Social Media Image Sizes

Most of the size charts online list forty dimensions, thirty of which nobody uses. These are the ones worth knowing, and the two rules that save more trouble than the numbers do.

The sizes

• Instagram post, square: 1080 × 1080.
• Instagram post, portrait: 1080 × 1350 - takes more room in the feed, which is why most brands now use it.
• Instagram story and Reels: 1080 × 1920.
• YouTube thumbnail: 1280 × 720, under 2MB.
• Facebook page cover: 820 × 312.
• LinkedIn banner: 1584 × 396.
• Display ads: 300 × 250, 728 × 90, 300 × 600.

Rule one: the interface sits on top of your design

A story has your profile at the top and a reply bar at the bottom. A LinkedIn banner has your photo over the lower left. A YouTube thumbnail has a duration badge in the bottom right. None of this is optional, so anything that matters belongs in the middle - not because the middle looks better, but because it is the only part guaranteed to be visible.

Rule two: uploading bigger does not make it sharper

Every platform re-compresses what you upload to its own dimensions. A 4000 pixel image is downscaled and recompressed, often looking worse than one supplied at the right size. Export at the listed size and stop there.

Test at thumbnail scale, not full size

The single most useful habit: shrink your design on screen until it is about the width of your thumb, then look at it. That is the size a feed shows it at. If the headline is not readable there, the answer is fewer words rather than bigger type - bigger type in a small space just wraps.

One design, several sizes

Every template on this site switches size in the editor and re-flows rather than cropping, so an Instagram post can become a story without a redesign. The wording almost always needs shortening as the canvas gets narrower, which is a content decision rather than a design one.

For print rather than a feed the trade-offs change completely - paper has no interface sitting on top of it, but it also gets two seconds in the hand. That is flyer territory.