compress-or-die.com

What is compress-or-die.com?

compress-or-die.com is a free, browser-based image compression site. Its promise is narrow and clear: shrink image files as much as possible while keeping the format's own settings under your control. The compressors run entirely on your device, so images are not uploaded to a server, and no account is required.

What you can actually do there

The site is organised as one tool per file format rather than a single generic compressor:

• JPEG — turns lossless PNGs or high-quality JPEGs into smaller lossy JPEGs.
• PNG — quantises and re-encodes PNGs, keeping full alpha transparency.
• GIF — shrinks animated GIFs, with lossy compression, dithering control and frame trimming.
• WebP — lossy and lossless WebP, with a separate quality setting for the alpha channel.
• JPEG XL and AVIF — listed as upcoming converters to newer formats.

There is also a drop-or-paste area: drop any image anywhere on the page and the matching compressor opens. The site states it is made in Germany and EU GDPR compliant, and that images never leave your device.

Who it suits

This is aimed at people who want to understand why a file got smaller, not just click "compress". If you are preparing web images and care about the difference between quantisation, dithering and alpha handling, the per-format approach is more useful than a one-size-fits-all tool. If you just want a quick automatic result with no decisions, a simpler compressor will feel faster.

A practical next step

Start with the format you already have. Export a copy of your image, run it through the matching compressor, and compare the result at 100% zoom against the original before replacing anything. Keep the original file until you are satisfied.

Trade-offs to weigh

• Local processing means privacy and no upload wait, but the speed depends on your own machine.
• Per-format tools give finer control, but you need to know which format you are working with.
• Smallest possible file size is the stated goal, so expect to trade some quality or transparency fidelity at aggressive settings.
• The rebuild is ongoing, so some format tools may not be live yet.

How do I compress a JPEG without uploading it to a server?

Use a browser-based compressor that processes the file locally on your device. On compress-or-die.com, the JPEG tool is described as running entirely in your browser, with the page stating that "nothing is uploaded" and "your images never leave your device." That is the key distinction from most online compressors, which upload your file to a server, compress it there, and send back a download link.

How to do it

1. Open the JPEG compressor on the site, or drop/paste the image anywhere on the page and it will open the matching compressor for that format.
2. Adjust the quality and other settings the tool exposes. The site's approach is to give you the format's own settings rather than a single generic slider, so you can trade file size against visible artifacts yourself.
3. Run the compression and download the result. The processing happens in your browser tab, so the file does not travel over the network.

Why local processing matters

• Privacy and confidentiality. Client photos, medical images, unreleased product shots, or anything under NDA never leave your machine. With an upload-based tool you are trusting a third party's retention policy.
• Speed and reliability. No upload or download round trip, so large files or slow connections are less painful. It also works if you are on a restricted network.
• No account friction. The page describes the tools as free and without an account.

Trade-offs to weigh

• Browser-based compression uses your device's CPU and memory. Very large images or batch jobs can be slower than a desktop application, and a low-powered laptop or phone may struggle.
• Because everything runs in the tab, closing or reloading the page mid-job can lose your work.
• Format coverage depends on the site's rebuild progress. The JPEG, PNG, GIF and WebP compressors are listed as live, while JPEG XL and AVIF are marked as coming soon, so check before relying on those.

Practical scenario

A freelance photographer preparing a client gallery on a laptop can compress dozens of JPEGs without uploading client work to an unknown server. If the images are numerous or huge, a desktop tool such as ImageOptim or a command-line encoder like MozJPEG may still be faster, but it requires installation. For occasional files and privacy-sensitive work, the in-browser route is the simpler choice.

As a next step, compare the compressed result at 100% zoom against the original before publishing. Local compression gives you privacy, not automatic quality — the settings you choose still determine how the image looks.

Which image format should I choose for the smallest file size?

For the smallest possible file size, the answer depends on whether you need transparency or animation, and whether you can accept lossy quality:

• JPEG – smallest for ordinary photos with no transparency.
• WebP – smallest when you need lossy or lossless compression with alpha transparency.
• AVIF and JPEG XL – next-generation formats that can beat both, but support is still limited.
• PNG – only when you need lossless quality with full alpha transparency.
• GIF – only when the image is animated and you cannot convert it.

For most people, the practical decision is simpler than the format list suggests. A photograph destined for a website should become a JPEG or WebP. A logo or screenshot with transparency should become a PNG or lossless WebP. An animated GIF should become an animated WebP or a heavily optimized GIF.

A quick comparison

The trade-off is compatibility. AVIF and JPEG XL produce the smallest files, but not every browser, editor or content management system handles them yet. WebP is the safer middle ground: broad browser support and a size advantage over JPEG and PNG.

A concrete scenario: you have a 4 MB product photo with a transparent background. Saving it as PNG keeps the transparency but may stay several megabytes. Converting it to lossy WebP with alpha can cut the file dramatically while preserving the transparency. If the background does not need to be transparent, dropping the alpha channel and using JPEG usually wins outright.

Image compression to the smallest file size possible offers one compressor per format, including JPEG, PNG, GIF and WebP, and states that processing happens entirely in your browser, so files are not uploaded. That matters if you are compressing client work, medical images or anything you would rather not send to a server. The site also notes that AVIF and JPEG XL tools are planned but not yet live, so for those formats you would currently need another tool.

A useful next step is to test the same image in two formats at comparable visual quality and compare the resulting file sizes. Start with WebP, then check whether JPEG or PNG does better for that specific image. If you control the audience and know their browsers support AVIF or JPEG XL, test those too, but keep a fallback format for everyone else.

