Vision-centric tasks on VideoThinkBench mini (test)
37.3Average ScoreClaude Sonnet 4.5
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Claude Sonnet 4.5Model Category=Vision-Language Models2025.11 | 37.3 | 40 | 34 | 60 | 75 | 75 | 83 | 5.7 | 0 | 0 | 0 | |
| Gemini 2.5 ProModel Category=Vision-Language Models2025.11 | 35.6 | 33 | 23 | 40 | 95 | 95 | 68 | 2.1 | 0 | 0 | 0 | |
| GPT-5 highModel Category=Vision-Language Models2025.11 | 35.5 | 39 | 30 | 23 | 98 | 80 | 85 | 0 | 0 | 0 | 0 | |
| Sora-2Model Category=Video Generation Models, Video Duration=10 seconds2025.11 | 31.6 | 50 | 35 | 25 | 80 | 35 | 53 | 2.8 | 35.3 | 0 | 0 | |
| Qwen3-VL-235B-A22BModel Category=Vision-Language Models2025.11 | 30.2 | 24 | 17 | 30 | 93 | 55 | 83 | 0 | 0 | 0 | 0 | |
| Nano Banana ProModel Category=Image Generation Models2025.11 | 29.8 | 24 | 30 | 35 | 85 | 50 | 73 | 0.7 | 0 | 0 | 0 | |
| Qwen3-VL-32BModel Category=Vision-Language Models2025.11 | 29.6 | 33 | 21 | 20 | 85 | 55 | 78 | 4.1 | 0 | 0 | 0 | |
| Qwen3-VL-PlusModel Category=Vision-Language Models2025.11 | 29.4 | 32 | 29 | 30 | 90 | 35 | 78 | 0 | 0 | 0 | 0 | |
| Veo 3.1Model Category=Video Generation Models2025.11 | 27.7 | 34.4 | 24.3 | 30 | 77.5 | 40 | 70 | 0.7 | 0 | 0 | 0 | |
| MiniMax Hailuo 2.3Model Category=Video Generation Models2025.11 | 25.9 | 36.7 | 34.3 | 27.5 | 72.5 | 45 | 42.5 | 0 | 0 | 0 | 0 | |
| Seedream 4.5Model Category=Image Generation Models2025.11 | 24.4 | 25.6 | 16.3 | 30 | 75 | 35 | 62.5 | 0 | 0 | 0 | 0 | |
| GPT Image 1.5Model Category=Image Generation Models2025.11 | 19.2 | 24.4 | 15 | 17.5 | 38 | 50 | 47.5 | 0 | 0 | 0 | 0 | |
| Qwen-Image-Edit-2511Model Category=Image Generation Models2025.11 | 14.9 | 30 | 23.8 | 27.5 | 25 | 35 | 7.5 | 0 | 0 | 0 | 0 | |
| MOVA-360pModel Category=Video Generation Models, Resolution=360p2025.11 | 13.4 | 23.3 | 25.7 | 25 | 45 | 0 | 15 | 0 | 0 | 0 | 0 | |
| Seedance 1.0 ProModel Category=Video Generation Models2025.11 | 12.4 | 22.2 | 24.3 | 35 | 25 | 10 | 7.5 | 0 | 0 | 0 | 0 | |
| MOVA-720pModel Category=Video Generation Models, Resolution=720p2025.11 | 11.8 | 32.2 | 18.6 | 25 | 30 | 0 | 12.5 | 0 | 0 | 0 | 0 | |
| BAGEL (Image Output)Model Category=Image Generation Models2025.11 | 7.7 | 24.4 | 12.5 | 25 | 5 | 0 | 10 | 0 | 0 | 0 | 0 | |
| Wan2.2-TI2V-5BModel Category=Video Generation Models, Resolution=480P, Video Duration=5 seconds2025.11 | 7.3 | 17.8 | 10 | 20 | 7.5 | 10 | 7.5 | 0.7 | 0 | 0 | 0 |