Image compression guide
What does image compression do?
Image compression means storing an image using fewer bytes. For photos, the most effective method is JPEG's lossy compression: details that are hard to perceive are simplified and the file size drops clearly. In lossless formats such as PNG, the pixels are preserved; the saving is usually smaller.
GörselKüçült re-encodes the image you upload at the quality you choose. If re-compressing in the same format would make the file larger or would not bring a meaningful saving, it does not change the file and keeps the original.
How do you choose the quality setting?
The quality value is a scale between 1 and 100; a higher value means more detail and a larger file. For photos published on the web, the 80-85 range usually delivers a clear size drop with a loss the eye does not notice.
In images with text, screenshots and thin lines, lowering the quality can create blur around the edges. For these images, choosing PNG or keeping the quality high gives a better result.
Resizing and the long-edge limit
Image size is measured in pixels. For example, if you are showing a 4000-pixel-wide photo at 800 pixels on a web page, the remaining resolution is wasted. Setting the long-edge limit to a value such as 1600 pixels shrinks the file while largely preserving the visible quality. With aspect-ratio-preserving resizing, the image is not stretched.
The difference between JPEG, PNG and WebP
JPEG is ideal for photos; it gives small files but does not support transparency. PNG is lossless and preserves transparency; it suits screenshots and logos but stays large for photos. WebP offers the advantages of both: it works in both lossy and lossless modes and is supported in modern browsers.
Saving the same image as WebP gives a smaller file than JPEG in most cases. If you need a fallback for old browsers, JPEG is the safe choice.
Transparency and background
Transparent areas are preserved in PNG and WebP outputs. When you convert an image to JPEG, those areas are placed on white because transparent regions have no background. If you need transparency, you should keep the output format as PNG or WebP.
The orientation of photos taken with a phone
Photos taken with a phone carry the capture orientation inside the file. If that information is ignored, the photo can appear sideways. When the 'Auto-correct phone orientation' option is on, the capture orientation is applied and the image is saved in the correct orientation.
Bulk processing and ZIP
You can upload more than one image at a time. Each image can be downloaded separately; if you want, you can take all the outputs as a single ZIP file. Bulk processing is ideal for shrinking product photos or a whole folder without dealing with them one by one.
File size before and after compression
The tool shows the original size, the new size and the saving rate on each output's label, so you can see at a glance how much each image shrank. When you process more than one image, the output cards carry the same information as a list.
When comparing, note the quality setting too: trying the same image at quality 82 and 60 and putting the results side by side is the most practical way to find the right balance for your own content.
Target file size for the web
There is no single magic number for page speed; the goal is to drop unnecessary bytes without hurting the visible quality. For hero images, 200-400 KB, and for list and card images, under 100 KB works comfortably for most sites.
When used together with the long-edge limit, both the pixel dimensions and the file size drop; the two together give a better result than lowering the quality alone. Always keep the original file as well; compression is for the working copy.
When should you not compress?
Compressing a high-resolution image prepared for print or an original file you keep for archiving creates irreversible quality loss. In these cases, the best approach is to keep the original file separately and compress a working copy for the web.