OpenVLM MMT-Bench - Geometrical Relativity: leaderboard
The "Geometrical Relativity" subtask of OpenVLM MMT-Bench. OpenCompass OpenVLM evaluation of MMT-Bench: multimodal tasks covering visual recognition, localization, OCR, counting, relation reasoning, visual math, and hallucination detection.
Metric: Score (%). Source: huggingface.co. 207 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 3 27B | 65 | #607 |
| 2 | Qwen 2 VL 2B | 65 | #1220 |
| 3 | Aquila-VL-2B | 65 | #1238 |
| 4 | Gemma 3 4B | 60 | #1089 |
| 5 | Qwen 2 VL 7B | 55 | #750 |
| 6 | Pixtral-12B | 55 | #808 |
| 7 | Gemma 3 12B | 55 | #671 |
| 8 | Llama 3.2 11B Instruct | 55 | #1126 |
| 9 | Gemini 1.5 Flash | 50 | #618 |
| 10 | Claude 3 Haiku | 50 | #803 |
| 11 | InternVL3-78B | 35 | #352 |
| 12 | InternVL3-14B | 35 | #505 |
| 13 | Llama 3.2 90B Vision Instruct | 35 | #622 |
| 14 | GPT-4.1 (2025-04-14) | 30 | #250 |
| 15 | MiniCPM-V-2.6 | 30 | #836 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmt-bench-geometrical-relativity · How It Works · Data refreshed daily, snapshot 2026-10-20.