Open LMM Reasoning - WeMath - RoteMemorization (Strict) — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 83 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4.1 Mini | 100 |
| 2 | InternVL2.5-2B | 81.9 |
| 3 | Llama 3.2 11B Instruct | 80 |
| 4 | Qwen 2 VL 2B | 75.4 |
| 5 | Aquila-VL-2B | 67.5 |
| 6 | InternVL2-8B | 65.1 |
| 7 | Qwen 2 VL 7B | 57.3 |
| 8 | Ovis2-8B | 55.6 |
| 9 | Gemma 3 4B | 53.7 |
| 10 | GPT-4.1 Nano | 50.6 |
| 11 | Gemma 3 12B | 42.5 |
| 12 | QVQ-72B-Preview | 40.7 |
| 13 | InternVL3-8B | 40.6 |
| 14 | Qwen 2 VL 72B | 40.5 |
| 15 | Grok 2 (1212) | 39.7 |
Interactive version: theaggregate.ai/benchmark?slug=open-lmm-reasoning-wemath-rotememorization-strict · How the rankings work · Data refreshed daily, snapshot 2026-07-22.