MEGA-Bench Task - Mensa IQ Test: leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude 3.5 Sonnet (20240620) | 55 |
| 2 | Claude 3.5 Sonnet (20241022) | 49.5 |
| 3 | Gemini 2.5 Pro (Preview 03-25) | 48.7 |
| 4 | Gemini 1.5 Flash (002) | 48 |
| 5 | Gemma 3 27B (IT) | 47.7 |
| 6 | Gemini 1.5 Pro (002) | 46.6 |
| 7 | Qwen 2 VL 72B | 44.3 |
| 8 | InternVL2.5-78B | 43.2 |
| 9 | InternVL3-78B | 40.7 |
| 10 | GPT-4o | 40.3 |
| 11 | Qwen 2 VL 7B | 39.2 |
| 12 | InternVL3-8B | 37 |
| 13 | InternVL3-38B | 34.8 |
| 14 | GPT-4o Mini | 33.5 |
| 15 | Gemma 3 12B (IT) | 33 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-mensa-iq-test · How It Works · Data refreshed daily, snapshot 2026-09-05.