MEGA-Bench Task - Geometry Descriptive — leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | InternVL3-78B | 35.7 |
| 2 | InternVL3-8B | 35.7 |
| 3 | Pixtral-12B | 35.7 |
| 4 | Gemma 3 12B (IT) | 28.6 |
| 5 | InternVL2-8B | 28.6 |
| 6 | Qwen 2 VL 72B | 28.6 |
| 7 | Qwen 2 VL 7B | 28.6 |
| 8 | InternVL3-38B | 28.6 |
| 9 | Gemini 2.5 Pro | 21.4 |
| 10 | Aquila-VL-2B | 21.4 |
| 11 | GPT-4o Mini | 21.4 |
| 12 | Claude 3.5 Sonnet (20240620) | 21.4 |
| 13 | Llama 4 Scout Base | 21.4 |
| 14 | Qwen 2 VL 2B | 14.3 |
| 15 | InternVL3-14B | 14.3 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-geometry-descriptive · How the rankings work · Data refreshed daily, snapshot 2026-07-22.