Choose a target size in KB and compress your JPG locally in the browser.
Most people compress images to "make them smaller." That works fine when you just need a webpage to load faster. But there's a long list of situations where an approximate size simply won't do:
Government and university portals often enforce strict upload limits — 100 KB for a passport photo, 200 KB for a scanned document, 500 KB for a thesis figure. Go over by even 1 KB and the form rejects your file.
Email attachments have hard caps. Gmail allows 25 MB total, but many corporate mail servers cap individual attachments at 5 MB or even 2 MB. When you're sending three product photos and a PDF, precision matters.
Auction and marketplace listings — platforms like eBay, Poshmark, and regional classifieds often specify maximum image sizes per upload. A 2.1 MB photo on a platform that rejects anything over 2 MB is useless.
Visa and immigration applications are notoriously strict. The US, UK, Schengen countries, and many others specify exact KB ranges for uploaded photos. Too small and the resolution fails; too large and the upload fails.
In all these cases, "approximately 200 KB" isn't the goal. You need to compress your image to a specific file size — and you need it done correctly on the first try.
The underlying technique is called iterative binary compression. Here's what that means in plain terms:
The tool starts by compressing your image at a medium quality setting — say, quality level 75 out of 100 (for JPEG). It measures the output. If the result is 320 KB but you wanted 200 KB, it knows quality 75 was too generous. It tries 55. If that produces 160 KB — too small — it tries 65. Each iteration narrows the gap between the actual output and your target.
This binary search approach typically converges within 5–8 iterations, which happens in under two seconds for most images. The final result lands within a few KB of your requested size.
Two critical details:
Format awareness. The tool respects the original format. A PNG with transparency won't be silently converted to JPEG (which would destroy the alpha channel). If the format can't reach the target size through quality reduction alone, the tool will let you know and suggest a format change.
Resolution isn't sacrificed unless necessary. Quality reduction is tried first because it preserves pixel dimensions. Only if the target size is unrealistically small for the given resolution will the tool suggest or apply downscaling.
The most common specific request we see is to compress images to 200 KB. This number shows up everywhere — job application portals, online forms, university submissions, professional certification uploads.
A 200 KB JPEG at 1920×1080 typically retains enough quality for web display, documents, and even printed materials at moderate DPI. For a passport-style photo (600×600), 200 KB is generous — the result will look nearly identical to the original.
If 200 KB is your specific target, Pixes offers a dedicated 200 KB compression tool that skips the manual size input entirely. Same precision, one fewer step.
Some portals are stricter. Government ID systems, certain Asian marketplace listings, and older university portals sometimes cap uploads at 100 KB.
At 100 KB, you're working with tighter constraints. A 2480×3508 A4 scan will show visible quality loss at this size. But a 800×600 web image? It'll compress to 100 KB with minimal visible difference. The key variable is always resolution relative to the target size.
For this specific threshold, the compress to 100 KB tool is optimized for that exact range and gives faster results when you know 100 KB is your limit.
Every compression involves a tradeoff: smaller files mean lower visual quality. But the relationship isn't linear, and most people misunderstand where the critical thresholds are.
JPEG quality 95 → 80: Barely noticeable difference to the human eye. File size drops by roughly 40–50%. This is the sweet spot for most use cases.
JPEG quality 80 → 60: Fine details start softening. Text in images becomes slightly less crisp. Photos still look good at normal viewing distance. File size drops another 30–40%.
JPEG quality 60 → 40: Noticeable artifacts appear — especially around sharp edges, text, and color gradients. Acceptable for thumbnails and preview images, but not for hero images or print.
JPEG quality below 40: Visible blocking, color banding, and detail loss. Use only when the file size requirement absolutely demands it.
When you set a target size, the tool navigates this tradeoff automatically. But understanding the ranges helps you set realistic targets. Asking for a 50 KB version of a 4000×3000 photo is technically possible — but the result will look rough.
Not all image formats compress the same way. This affects whether your target size is achievable:
JPEG is the most predictable format for size-targeted compression. Quality adjustment maps cleanly to file size, and the range is wide. A single image can typically be compressed from 5% to 95% of its original size by adjusting quality alone.
PNG is trickier. PNG compression (which is lossless) works by finding patterns in pixel data. A screenshot with large flat-color areas compresses well. A photograph with continuous tones compresses poorly — sometimes a "compressed" PNG is actually larger than the JPEG equivalent. If you need a specific small size from a photo-heavy PNG, converting to JPEG is usually the practical path.
WebP offers the widest compression range with good quality retention. A WebP at quality 60 often looks as good as a JPEG at quality 75, at a smaller file size. If your target platform supports WebP, it's the most efficient choice for hitting tight size targets.
AVIF compresses even further but is slower to encode and not universally supported. For aggressive size targets on modern browsers, it's worth considering.
Before tools like this existed, the workflow was painful:
Open image in an editor → Export as JPEG at quality 60 → Check file size → 287 KB (too big) → Re-export at quality 45 → Check file size → 156 KB (too small) → Try quality 52 → 203 KB (close, but maybe not close enough) → Try quality 50 → 194 KB...
Each cycle takes 30–90 seconds depending on the editor and file size. For a single image, that's annoying. For a batch of 20 product photos that all need to be under 150 KB? That's an hour of your life spent on a mechanical task.
Automating this with binary search eliminates the back-and-forth. The same batch of 20 images takes under a minute — upload, set target, download. The general JPG compressor handles batch work well when you need to reduce multiple files with consistent quality.
No tool is magic. Here are the real constraints:
Unrealistic targets. If you try to compress a 5000×4000 photo to 20 KB, the result will be unusable. The tool will compress it as much as possible, but it can't create information that isn't there. As a rough guide: target at least 1 KB per 1000 pixels of total resolution (width × height ÷ 1000) for acceptable JPEG quality.
Already-compressed images. If your image is already heavily compressed, there's little room to reduce further. A JPEG saved at quality 30 can't lose much more detail without becoming unrecognizable.
Transparency requirements. JPEG doesn't support transparency. If your image has an alpha channel and you must keep it, you're limited to PNG or WebP — both of which have less predictable size-to-quality ratios for photographic content.
Animated images. GIF and animated WebP files have frame-based complexity. File size depends on frame count, duration, and per-frame content. The target-size tool works on static images.