YouTube Thumbnail Size: The Complete Dimensions Guide

The correct size is 1280 x 720 pixels, 16:9 aspect ratio, maximum file size 2MB, in JPG, PNG, or non-animated GIF format. YouTube's stated minimum width is 640 pixels, but nobody serious uploads at the minimum. Every thumbnail we ship at ClickTheory goes out at exactly 1280 x 720 because that is what YouTube's player and feed are built around.

A few things trip people up. First, the 2MB limit is real and YouTube will reject anything over it with an unhelpful error. Second, aspect ratio matters more than raw resolution. A 1920 x 1080 image works fine because it is still 16:9 and YouTube downscales it, but a 1280 x 800 image gets cropped or letterboxed and suddenly your subject's forehead is gone. Third, verify your account. Unverified channels cannot upload custom thumbnails at all, and I have had two onboarding calls where that was the entire problem.

If you want to see what properly sized, properly composed thumbnails look like in practice, our work is all built to this spec, one custom image per video, never templates.

Because almost nobody sees your thumbnail at 1280 x 720. On a phone in the home feed, your thumbnail renders at roughly 160 pixels wide. In the suggested sidebar on desktop it is about 168 pixels. That is an eighth of the canvas you designed on, and it is where the click decision actually happens.

This is the single most common failure I see when creators send me their old thumbnails. They look great full size in Photoshop. Then you shrink them to feed size and the text turns to noise, the face becomes a smudge, and the clever background detail disappears entirely. Design at 1280 x 720, but judge at 160 wide. At ClickTheory we check every draft at feed size before it goes to the client, and we test styles on our own YouTube channels first, so we know which compositions hold up when compressed.

The practical rules that fall out of this: one focal point, three to four words of text maximum, and contrast you can read from across the room. Small-size legibility is also most of what separates a thumbnail that lifts CTR from one that just exists. I wrote more about that mechanism in how to increase YouTube CTR.

YouTube overlays its own UI on top of your image, and if you put anything important where the UI sits, it gets covered. The big one is the duration stamp in the bottom right corner, which eats roughly a 140 x 50 pixel patch. Never put text or a logo there. On hovered previews and in some players, progress bars and watched-marker lines run along the bottom edge too.

My working rule is a margin of about 5 percent on all sides, so roughly 64 pixels left and right and 36 pixels top and bottom, with the bottom right corner treated as fully off limits. Keep faces and headline text inside that inner box. Edge-to-edge backgrounds are fine, edge-to-edge information is not.

There is a second, sneakier safe zone problem: your text sitting where the video title wraps underneath. If your thumbnail text just repeats the title, you have wasted the most valuable pixels on the platform saying the same thing twice. Thumbnail and title should be two halves of one hook, which is why every ClickTheory pack includes a title consultation with each thumbnail. Faces need safe zone thinking too, since a cropped chin reads very differently from a cropped forehead, and I go deep on that in faces in YouTube thumbnails.

Use PNG while designing, and upload whichever version stays under 2MB and looks clean. In practice that usually means a high-quality JPG for photographic thumbnails and PNG for flat graphic styles with hard edges and big type, because JPG compression puts ugly artifacts around sharp text edges.

Here is the part most guides skip: YouTube re-compresses your upload anyway. Whatever you send gets run through YouTube's own pipeline, so an over-optimized 90KB upload gives the algorithm less to work with and comes out mushier than a clean 1.5MB file. I do not chase tiny file sizes. I upload the best-looking file under the 2MB cap and let YouTube handle delivery.

Two more format notes. Animated GIFs are technically accepted as a format but the animation is stripped, so there is no trick there. And WebP is not accepted for upload even though YouTube serves your thumbnail as WebP after processing, which confuses people who try to re-upload a downloaded thumbnail and get rejected. If a thumbnail keeps failing to upload, it is almost always one of three things: over 2MB, wrong format, or an unverified account.

Yes, and this catches podcast channels constantly. Shorts are vertical, 9:16, and 1080 x 1920 is the size to design at if your channel is eligible to upload custom Shorts thumbnails. A 16:9 thumbnail dropped into a Shorts context gets cropped to a useless center strip.

For video podcasts, you are usually publishing full 16:9 episodes plus vertical clips, which means every episode needs both a 1280 x 720 episode thumbnail and vertical art for the clips. That doubles the design surface, and it is exactly the workflow our sister service handles: Pixel Samy edits video podcasts, episodes from $199 and clips at $25, while ClickTheory covers the thumbnail side.

Also remember your 1280 x 720 image does double duty in places you do not control: embedded players on blogs, social link previews, and search results all pull the same file at different rendered sizes. One image, a dozen render sizes. That is another argument for the boring, robust composition rules above, because you cannot art-direct every surface, you can only make an image that degrades gracefully. Music channels have their own version of this problem with cover art conventions, which I cover in thumbnail design for music artists.

Every thumbnail we deliver is 1280 x 720, under 2MB, checked at feed size, with safe zones respected, as a baseline that is not negotiable. On top of that spec, each image is custom built for its specific video. We do not resize a template and swap the text, because templates are exactly the thing that dies at 160 pixels wide.

The packaging is simple: one pack is $250 for 8 custom thumbnails, which works out to about $31 each. One thumbnail per video, unlimited revisions, a dedicated designer who learns your channel, and a title consultation with every image. Turnaround is 48 to 72 hours per thumbnail, with a rush option at +$20 when a video cannot wait. If you want to test two compositions against each other, an A/B variant is +$15, and testing is genuinely where sizing discipline pays off, because a variant that reads better at small size will usually win the test. I walk through that process in the YouTube thumbnail A/B testing guide.

We are based in Dubai and work with creators worldwide, fully async. Full details on packs and add-ons are on pricing and services.

Go back when an old video still gets impressions but the thumbnail breaks one of the rules in this guide. Swapping a thumbnail does not reset a video's performance history, it just changes what future impressions see, so fixing a wrongly cropped or illegible thumbnail on a video that still surfaces in search or suggested is close to free CTR.

My triage order: first, anything uploaded before you knew about the 16:9 crop, since stretched or letterboxed images look broken in the feed. Second, thumbnails with text in the bottom right corner under the duration stamp. Third, anything with more than five words of text, because that text was never being read anyway. Do not bother refreshing videos with near-zero impressions, since a better image shown to nobody changes nothing.

I keep a fuller decision framework, including how long to wait after a swap before judging it, in when to refresh YouTube thumbnails. And if the backlog is large, this is a sensible first pack to send us, since 8 fixes at $31 each on videos with proven impressions is the highest-probability spend in thumbnail design.