dates preserved, this is the wrong tool — keep your originals.
No. All compression runs locally in your browser using the Canvas API, so your images never leave your device.
You can compress common formats like JPG, PNG and WebP, and export a smaller optimized version to download.
You control the quality level, so you can balance file size against visual quality yourself. For web use, you can often shrink files a lot with no visible difference.
What this tool bases its numbers on, and how far those numbers go.
The image is decoded to a canvas, optionally scaled so its long edge meets your maximum, and re-encoded as JPEG or WebP at the quality factor you set. Quality controls how much colour and high-frequency detail the encoder discards at a given resolution; the maximum size reduces the pixel count itself.Halving both dimensions quarters the pixels: 4000 × 3000 = 12,000,000 becomes 2000 × 1500 = 3,000,000. That is usually a far bigger saving than lowering quality at full resolution.