Open LMM Reasoning - MMMath - Difficulty-hard — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | InternVL2-8B | 9 |
| 2 | Qwen 2 VL 72B | 8.6 |
| 3 | Qwen 2 VL 7B | 6 |
| 4 | Qwen 2 VL 2B | 3.3 |
| 5 | Llama 3.2 11B Instruct | 2.5 |
Interactive version: theaggregate.ai/benchmark?slug=open-lmm-reasoning-mmmath-difficulty-hard · How the rankings work · Data refreshed daily, snapshot 2026-07-22.