Skip to content
Processing locally — files never leave your device

Compress Images Online

Compress JPG, PNG, and WebP images with a quality slider. Compare before / after sizes side-by-side.

How to use Image Compressor

  1. Drop one or more images into the upload area, or click to browse.
  2. Adjust the quality slider — lower values shrink the file at the cost of fidelity.
  3. Optionally cap the maximum width or output format.
  4. Click "Compress" — the tool processes each image locally.
  5. Download individual results, or grab them all as a ZIP.

Compress images locally — free, instant, and private

Image compression is a CPU-bound task, and server-based compressors charge for the cycles or mine your uploads. Modern browsers expose the same encoding routines desktop apps use, and your phone or laptop has plenty of idle CPU. This tool decodes each image to raw pixels, re-encodes it at your chosen quality inside a Web Worker (so the page never freezes), and hands you the result — the photo bytes never touch a server. A side effect worth knowing: re-encoding from raw pixels strips EXIF metadata, including GPS coordinates, which is usually what you want before posting a photo publicly.

Lossy vs. lossless, in one minute

Lossy formats (JPEG, lossy WebP) shrink files by discarding detail the eye barely registers: fine high-frequency texture is quantized away and colour information is stored at lower resolution than brightness, because human vision is sharper at luminance than at colour. Lossless formats (PNG, lossless WebP) only repack the existing data more efficiently, so every pixel survives exactly — and the savings are correspondingly modest. That asymmetry explains the most common surprise here: a photograph can drop 70% as a JPEG, while an already-optimised PNG barely moves.

What the quality slider does, with real numbers

The slider controls how aggressively the encoder quantizes detail. The relationship to file size is steep at the top and flat at the bottom: going from quality 1.0 to 0.8 typically halves a photo or better, while going from 0.5 to 0.3 saves little and looks visibly worse. As a concrete example, a 4 MB 12-megapixel JPEG straight off a phone usually lands around 1–1.5 MB at quality 0.8 with no visible difference at normal viewing size, and around 600–800 KB at 0.6, where flat gradients like skies start to show faint banding. Screenshots and UI graphics behave differently — their hard edges generate exactly the high-frequency detail lossy encoders sacrifice, so keep those as PNG or lossless WebP.

Resize first, then compress

Pixel count dominates file size. A 4,000-pixel-wide photo displayed in an 800-pixel slot carries 25× more pixels than the screen will ever show; capping the width often saves more than any quality setting. The max-width option here does both in one pass — downscale, then encode — which is the right order, since scaling after lossy encoding bakes the artifacts in.

Choosing an output format

  • JPEG — best for photographs when compatibility matters everywhere, including old email clients. Quality 0.7–0.85 is the sweet spot for the web.
  • WebP — roughly 25–35% smaller than JPEG at equivalent quality, handles gradients better, and supports transparency. Supported by every modern browser.
  • PNG— lossless. Use for screenshots, line art, and anything needing pixel-perfect fidelity or transparency. Don't expect dramatic size reductions.

When not to compress

Keep originals untouched if you'll edit them again — every lossy re-encode compounds the loss — and never replace an archival copy with a compressed one. Skip compression for images headed to print, where the artifacts invisible on screen can show at 300 DPI. And if a file is already small for its job, leave it alone: re-encoding a well-optimised 80 KB thumbnail buys nothing.

Related tools

Frequently asked questions

Are my photos uploaded?
No. The re-encoding runs inside a Web Worker in your own tab, which keeps the page responsive and keeps the image bytes off the network entirely — a folder of photos is squeezed at the speed of your CPU, not your upload link.
How much can I shrink an image?
Typical results: 60–80% smaller for photographs at quality 0.7, with no perceptible quality loss. Screenshots and graphics with sharp edges compress less aggressively because lossy formats trade detail for size.
Which formats are supported?
JPEG, PNG, and WebP — both as input and output. WebP usually gives the smallest file at equivalent quality and is supported by every modern browser.
Why does my PNG get bigger when I compress it?
PNG is lossless. If your input is already optimised, re-encoding it as PNG won't shrink it — and may add metadata. Convert to WebP or JPEG for serious size reductions.
Is there a file-size limit?
Practically, the limit is your browser tab's memory. Phones can handle 20–30 MB images comfortably; desktops easily handle 100 MB+. Very large RAW images may need to be downscaled first.
Does compressing remove EXIF data like GPS location?
Yes. The image is decoded to raw pixels and re-encoded from scratch, so camera metadata — GPS coordinates, device model, timestamps — is not carried into the output. That's a privacy bonus when sharing photos publicly, but keep the original if you need that data.
What quality setting should I use?
For photos going on the web, 0.7–0.8 is the sweet spot: large savings with no visible difference at normal viewing size. Drop to 0.6 only if you're fighting a hard size limit, and expect mild blockiness in flat areas like skies. Above 0.9 the file grows quickly for gains you won't see.
Does compressing an image twice make it worse?
With lossy formats, yes — each re-encode discards a little more detail, a problem known as generation loss. One pass at quality 0.7 looks better than two passes at 0.85. Always compress from the original rather than re-compressing an already-compressed copy.

More tools you might find useful in the same flow.

Built by Muhammad Tahir · About