OpenVLM MMT-Bench - Geometrical Relativity — leaderboard
Metric: Score (%). Source: huggingface.co. 207 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 2 VL 2B | 65 |
| 2 | Aquila-VL-2B | 65 |
| 3 | Gemma 3 27B | 65 |
| 4 | Gemma 3 4B | 60 |
| 5 | Qwen 2 VL 7B | 55 |
| 6 | Pixtral-12B | 55 |
| 7 | Llama 3.2 11B Instruct | 55 |
| 8 | Gemma 3 12B | 55 |
| 9 | Gemini 1.5 Flash | 50 |
| 10 | Claude 3 Haiku | 50 |
| 11 | InternVL3-78B | 35 |
| 12 | InternVL3-14B | 35 |
| 13 | Llama 3.2 90B Vision Instruct | 35 |
| 14 | MiniCPM-V-2.6 | 30 |
| 15 | GPT-4.1 (2025-04-14) | 30 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmt-bench-geometrical-relativity · How the rankings work · Data refreshed daily, snapshot 2026-07-22.