Wan v2.6 Crushes AnimateDiff on Prompted Video

Wan v2.6 Image to Video scored 15.8 on prompted video generation, more than doubling AnimateDiff's 7.3, according to benchmark results. While AnimateDiff maintained temporal consistency, it defaulted to generic imagery rather than executing specific prompts. Wan v2.6 consistently translated written direction into visible, cinematic choices, such as depicting a courier in a narrow wet alley at blue hour wearing a rain cape and carrying a pastry box.

AnimateDiff’s 7.3 versus Wan v2.6 Image to Video’s 15.8 is not a close result; it’s a separation in kind. Model A can hold a shot together over time, but in both tests it defaulted to safer, more generic imagery instead of executing the prompt. Model B consistently translated specific written direction into visible, cinematic choices on screen. In Cinematic motion , Wan v2.6 did the real directing work: a courier in a narrow wet alley at blue hour , wearing the rain cape , carrying pastry box...