OpenVLM MMBench V1.1 CN - Social Relation: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 284 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | InternVL2.5-78B | 91.2 |
| 2 | InternVL2-8B | 90.1 |
| 3 | InternVL3-8B | 89 |
| 4 | Qwen 2 VL 72B | 89 |
| 5 | Qwen 2 VL 7B | 89 |
| 6 | InternVL3-38B | 89 |
| 7 | Gemini 1.5 Flash (002) | 89 |
| 8 | Gemma 3 12B | 89 |
| 9 | GPT-4o ChatGPT | 89 |
| 10 | Gemma 3 4B | 87.9 |
| 11 | GPT-4.1 (2025-04-14) | 87.9 |
| 12 | Gemini 2.5 Pro | 86.8 |
| 13 | GPT-5 | 86.8 |
| 14 | Llama 3.2 90B Vision Instruct | 86.8 |
| 15 | Aquila-VL-2B | 85.7 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmbench-v1-1-cn-social-relation · How It Works · Data refreshed daily, snapshot 2026-09-05.