MEGA-Bench Task - FunQA Unexpected Action Creative Video: leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 46 |
| 2 | Gemini 1.5 Pro (002) | 42 |
| 3 | Claude 3.5 Sonnet (20241022) | 38.7 |
| 4 | Gemini 1.5 Flash (002) | 34.7 |
| 5 | Llama 4 Scout Base | 34.7 |
| 6 | Qwen 2 VL 7B | 33.3 |
| 7 | Claude 3.5 Sonnet (20240620) | 32.7 |
| 8 | Qwen 2 VL 72B | 31.3 |
| 9 | GPT-4o Mini | 31.3 |
| 10 | InternVL3-38B | 30.7 |
| 11 | Gemma 3 27B (IT) | 29.3 |
| 12 | GPT-4o | 28.7 |
| 13 | Gemma 3 12B (IT) | 28 |
| 14 | Gemma 3 4B (IT) | 27.3 |
| 15 | InternVL2.5-78B | 26.7 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-funqa-unexpected-action-creative-video · How It Works · Data refreshed daily, snapshot 2026-09-05.