OpenVLM AI2D - Solar System: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 284 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 100 |
| 2 | InternVL3-8B | 100 |
| 3 | GPT-5 | 97.2 |
| 4 | InternVL3-78B | 97.2 |
| 5 | Qwen 2 VL 72B | 97.2 |
| 6 | GPT-5 Mini (2025-08-07) | 97.2 |
| 7 | InternVL2-8B | 94.4 |
| 8 | InternVL3-38B | 94.4 |
| 9 | InternVL2.5-78B | 94.4 |
| 10 | GPT-4.1 (2025-04-14) | 94.4 |
| 11 | GPT-4o ChatGPT | 94.4 |
| 12 | Qwen 2 VL 7B | 91.7 |
| 13 | Pixtral-12B | 91.7 |
| 14 | Llama 3.2 11B Instruct | 91.7 |
| 15 | Gemma 3 12B | 91.7 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-ai2d-solar-system · How It Works · Data refreshed daily, snapshot 2026-09-05.