AnimationBench - Personality Consistency: leaderboard
Whether facial acting, posture and body language of a generated cartoon character stay faithful to its personality, on AnimationBench, an animation image-to-video benchmark (2026) built on IP preservation and the Twelve Basic Principles of Animation, about 30 generated videos per dimension, scored by Qwen3-VL-Max yes/no question sets. Percent; higher is better.
Metric: IP personality consistency: whether facial acting, posture and body language stay faithful to the character's personality (%), animation-style image-to-video generation from the AnimationBench close-set suite (fixed source images and prompts, about 30 videos per dimension per model), scored by Qwen3-VL-Max yes/no question sets aggregated per video; higher is better. Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 8 models tracked. Updated 14 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Veo 3.1 | 91.61 | |
| 2 | Kling 2.6 | 91.61 | |
| 3 | Seedance 1.0 Pro | 90.91 | |
| 4 | Wan2.2-I2V-A14B | 89.58 | |
| 5 | Seedance 2.0 | 88.33 | |
| 6 | FramePack | 88.33 | |
| 7 | HunyuanVideo-1.5 | 86.53 | |
| 8 | Sora 2 Pro | 86.01 |
Interactive version: theaggregate.ai/benchmark?slug=animationbench-personality-consistency · How It Works · Data refreshed daily, snapshot 2026-11-22.