Viral YouTube thumbnail, 16:9, comparing three AI video models on the same face. Layout: three vertical panels side by side, each showing the SAME old fisherman's weathered face in the rain at dusk as rendered by a different model — image 1 = left panel, image 2 = middle panel, image 3 = right panel. Show the full face with both eyes clearly visible looking into the camera in every panel (zoom out if needed), keep each model's distinct look (Kling cooler and darker, FLUX detailed wet skin, Seedance warm close-up), but brighten them, add contrast and crisp detail: deep wrinkles, raindrops on the skin, piercing eyes. Thin bright white glowing dividers with a big yellow lightning-bolt "VS" on each. Huge bold white condensed sans-serif labels at the top with thick black outline: "KLING 4", "FLUX 3", "SEEDANCE 2.5". Clean, bold, readable at small size like images 4 and 5, no other text. Small dark rounded pill with the word "Doitong" bottom right.
