MEGA-Bench - Application: Perception — leaderboard
Metric: Average Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 66.7 |
| 2 | Gemini 2.0 Flash (Preview) | 57.2 |
| 3 | Claude 3.5 Sonnet (20241022) | 55.6 |
| 4 | GPT-4o | 55.1 |
| 5 | Claude 3.5 Sonnet (20240620) | 53.2 |
| 6 | InternVL3-38B | 52.6 |
| 7 | Gemini 1.5 Pro (002) | 52.5 |
| 8 | InternVL3-78B | 52 |
| 9 | Qwen 2 VL 72B | 51.8 |
| 10 | InternVL2.5-78B | 51.7 |
| 11 | InternVL3-14B | 47.1 |
| 12 | Gemini 1.5 Flash (002) | 46.4 |
| 13 | Gemma 3 27B (IT) | 45.5 |
| 14 | GPT-4o Mini | 43.6 |
| 15 | InternVL3-8B | 42.8 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-application-perception · How the rankings work · Data refreshed daily, snapshot 2026-07-22.