Skip to content

Benchmark

VBench-Long Quality Score

Scores are percentages.

Models measured
9
Scores recorded
9
Sources
1
IQ weight
0.96

Score distribution

84.7%
Mean
85%
Median
1.5%
Std dev
82.6%
10th percentile
85.7%
90th percentile
82.3%
Min
82.3%87.3%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.96 for video-generation models, measured over 8 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Vidu Q12025-04-17· VBench data docs87.3%
2Veo 3· VBench data docs85.7%
3Sora· VBench data docs85.5%
4Hunyuan Video2025-05-22· VBench data docs85.1%
5Kling V1.6· VBench data docs85%
6Video 01· VBench data docs84.8%
7Vidu Q4· VBench data docs83.8%
8Mochi 1· VBench data docs82.6%
9LTX Video5s 768×512· VBench data docs82.3%

Data credits: VBench (Vchitect), MIT (licence)