OpenVLM AI2D - Water, Carbon, and Nitrogen Cycle: leaderboard
OpenCompass OpenVLM AI2D subcategory score for diagram questions about water, carbon, and nitrogen cycles.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Around February 2027. 284 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 | 77.3 | #98 |
| 2 | InternVL3-38B | 77.3 | #405 |
| 3 | InternVL3-78B | 77.3 | #352 |
| 4 | Gemini 2.5 Pro | 75 | #152 |
| 5 | GPT-4.1 (2025-04-14) | 75 | #250 |
| 6 | InternVL3-14B | 75 | #505 |
| 7 | InternVL2.5-78B | 75 | #431 |
| 8 | Claude 3.5 Sonnet (20241022) | 72.7 | #291 |
| 9 | InternVL3-8B | 72.7 | #620 |
| 10 | Qwen 2 VL 72B | 72.7 | #460 |
| 11 | Claude 3.5 Sonnet | 72.7 | #350 |
| 12 | GPT-4.5 | 72.7 | #274 |
| 13 | GPT-4o ChatGPT | 70.5 | |
| 14 | GPT-5 Nano | 68.2 | #425 |
| 15 | InternVL2-8B | 68.2 | #817 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=openvlm-ai2d-water-carbon-and-nitrogen-cycle · How It Works · Data refreshed daily, snapshot 2026-10-20.