Open LMM Reasoning - WeMath - InsufficientKnowledge (Loose) — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 83 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4.1 Mini | 97.7 |
| 2 | Llama 3.2 11B Instruct | 80 |
| 3 | Qwen 2 VL 2B | 65.9 |
| 4 | InternVL2.5-2B | 63.8 |
| 5 | Aquila-VL-2B | 54.9 |
| 6 | Qwen 2 VL 7B | 50.1 |
| 7 | InternVL2-8B | 47 |
| 8 | Gemma 3 4B | 43.6 |
| 9 | Ovis2-8B | 40.4 |
| 10 | Gemma 3 12B | 39.6 |
| 11 | Qwen 2 VL 72B | 37.3 |
| 12 | GPT-4.1 Nano | 36 |
| 13 | Grok 2 (1212) | 34.7 |
| 14 | InternVL2.5-78B | 32.8 |
| 15 | QVQ-72B-Preview | 32.8 |
Interactive version: theaggregate.ai/benchmark?slug=open-lmm-reasoning-wemath-insufficientknowledge-loose · How the rankings work · Data refreshed daily, snapshot 2026-07-22.