Thumbnail A/B Comparison Tool
You cannot judge two thumbnails at full size on a big monitor. Put them side by side at the widths a viewer actually sees and the winner usually becomes obvious.
100% free · runs in your browser · nothing is uploaded
Drop two versions of the same thumbnail. Both are read locally — nothing is uploaded, so you can compare artwork for a video that has not been published.
Side by side, at the sizes people actually see
Desktop home feed · 360px
Mobile feed · 246px
Up-next sidebar · 168px
What the pixels say
| Measure | A | B | Reads as |
|---|
These are measurements, not predictions. They tell you which image is louder in a feed — only a real test on your own audience tells you which one wins.
Deciding between two thumbnails
Most thumbnail decisions get made the worst possible way: two files open in a viewer at 1280 pixels wide, on a colour-accurate monitor, by the person who made them. At that size both look fine. The comparison that matters is at 168 pixels, in a row of eight competing images, glanced at for a third of a second.
This puts your two versions in exactly that comparison — the desktop home feed at 360 pixels, the mobile feed at 246, the up-next sidebar at 168 — and lets you apply the two filters that strip a thumbnail back to what a viewer's peripheral vision actually processes.
The squint test and grayscale
Blur removes detail and leaves composition. Whichever version still says "person, reaction, one big thing" when blurred is the one that reads in a scrolling feed. Grayscale removes colour and leaves tonal contrast — if a thumbnail only separates from its background because of a colour difference, it collapses in grayscale, and it will collapse next to a brightly coloured competitor too.
What the numbers mean
Four measurements are taken from the pixels of each image:
- Brightness — average luminance. Very dark thumbnails disappear against a dark-mode feed; blown-out ones lose their text.
- Contrast — how widely the tones spread. This is the single best predictor of whether a subject separates from its background at small size.
- Colour punch — average saturation. Muted images are easy to scroll past, whatever their composition.
- Busyness — how much of the frame is fine detail and edges. High numbers mean the image turns to mush when shrunk.
These are measurements, not predictions. They tell you which image is louder. Whether louder wins for your audience is a question only a real test on your own channel can answer — and YouTube's own Test & compare feature in Studio is the place to run it. Use this tool to pick the two candidates worth testing, not to skip the test.
When you have a winner, check it once more at full detail in the thumbnail preview, and read the guide to Test & compare for how to run the real experiment.
Last updated August 15, 2026.
Frequently asked questions
How do I A/B test a YouTube thumbnail?
A real A/B test happens in YouTube Studio's Test & compare, which serves each version to a slice of your audience and measures watch time. This tool is the step before that: it shows you both versions at real feed sizes so you can pick which two are actually worth testing.
Which thumbnail is better?
The one that still reads when blurred. That single check catches most bad thumbnails — if the subject and message survive the blur, the image is doing its job at 168 pixels. The measurements below the previews back that judgement up with contrast, colour and busyness numbers.
Why compare at 360, 246 and 168 pixels?
Those are the rendered widths YouTube uses for the desktop home feed, the mobile feed and the up-next sidebar. Your thumbnail is almost never seen at 1280 pixels, so judging it there tells you very little.
Are my thumbnails uploaded anywhere?
No. Both images are read from your disk by your own browser and analysed with Canvas on your machine. Nothing is sent to us, so comparing artwork for an unpublished video is safe.
Can I compare more than two?
Two at a time — that is what makes a comparison decisive. To narrow a set of four, run two pairs and then compare the winners.
Stop designing thumbnails. Start generating them.
Describe your video, pick your face, and Thumblore returns click-ready 1280×720 thumbnails in seconds — free to start.
Try Thumblore freeMore free tools
Video Frame Grabber
Scrub your video, grab any frame, and save it at 1280×720 — or auto-grab six across the whole clip.
Open tool →YouTube Title Checker
See where your title gets cut off on every surface, and check length, keyword placement and caps.
Open tool →YouTube Description Generator
Build a description that ranks — keyword above the fold, valid chapters, clean hashtags and tags inside 500 characters.
Open tool →