Resize Image

An image's dimensions are how many pixels wide and tall it is; its file size is how many bytes those pixels take to store. The two are related but not the same thing, and confusing them is the most common reason people end up with images that are still too heavy after 'making them smaller'.

Resizing attacks the first number. Halving both width and height leaves a quarter of the pixels, and the file usually shrinks by a similar proportion — a much larger effect than any compression setting can achieve on its own. This is why resizing first and compressing second gives dramatically better results than compressing alone.

The reference point that matters is the space the image actually occupies on screen. A 4000-pixel-wide photograph shown in a 800-pixel column is delivering twenty-five times more data than the display can use, and no amount of clever compression recovers that waste.

Making an image smaller is a well-behaved operation. The resampler averages groups of pixels down into fewer, and the result is generally sharper and cleaner than the original at its new size. You can downscale aggressively without visible harm.

Enlarging is a fundamentally different problem, because the detail you are asking for was never captured. Interpolation can only guess at values between the pixels that exist, so edges go soft and fine texture turns to mush. A 20–25% enlargement is usually acceptable; doubling is visibly degraded.

If you truly need a larger image, an AI upscaler is the only approach that helps, because it synthesises plausible detail from a trained model rather than stretching what is there. Even then, it is inventing information, which matters if the image is evidentiary or technical.

Aspect ratio is the relationship between width and height. Keep it locked and your image scales proportionally — a 3:2 photo stays 3:2 and simply gets smaller. Unlock it and you can force any dimensions you like, at the cost of stretching or squashing everything in the frame. Faces make this obvious immediately; landscapes hide it for a while and then look subtly wrong.

When you need an image to fit an exact box that does not match its proportions, the right sequence is crop then resize, not stretch. Cropping discards the parts outside your target shape, leaving the remaining content undistorted, and resizing then brings it to the pixel dimensions you need.

Setting dimensions once and applying them across a folder is the normal case for product catalogues, gallery uploads and any site with a fixed image slot. Percentage mode is often the better choice for a mixed batch, because it scales each image relative to its own size rather than forcing everything to identical dimensions — useful when portrait and landscape shots are mixed together.

Everything is processed in your browser and delivered as a single ZIP, so a hundred images is one download rather than a hundred. Nothing is uploaded unless an image is too large for a browser tab to paint — which depends on its resolution rather than its file size.