Benchmark
VBench
Evaluate generated videos for subject consistency, motion, visual quality, and prompt alignment.
Published leaderboard
Explore scores and evaluation settings on the publisher’s leaderboard.
Open leaderboardCoverage
The original text-to-video benchmark covers 16 dimensions, including subject consistency, motion smoothness, temporal flickering, and spatial relationships.
Methodology
Generate videos using the published prompt suite and apply the official evaluator for each dimension. The authors compare automated measurements with human preferences; inspect dimension-level results as well as aggregates.