OpenVLM MME - Position: leaderboard
OpenCompass OpenVLM MME subcategory score for spatial position understanding in images.
Metric: Score. Source: huggingface.co. Saturation forecast: Estimated already saturated. 235 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | InternVL3-78B | 180 | #352 |
| 2 | InternVL3-14B | 168.3 | #505 |
| 3 | InternVL3-38B | 165 | #405 |
| 4 | Qwen 2 VL 72B | 165 | #460 |
| 5 | InternVL3-8B | 158.3 | #620 |
| 6 | Qwen 2.5 VL 7B | 158.3 | #643 |
| 7 | Qwen 2 VL 7B | 155 | #750 |
| 8 | GPT-4.1 (2025-04-14) | 153.3 | #250 |
| 9 | InternVL2.5-78B | 151.7 | #431 |
| 10 | GPT-4.1 Mini | 146.7 | #355 |
| 11 | MiniCPM-V-2.6 | 146.7 | #836 |
| 12 | InternVL2-8B | 143.3 | #817 |
| 13 | Moonshot v1 8K | 143.3 | #617 |
| 14 | Grok 2 (1212) | 140 | #496 |
| 15 | Pixtral-12B | 135 | #808 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mme-position · How It Works · Data refreshed daily, snapshot 2026-10-20.