Score thumbnails 0-100 across 10 visual dimensions and preview them in a real YouTube card - before you publish.
Drop a thumbnail on the left to get its 0-100 score, dimension breakdown, and YouTube preview.
Scores are heuristic measurements of your image pixels against published YouTube CTR best practices. They do not predict YouTube's actual recommendation algorithm, which is proprietary and considers audience, topic, and history. Use the score as a design checklist, not a guarantee.
A thumbnail's only job is to earn the click — CTR is YouTube's strongest post-upload signal. This tester reads the actual pixels of your image (downscaled onto a Canvas in your browser) and scores 10 visual dimensions, each weighted by how much it predicts a high-click thumbnail: contrast and text readability at 15% each, subject prominence and mobile visibility at 12%, then composition, color discipline, brightness, badge safety, aspect ratio, and text balance. The weighted sum is the 0–100 score. Because text must survive at 335×186 on a phone screen, the strongest single predictor is whether large, high-contrast words still read after downscaling.
Each dimension is measured on a downscaled canvas, multiplied by its weight, and summed. Text readability and contrast carry the most weight — they predict the click — while the timestamp-badge corner is checked for safety.
Leo runs a cooking channel. For the same recipe video he has two thumbnail drafts — a moody photo-only shot and a bold text-led version — and wants to pick one before publishing.
Score your YouTube thumbnail 0-100 across 10 visual dimensions and preview it in a real YouTube search card at 335x186. Compare candidates side by side. 100% in your browser.
The YouTube Thumbnail Tester grades any thumbnail from 0 to 100 using ten weighted dimensions measured on the actual pixels of your image: color contrast, text readability, subject prominence, rule-of-thirds composition, color discipline, brightness, duration-badge safety, mobile-size visibility at 335x186 pixels, aspect ratio, and text balance. Upload up to four candidates and it scores all of them side by side, crowns the strongest, and lists the three highest-impact fixes for each. A simulated YouTube search-result card shows exactly how your artwork reads next to real titles and channel avatars at feed size. Every calculation runs locally in your browser - nothing is uploaded.
New creators checking their first thumbnails before the upload button, established channels running structured A/B tests across a video catalog, thumbnail designers who want objective measurements instead of taste arguments, and agencies auditing click-through potential across many client channels at once. It is also the natural second step after the YouTube Thumbnail Generator: generate a candidate, test it, tweak, re-test.
(1) Drag in one thumbnail, or drop up to four to compare. (2) The engine downsamples your image, measures luminance distribution, edge density by band, color variety, and how much detail survives the 335x186 phone-size downscale, then weights the ten dimensions into one score. (3) Read the dimension breakdown - each low score comes with a concrete fix, such as moving detail out of the bottom-right corner where the duration badge sits, or enlarging text that disappears at feed size. (4) Pick the winner, export, publish. Re-test any time; your images never leave the tab.
Your thumbnail is graded 0-100 across ten weighted dimensions measured on the actual pixels: color contrast, text readability, subject prominence, composition, color count, brightness, edge safety around the timestamp badge, mobile-size visibility at 335x186, aspect ratio, and text-to-area ratio. Low-scoring dimensions come with specific fix suggestions.
No. Scoring runs entirely in your browser with Canvas - the image never leaves your device. This matters for unpublished videos, where a leaked thumbnail would spoil a launch. Compare candidates freely; nothing is stored after you close the tab.
Most YouTube views happen on phones, where a thumbnail renders at roughly 335x186 pixels. Fine detail and small text disappear at that size. The tester reproduces that exact size so you can catch readability problems before publishing rather than after the click-through rate data arrives.
What do you call a crab that plays baseball?
No paywalls, no signups, no data sold. Built by a solo developer who believes useful tools should be accessible to everyone.
☕Support me on Ko-fi— keep tools free100% of proceeds go towards hosting & building more free tools.