WBench - Video Quality: leaderboard
The video quality score of WBench (2026), a multi-turn benchmark for interactive video world models: text-driven, camera-controlled and action-conditioned video world models extend a scene turn by turn under their own controls. The score is the mean of six video-quality metrics (aesthetic, imaging, flickering, dynamic degree, smoothness, HPSv3), on a 0-100 scale; higher is better.
Metric: Score (0-100): the mean of six video-quality metrics (aesthetic, imaging, flickering, dynamic degree, smoothness, HPSv3), on the 158 navigation cases, multi-turn generation, each model under its own text, camera or action control; higher is better. Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 20 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Seedance 1.5 | 82.1 | |
| 2 | Wan2.7 | 81.5 | |
| 3 | Kling 3.0 | 81.4 | |
| 4 | LingBot-World | 78.9 | |
| 5 | HY-World 1.5 | 78.1 | |
| 6 | YUME 1.5 | 77.6 | |
| 7 | HY-Video 1.5 | 77.6 | |
| 8 | Happy Oyster | 77.3 | |
| 9 | LTX-2.3 | 77.1 | |
| 10 | Infinite-World | 77 | |
| 11 | Matrix-Game 3.0 | 75.5 | |
| 12 | LongCat-Video | 75.4 | |
| 13 | Genie 3 | 75.2 | |
| 14 | Kairos 3.0 | 74 | |
| 15 | Matrix-Game 2.0 | 73.8 |
Interactive version: theaggregate.ai/benchmark?slug=wbench-video-quality · How It Works · Data refreshed daily, snapshot 2026-11-21.