How can I shrink an animated GIF while keeping smooth motion?

Shrink an animated GIF by reducing what each frame stores and how many frames you keep, rather than only lowering overall quality. Smooth motion depends on frame rate and consistent timing, so the priority is: trim frames you don't need, cut the color palette carefully, then apply lossy compression with controlled dithering.

On Image compression to the smallest file size possible, the GIF compressor is described as handling animated GIFs with lossy compression, dithering control and frame trimming — all running in the browser, with images never uploaded. That combination matches the workflow above: frame trimming addresses motion length, dithering control addresses the banding that appears when you shrink the palette, and lossy compression reduces per-frame data.

A practical order of operations

1. Trim first. Remove duplicate or near-duplicate frames and any lead-in or tail you don't need. Motion looks smoother when the remaining frames are evenly spaced than when a long clip is squeezed at low quality.
2. Reduce dimensions before frame rate. A GIF at 600 px wide with 12 frames per second usually looks better than one at 1200 px with 24 fps. Viewers notice softness less than stutter.
3. Lower the frame rate in steps. Try dropping from 25 to 15 or 12 fps and compare. Below roughly 10 fps, motion reads as a slideshow regardless of image quality.
4. Cut the palette gradually. GIF is limited to 256 colors per frame; fewer colors mean smaller files but more banding in gradients and skin tones. Check flat-color graphics and text first, since they survive palette reduction well.
5. Use dithering deliberately. Dithering hides banding but adds noise, and noise compresses poorly. For smooth gradients, a light dither often beats none; for flat illustrations, turn it off.
6. Apply lossy compression last. Once the frame count, size and palette are settled, lossy settings give the final size reduction with the least visible harm.

When GIF is the wrong target

If the animation has many colors, photographic gradients or long duration, converting to a modern format will beat any GIF tuning. Animated WebP and AVIF both support full color and better compression, and the same site lists WebP as available and AVIF as in progress. The trade-off is compatibility: GIF plays essentially everywhere, including in email clients and older chat apps, while WebP and AVIF need a fallback for some contexts.

Quick decision guide

• Flat colors, icons, short loops: aggressive palette reduction and no dithering.
• Screenshots or UI demos: moderate palette, trim aggressively, keep frame timing even.
• Photographic or gradient-heavy motion: expect visible limits; test WebP or AVIF instead.
• Strict compatibility needed: stay with GIF and accept a larger file.

A useful next step is to compress one GIF twice — once with frame trimming only, once with trimming plus palette reduction — and compare them side by side at actual display size. Judging at 100% zoom on a desktop often exaggerates problems that disappear in the feed or chat window where the GIF will actually appear.

Can I compress PNG images with transparency without visible quality loss?

Yes. The PNG compressor at Image compression to the smallest file size possible is built for exactly that case: it re-encodes PNGs with quantization tuned to avoid visible loss while keeping full alpha transparency. Because it runs entirely in your browser, your PNG never leaves your device, which matters if you are compressing client artwork, screenshots of internal tools, or product mockups under an NDA.

What "no visible loss" actually means

PNG is lossless by nature, so any real size reduction comes from reducing the color palette or precision of the pixel data. The trade-off is:

• Lossless PNG optimization (recompression, filter and metadata cleanup): no pixel changes at all, but savings are usually modest, often single-digit percentages.
• Lossy PNG quantization (reducing color depth, sometimes with dithering): much larger savings, and at sensible settings the difference is invisible on screen — but it is a real change to the file.

If you need byte-identical pixels for archival or print, use the lossless route. If the PNG is destined for a website, app UI, or slide deck, lossy quantization at a conservative setting is usually the better call.

A practical workflow

1. Keep your original PNG untouched as the master.
2. Compress a copy, then compare it side by side with the original at 100% zoom, not just at thumbnail size.
3. Check the tricky areas specifically: soft drop shadows, semi-transparent overlays, thin 1px borders, and gradients. These are where quantization shows up first as banding or blotchiness.
4. Inspect the alpha channel separately. Transparency is often where lossy PNG tools fail, producing halos or jagged edges around cut-out subjects. The tool's stated goal of keeping full alpha transparency is the feature to verify here.
5. Only then push the setting further until artifacts appear, and step back one notch.

When PNG is the wrong format

If your image has no transparency and no sharp text or line art, a lossy format will beat even a well-quantized PNG by a wide margin. For photographs, JPEG remains the pragmatic default; for modern web delivery, WebP and AVIF typically win on size at equivalent quality. The same site offers per-format tools for JPEG, WebP, GIF and, in progress, JPEG XL and AVIF, so you can compare outputs from one place rather than juggling separate services.

One caveat worth planning around: the project is being rebuilt format by format after a change of maintainer, so not every compressor is live yet. Check which tools are currently available before you build a workflow around a specific one.

A quick decision rule: transparency or crisp edges → PNG, compressed conservatively and checked at 100%. Photographic content without transparency → JPEG, WebP or AVIF. Archival master → lossless only.

What happened to the original compress-or-die website?

The original compress-or-die was created and run by Christoph Erdmann from 2016 to 2026, but he had to stop maintaining it and the site was taken down. It is now operated by Timo Strauß, a friend of Christoph's, who is rebuilding the compressors with modern web technologies and bringing them back format by format—some tools are live while others are still marked "Soon." Timo notes that the project partly returned because of the friendly messages and requests received after the old site went down.

If you used the site before, expect a partial rather than complete restoration: check the compressor list for the format you need before relying on a bookmark, and use the feedback page to report bugs or suggest features.