LLM Stats (DynaMath): leaderboard
Dynamic mathematics benchmark testing robust problem solving on generated and refreshed math questions designed to reduce memorization.
Metric: Score (%). Source: llm-stats.com. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.6 Plus | 88 | #120 |
| 2 | Qwen 3.5 27B | 87.7 | #187 |
| 3 | Qwen 3.5 122B A10B | 85.9 | #173 |
| 4 | Qwen 3.6 27B | 85.6 | #170 |
| 5 | Qwen 3.5 35B A3B | 85 | #256 |
| 6 | Seed 2.1 Pro | 73.1 | #67 |
| 7 | Seed 2.1 Turbo | 68.1 | #86 |
| 8 | Seed 1.8 | 61.5 | #139 |
| 9 | Seed 2.0 Mini | 58.9 | #276 |
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-dynamath · How It Works · Data refreshed daily, snapshot 2026-10-20.