DVBench - Animation: leaderboard
DVBench accuracy on questions about how a data video's charts move: transitions, annotations, camera and pacing.
Metric: Accuracy (%). Source: bomiaowang.github.io. Saturation forecast: Around February 2027. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Kimi K2.5 | 72.18 | #152 |
| 2 | Gemini 3.1 Pro (Preview) | 69.92 | #64 |
| 3 | Claude Sonnet 4.6 | 68.42 | #94 |
| 4 | GPT-5.4 | 63.91 | #83 |
| 5 | Qwen 3.5 397B A17B | 62.41 | #154 |
| 6 | Qwen 3.5 27B | 59.4 | #192 |
| 7 | Qwen 3.5 122B A10B | 58.65 | #184 |
| 8 | Gemma 4 31B (IT) | 57.89 | #178 |
| 9 | Qwen 3.5 9B | 54.89 | #391 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=dvbench-animation · How It Works · Data refreshed daily, snapshot 2026-11-05.