VMMU - Biology: leaderboard
Metric: Accuracy (%) on the VMMU Biology questions (Vietnamese page-image multiple choice); higher is better. Source: arxiv.org. Saturation forecast: Around December 2026. 17 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro | 80.94 |
| 2 | GPT-5 (Medium) | 72.73 |
| 3 | Claude Opus 4.6 (Thinking) | 71.26 |
| 4 | O3 (Medium) | 67.16 |
| 5 | Gemini 2.5 Flash | 63.05 |
| 6 | GPT-4.1 | 44.87 |
| 7 | Claude Sonnet 4 | 44.87 |
| 8 | Qwen 2.5 VL 32B Instruct | 42.82 |
| 9 | Qwen 2.5 VL 72B Instruct | 40.47 |
| 10 | Llama 4 Scout | 38.71 |
| 11 | Gemma 3 27B | 30.79 |
| 12 | Mistral Medium 3 | 29.33 |
| 13 | Mistral Small 3.2 | 28.15 |
| 14 | Llama 4 Maverick | 27.57 |
| 15 | aya-vision-8B | 24.93 |
Interactive version: theaggregate.ai/benchmark?slug=vmmu-biology · How It Works · Data refreshed daily, snapshot 2026-09-29.