About Image Compressor
High-resolution images eat bandwidth and storage, slowing down pages and blowing email attachment limits. This tool compresses PNG, JPEG and WebP images locally in your browser: drag in files for batch processing, tune the quality slider to balance size and clarity, and see before/after file sizes at a glance. Stripping EXIF metadata (location, camera model, capture time) also trims a little extra weight — something to be aware of if you need to keep GPS or camera info for archival. A common misconception is that you must shrink resolution to shrink the file; in practice most photos look great at 80% JPEG quality, so tune quality first to cut size dramatically while keeping full pixel dimensions. The original file is never modified — every compression produces a new file you can save and swap in. For example, dropping a 500KB PNG and tuning quality compresses it to about 120KB, great for page performance. Prefer compressing above-the-fold images first for the most visible loading-time gains with minimal visible quality loss.
A common misuse is pushing quality down blindly: going from one hundred to eighty often shrinks the file sharply with little visible change, while eighty to sixty degrades the image clearly for a limited gain, so returns diminish fast. Find the quality knee first and then consider lowering resolution, which usually contributes more to size reduction and is often acceptable on screen, though never for assets meant to be zoomed.
The split depends on how the asset ships. Bitmap compression is fine for on-screen use, but keep a lossless master and compress only derived copies when high-resolution preview or print matters, since compression cannot be undone. With a lossy target, note that repeated compress-edit-compress cycles accumulate generation loss, so the workflow should be master then derive each time, never editing a compressed result further.
Keep masters and derivatives in separate places with names that show the relationship. Without that distinction, originals get overwritten and only the compressed version survives, a loss no further compression can undo. When handing files to a third party, state that they receive derivatives so nobody edits and re-exports on top of them and stacks generation loss until the asset is unusable.
A final premise is the viewing condition: judging compression must happen under the target condition, since an image for the web looks poor when enlarged on a large screen while one for print looks fine on screen. Match the check to the use, viewing web images at common display widths and print images at the resolution their print size needs, or the conclusion is meaningless and either over- or under-compresses.
Compared with resizing, compression keeps pixel dimensions intact, which matters when a layout depends on exact width and height and only quality or colour depth can give the size back. In fixed layouts such as precisely aligned cards compression is the only option, while resizing is more effective wherever the layout can adapt.
How It Works Internally
Lossy compression trades high-frequency detail for size, and the return diminishes: dropping quality from one hundred to eighty shrinks the file sharply with little visible change, while going from eighty to sixty visibly degrades it for a limited gain. Find the quality knee first and then consider resolution, which pays better than pushing quality down further.
A 2 MB photo at quality 100At quality 80 about 500 KB with little visible change | at 60 about 350 KB with clearly visible detail lossMeasured on This Page
| Compression works | 410.8 KB dropped to 2.4 KB (99% saved), and PIL decoded the output cleanly at 900x700 |
|---|---|
| Size change explained | The achievable ratio depends on the source and quality setting: the same page reported 0% savings on a flat PNG, because such images are already at PNG's floor and cannot shrink further (noted) |
| Dimensions untouched | Compression changes encoding quality, not pixel count — resizing needs a separate scaling tool |
| Page feedback | Original size, compressed size and savings percentage are shown after upload, with a direct download |
How much a compressor saves is entirely a property of the source: phone photos often shed 60–90%, while screenshots, flat colour and already-optimised PNGs save nothing — measured here as 0% on a flat PNG versus under 1% of the original on a noisy JPEG. So 'it did not get smaller' rarely means a broken tool; it means no redundancy is left. Three cautions: JPEG is lossy and repeated passes accumulate artefacts, so keep the original; compression never changes pixel dimensions; and transparent artwork flattened into JPEG turns white or black, so keep PNG when alpha matters, accepting the larger file.
How to Use
- Open "Image Compressor"
- Drop one or more PNG, JPEG or WebP files
- Move the quality slider and compression reruns
- Choose auto, JPEG or WebP as the output format
- Download each image, or use "Download All"
Use Cases
- Optimize web images — Compress banners and product photos to shrink page weight and speed up loading.
- Slim email attachments — Compress photos below email attachment size limits to avoid send failures.
- Save cloud storage — Batch-compress images before uploading to cloud drives to save space and sync traffic.
- Social media fit — Compress images within the size limits of WeChat, Twitter and other platforms for quick sharing.
- E-commerce optimization — Compress product hero images while keeping clarity, balancing display quality and load speed.
FAQ
Does compression lose quality?
JPEG and WebP use lossy compression - lower quality means smaller size but more artifacts. PNG is lossless, so size reduction is limited. Preview at 80% quality first, then adjust.
How do I balance quality and size?
75-85% JPEG quality usually cuts weight noticeably with little visible loss. Start at 80%, then nudge down only if your target size demands it.
Does compression change dimensions?
By default it only adjusts compression and keeps the original resolution. To also scale down, process the image with our resize tool first.
Are EXIF and GPS data kept?
No. Compression strips EXIF metadata (location, camera, time) to reduce size. If those details matter for archives, back up the original first.
Which formats are supported?
PNG, JPEG and WebP for both input and output, with cross-format compression (e.g. PNG to WebP) also available.
Can I batch compress?
Yes. Drag in multiple images to batch-compress them, each showing its own before/after size comparison.
Quality drops visibly after compression — how do I balance it?
Quality is driven by the slider; 60–75 suits most screenshots and photos, below 40 blocks appear. Step down in fives from 70 rather than jumping to the minimum. For text-heavy screenshots prefer PNG or lossless WebP — JPEG ringing blurs small text.
The file got larger after compressing?
Two cases: the source was already an optimised WebP/AVIF, so re-encoding only adds loss and size; or a transparent PNG became a JPEG, where transparency flattens to a background and savings are small. Keep the original format and tune quality in the first case; switch to lossy PNG or WebP in the second.