OpenVLM MME - Commonsense Reasoning: leaderboard
OpenCompass OpenVLM MME subcategory score for commonsense reasoning over images.
Metric: Score. Source: huggingface.co. Saturation forecast: Estimated already saturated. 235 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2 VL 72B | 178.6 | #460 |
| 2 | InternVL3-78B | 175.7 | #352 |
| 3 | InternVL2.5-78B | 174.3 | #431 |
| 4 | InternVL3-38B | 172.1 | #405 |
| 5 | GPT-4.1 (2025-04-14) | 166.4 | #250 |
| 6 | GPT-4.1 Mini | 165.7 | #355 |
| 7 | InternVL3-14B | 165.7 | #505 |
| 8 | Gemini 2.0 Flash | 160 | #342 |
| 9 | Claude 3.7 Sonnet | 157.1 | #249 |
| 10 | Grok 2 (1212) | 156.4 | #496 |
| 11 | Moonshot v1 8K | 155.7 | #617 |
| 12 | Gemma 3 27B | 155 | #607 |
| 13 | Claude 3.5 Sonnet (20241022) | 152.1 | #291 |
| 14 | Qwen 2 VL 7B | 152.1 | #750 |
| 15 | Gemini 1.5 Pro | 150.7 | #445 |
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-commonsense-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-20.