MEGA-Bench — leaderboard
TIGER-Lab multimodal evaluation with 500+ tasks across vision-language understanding, generation, and reasoning with 44 models.
Metric: Overall Score (%). Source: huggingface.co. Status: saturation imminent. 51 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 64.7 |
| 2 | O1 | 58 |
| 3 | Gemini 2.0 Flash (Preview) | 55.3 |
| 4 | Claude 3.5 Sonnet (20241022) | 54.3 |
| 5 | GPT-4o | 54.2 |
| 6 | Claude 3.5 Sonnet (20240620) | 52.1 |
| 7 | Gemini 1.5 Pro (002) | 49.6 |
| 8 | InternVL3-78B | 47.1 |
| 9 | Qwen 2 VL 72B | 46.8 |
| 10 | InternVL3-38B | 45.8 |
| 11 | InternVL2.5-78B | 45.6 |
| 12 | Gemini 1.5 Flash (002) | 43.8 |
| 13 | Gemma 3 27B (IT) | 43.7 |
| 14 | GPT-4o Mini | 43.1 |
| 15 | InternVL3-14B | 42.5 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench · How the rankings work · Data refreshed daily, snapshot 2026-07-22.