Open LMM Reasoning - DynaMath - Subject-algebra: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 100 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4.1 Mini | 82.4 |
| 2 | Gemini 2.0 Flash | 80.4 |
| 3 | Doubao-1.5-Pro | 78.4 |
| 4 | GPT-4o ChatGPT | 76.5 |
| 5 | GPT-4.1 (2025-04-14) | 72.5 |
| 6 | Claude 3.7 Sonnet | 70.6 |
| 7 | InternVL3-38B | 68.6 |
| 8 | InternVL3-78B | 66.7 |
| 9 | Claude 3.5 Sonnet (20241022) | 62.7 |
| 10 | Gemini 1.5 Pro (002) | 58.8 |
| 11 | QVQ-72B-Preview | 58.8 |
| 12 | Gemma 3 27B | 54.9 |
| 13 | InternVL2.5-78B-CoT | 51 |
| 14 | InternVL3-8B | 45.1 |
| 15 | Grok 2 (1212) | 45.1 |
Interactive version: theaggregate.ai/benchmark?slug=open-lmm-reasoning-dynamath-subject-algebra · How It Works · Data refreshed daily, snapshot 2026-09-05.