MathArena - ARXIVLEAN June: leaderboard
Metric: Accuracy (%). Source: matharena.ai. 13 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-6 (Max) | 65.22 |
| 2 | Qwen 3.8 Max | 45.65 |
| 3 | GPT-5.6 Sol | 39.13 |
| 4 | Claude Opus 5 (Max) | 32.61 |
| 5 | Muse Spark 1.3 | 28.26 |
| 6 | Claude Opus 4.8 (Max) | 23.91 |
| 7 | Gemini 3.7 Flash | 21.74 |
| 8 | Gemini 3.8 Flash | 19.57 |
| 9 | Claude Fable 5 (Max) | 15.22 |
| 10 | Gemini 3.1 Pro (Preview) | 10.87 |
| 11 | GLM-5.2 | 8.7 |
| 12 | Step 3.7 Flash | 0 |
Interactive version: theaggregate.ai/benchmark?slug=matharena-arxivlean-june · How It Works · Data refreshed daily, snapshot 2026-09-19.