OpenVLM AI2D - Rock Strata — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 284 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | InternVL3-78B | 95.1 |
| 2 | Qwen 2 VL 72B | 92.7 |
| 3 | GPT-4.5 | 92.7 |
| 4 | GPT-5 Mini (2025-08-07) | 92.7 |
| 5 | Gemini 2.5 Pro | 90.2 |
| 6 | InternVL2.5-78B | 90.2 |
| 7 | GPT-5 (2025-08-07) | 90.2 |
| 8 | InternVL3-38B | 87.8 |
| 9 | GPT-4.1 (2025-04-14) | 87.8 |
| 10 | GPT-4o ChatGPT | 87.8 |
| 11 | InternVL3-14B | 85.4 |
| 12 | GPT-5 Nano | 85.4 |
| 13 | Claude 3.7 Sonnet | 82.9 |
| 14 | Gemini 1.0 Pro | 82.9 |
| 15 | InternVL3-8B | 80.5 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-ai2d-rock-strata · How the rankings work · Data refreshed daily, snapshot 2026-07-22.