OpenVLM MMBench V1.1 CN - Attribute Recognition: leaderboard
OpenCompass OpenVLM MMBench V1.1 Chinese split accuracy for attribute recognition.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 284 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | InternVL2.5-78B | 98.9 | #431 |
| 2 | InternVL3-38B | 96.6 | #405 |
| 3 | Grok 2 (1212) | 96.6 | #496 |
| 4 | GPT-5 Mini (2025-08-07) | 95.5 | #172 |
| 5 | GPT-5 | 94.4 | #98 |
| 6 | InternVL3-8B | 94.4 | #620 |
| 7 | InternVL3-78B | 94.4 | #352 |
| 8 | Gemini 2.5 Pro | 93.3 | #152 |
| 9 | GPT-4.1 (2025-04-14) | 93.3 | #250 |
| 10 | MiniCPM-V-2.6 | 92.1 | #836 |
| 11 | InternVL2-8B | 91 | #817 |
| 12 | Qwen 2 VL 72B | 91 | #460 |
| 13 | Gemini 1.5 Flash | 91 | #618 |
| 14 | InternVL2.5-2B | 91 | #1249 |
| 15 | GPT-4o ChatGPT | 91 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmbench-v1-1-cn-attribute-recognition · How It Works · Data refreshed daily, snapshot 2026-10-20.