OpenVLM MMBench V1.1 CN - Overall: leaderboard
OpenCompass OpenVLM MMBench V1.1 Chinese split accuracy for overall evaluation.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 284 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 88.1 | #152 |
| 2 | InternVL3-78B | 87.9 | #352 |
| 3 | InternVL2.5-78B | 87.8 | #431 |
| 4 | InternVL3-38B | 86.6 | #405 |
| 5 | GPT-5 | 86.4 | #98 |
| 6 | GPT-4.1 (2025-04-14) | 86.2 | #250 |
| 7 | GPT-5 Mini (2025-08-07) | 86.1 | #172 |
| 8 | Qwen 2 VL 72B | 85.8 | #460 |
| 9 | GPT-4o ChatGPT | 85.2 | |
| 10 | InternVL3-14B | 83.3 | #505 |
| 11 | GPT-4.5 | 83.3 | #274 |
| 12 | Claude 3.5 Sonnet (20241022) | 82.5 | #291 |
| 13 | Grok 2 (1212) | 82.5 | #496 |
| 14 | Qwen 2.5 VL 7B | 82.1 | #643 |
| 15 | InternVL3-8B | 81.9 | #620 |
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-overall · How It Works · Data refreshed daily, snapshot 2026-10-20.