Image Sizes for Social Media and the Web: A Practical Resizing Guide

You post a photo that looks flawless on your iPhone, and an hour later it's on your feed looking soft, smeared, and vaguely disappointing. What happened between your camera roll and the timeline? In a word: recompression. Every social platform re-encodes every image you upload — and the further your file is from what the platform wants, the more aggressively it gets mangled. Understanding a handful of numbers, and resizing before you upload, is the difference between photos that survive the pipeline and photos that get chewed up by it.

Why platforms recompress everything you upload

Social networks serve billions of images a day, mostly to phones on mobile connections. Storage and bandwidth at that scale are enormous costs, so every platform runs uploads through its own pipeline: strip metadata, resize to internal maximums, and re-encode at an aggressive quality setting. You cannot opt out. What you can control is the input. When you upload a 48-megapixel, 12 MB original, the platform's automated resizer does severe surgery in one violent step. When you upload a file already near the platform's target dimensions and a reasonable file size, the pipeline has little to do — and your photo comes out looking close to what went in.

That's the whole game: do the resizing yourself, with care, before the platform does it for you, without any.

Decoding the numbers: HD, Full HD, and 4K

Resize presets in most tools are named after display resolutions, and it's worth knowing what they actually mean:

Notice the gap between even 4K and a modern phone camera's 12–48 megapixels. A photo at Full HD is already more resolution than most feeds will ever display, since platforms commonly cap the long edge of images somewhere between 1080 and 2048 pixels. Uploading more than that doesn't add sharpness — it just gives the recompression algorithm more work to do, often with worse results.

Practical sizing rules for social feeds

Exact pixel specs shift as platforms redesign, so instead of memorizing a chart that will be stale next quarter, internalize the durable rules:

• Aim for 1080–2048 pixels on the long edge. This range matches or slightly exceeds what major feeds display, giving the platform an easy job.
• Mind the aspect ratio. Feeds favor squares (1:1), portraits around 4:5, and landscapes around 16:9; stories and vertical formats want 9:16. If your image is far outside the expected ratio, the platform will crop it — usually badly. Decide the crop yourself.
• Keep files comfortably under platform limits. A Full HD JPG at 80% quality lands around a few hundred kilobytes to 1 MB — small enough that most pipelines pass it through gently.
• Upload JPG for photos. Feeds are built around it. Exotic inputs get transcoded anyway; see our format guide for when PNG or WebP make sense elsewhere.
• Sharpness reads as quality. Downscaling slightly softens images; a photo resized to the display size you expect, rather than left huge, avoids the double-downscale (yours plus theirs) that produces mush.

Sizing for the web beyond social

The same logic applies to blogs, marketplaces, forums, and portfolios — except you often are the pipeline. Oversized images are the single most common cause of slow web pages. For content images, 1200–2000 pixels wide covers everything from mobile to large desktop layouts; hero banners might justify 2400. Pair sensible dimensions with 70–85% JPG quality (or WebP) and images stop dominating your page weight. If a listing site or CMS enforces upload caps — 1 MB and 2 MB are common — resizing to Full HD before compressing almost always gets you under the bar without visible damage, as covered in our file-size guide.

Batch thinking: the ten-photo problem

Nobody resizes one photo. You come home from a trip, an event, or a product shoot with dozens, and the destination — a carousel post, a listing, an album share — needs all of them consistent: same long edge, same format, same rough file size. Doing that one image at a time in an editor is exactly the kind of chore that makes people give up and upload originals, surrendering the results to the platform's compressor. Batch tools that apply one preset across a set of images turn a twenty-minute slog into a single action, and consistency across a set is itself a quality signal — a carousel where every frame matches feels deliberate.