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: saturated. 51 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 64.7 |
| 2 | O1 | 58 |
| 3 | Claude 3.5 Sonnet (20241022) | 54.3 |
| 4 | GPT-4o | 54.2 |
| 5 | Claude 3.5 Sonnet (20240620) | 52.1 |
| 6 | Gemini 1.5 Pro (002) | 49.6 |
| 7 | InternVL3-78B | 47.1 |
| 8 | Qwen 2 VL 72B | 46.8 |
| 9 | InternVL3-38B | 45.8 |
| 10 | InternVL2.5-78B | 45.6 |
| 11 | Gemini 1.5 Flash (002) | 43.8 |
| 12 | Gemma 3 27B (IT) | 43.7 |
| 13 | GPT-4o Mini | 43.1 |
| 14 | InternVL3-8B | 37.9 |
| 15 | Gemma 3 12B (IT) | 37.6 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench · How It Works · Data refreshed daily, snapshot 2026-09-05.