LinkedIn Post Image Size, and Which Format Gets Engagement.

The short answer

For LinkedIn post image size, HubSpot’s LinkedIn guidance asks for images larger than 552 by 276 pixels, a shape between 3:1 and 2:3, PNG, JPG or GIF, and 10 MB at most, up to nine per post. For engagement, post portrait: in our study, portrait image posts earned 13% more than the typical post.

Why the shape matters more than the pixels

Founder social on LinkedIn runs on the feed post, and the image is the first thing a reader sees. Getting the file accepted is the easy part. Choosing the shape is where the reactions and comments are won or lost.

The specs below get an image onto LinkedIn without being cropped. What matters more is what each shape, and each kind of post, earns once it is there. Alongside when you post, it is one of the cheapest things to get right.

The specs LinkedIn will accept

HubSpot’s guidance for LinkedIn image posts sets out these rules:

• Use images larger than 552 pixels wide by 276 pixels high.
• Keep the shape between 3:1 (wide) and 2:3 (tall). A wider image is centred and cropped.
• Use PNG, JPG, JPEG or GIF.
• Keep each file to 10 MB or less.
• Add up to nine images to one post.

HubSpot also notes that once an image is uploaded, it cannot be resized. LinkedIn’s own developer documentation for its Images API lists JPG, GIF and PNG, and a limit of under 36,152,320 pixels per image.

How we measured what each shape earns

On 10 September 2026 we collected every public post from 346 prominent LinkedIn writers, 74,119 posts in all. The figures here use the 39,858 posts from the 187 authors whose typical post earns 100 or more reactions and comments.

Engagement means reactions plus comments, the two counts LinkedIn shows on a post. The typical post in that set, the one in the middle, earns 320. Each figure below compares a group of posts with that typical post: 1.13 means the group’s middle post earns 13% more.

A portrait image is one taller than it is wide, a landscape image one wider than it is tall. The study reads that from the image’s pixel dimensions.

Portrait earns the most

Portrait images came out on top: 1.13, about 13% more reactions and comments than the typical post. They were also the most common shape, at 19,681 posts.

Portrait posts drew more conversation too. Comments made up 22.3% of their engagement, against 18.8% across the whole set.

Measured per follower, which strips out how big the author’s audience is, portrait posts ran 22% above the set. Square images ran higher on that measure, at 35% above, so portrait does not win every way you cut it.

Square and landscape

Square images earned 0.93, about 7% less than the typical post, across 4,262 posts.

Landscape did worst of the three shapes: 0.81, about 19% less than the typical post. It was also below the set per follower, at 18% under, and comments made up only 14.4% of its engagement.

A post with no image at all earned 0.87, about 13% less than the typical post, and the least per follower of any shape, 31% below the set.

An image beats text on its own

The cleanest comparison is the same author against themselves, which removes the difference between big and small accounts. Across 108 authors, a text-only post earned 0.39 of what the same author’s image posts earned, well under half. Text only did better for just 14 of those 108 authors.

That is a strong case for adding an image. If a point is worth posting, it is worth a portrait image.

More images can help too. Posts with four or more images earned 1.15, about 15% more than the typical post, against 1.05 for posts with one image. Only 959 posts carried four or more, so treat that as a lead worth testing, not a rule.

Never post a link card

A link card is a post that shows as an article preview. It was the weakest format on every measure.

Across 41 authors, a link card earned 0.52 of the same author’s image posts, about half. Only 7 of those 41 authors did better with the link card. Across the whole set it earned 0.51, about half the typical post, and 68% below the set per follower.

A habit of posting bare links is exactly the kind of leak an NBK audit finds early.

Documents start conversations

A document post is a carousel: a set of slides the reader swipes through. Its engagement came in slightly below the typical post, at 0.93.

What it does well is comments. Comments made up 27.5% of a document post’s engagement, the highest share of any format and about 47% above the set. If the aim of a post is a discussion, a carousel is the format to use.

Video looks strong until you account for audience

Video earned 1.23, about 23% more than the typical post, the highest raw figure of any format.

Per follower it fell to 0.84, 16% below the set, and comments made up just 12.6% of its engagement. The study notes that a raw figure counts audience as much as craft, and video’s drop once audience size is stripped out fits that.

Reposts and quote posts

A repost earned 0.83, about 17% less than the typical post. A quote post, a repost with the author’s own words above it, earned 0.20, a fifth of the typical post: a middle figure of 64 reactions and comments against 339 for an original post.

If something is worth sharing, it is worth a post of your own with your own image.

What the numbers cannot tell you

Every figure here is a comparison across accounts of very different sizes. None proves that a shape causes the result, and the study says so plainly.

The writers studied are prominent, not typical LinkedIn members. LinkedIn does not publish impressions, so nothing here measures reach. And the study recorded whether a post had an image and what shape it was, not what was in it.

Next step

Make portrait your default, keep landscape for when the picture demands it, and stop posting bare links. If your LinkedIn output feels busy but not effective, start with an NBK audit. For the other half of the job, see our guide to the LinkedIn posting schedule.