How to Resize Images for Social Media Without Distortion

Dimensions are how many pixels the image contains. File weight is how many bytes those pixels occupy after encoding. They are related, but not proportionally, and the same word covers both.

A 4,000-pixel-wide photograph saved at 200 KB and an 800-pixel-wide screenshot saved at 3 MB have opposite problems. The photograph needs resizing; the screenshot needs a different format or encoder, because its pixel count is already small. Deciding which problem you have first saves a lot of pointless slider-dragging.

The thing that visibly ruins a social image is almost never a slightly wrong pixel count. It is a wrong shape. Platforms fit your image into a frame with a fixed aspect ratio, and when your ratio does not match theirs they crop it, letterbox it, or scale it into the frame — which is where subjects lose their heads and logos lose their edges.

Get the ratio right and the exact pixel dimensions matter much less, because the platform will rescale to its own display size anyway. Use the Image Resizer with its aspect-ratio control when the whole picture should stay and only the dimensions should change, and the Image Cropper when the composition itself needs to change shape.

Almost every article on this topic publishes a chart of exact pixel sizes for each network. We deliberately do not, for a simple reason: those specifications change without announcement, and a chart is a snapshot of the day it was written. A guide that looks authoritative and is quietly eighteen months stale is worse than no chart at all, because you will not know which rows went wrong.

Aspect ratios are far more stable than pixel dimensions, so working in ratios survives platform changes that break a pixel table. When you do need an exact figure, read it from the platform's own current documentation at the moment you publish, and treat any third-party chart — including one on a site like this — as a hint rather than a source.

This is the assumption worth testing, so we measured it. Five fixtures were cropped with the production FiloTool cropper in Playwright Chromium on 2026-08-13, with bytes taken from the actual source and download files. Every output had fewer pixels than its input. Three of the five came out larger:

Resizing downward discards information, which is straightforward. Resizing upward has to invent it. FiloTool resizes with pica, which performs proper resampling rather than naive pixel doubling, so an enlargement looks smoother than a browser stretching the image in a layout — but smoother is not sharper. No resampling filter recovers detail the sensor never recorded.

If a source image is genuinely too small for its destination, the honest options are to obtain a larger original or to accept the softness. The Image Upscaler can help a marginal case, but treat any large enlargement as a compromise rather than a fix.

Resizing re-encodes the image, so the format you land in affects the result as much as the dimensions do. Converting to JPG on the way out will flatten any transparency onto a background colour, and a photograph and a screenshot respond very differently to the same format at the same quality.

Rather than repeat that reasoning here, the format decision guide covers which format suits which kind of image and shows measurements where the ranking inverts between a photograph and a graphic.

Because the variable parts are the platform specifications and the encoder behaviour, build the routine around checking rather than remembering.

Resizing is a re-encode, not an edit in place. Output is written at quality 0.95 by default, so a JPG round trip is lossy even when the dimensions are unchanged, and repeated exports compound that. Transparency survives only into formats that can store it. Animated sources are not preserved as animations by the still-image paths.

Processing happens in your browser, so there is no upload step and no server queue — practical limits come from your device's memory and the size of the image, not from your connection speed.