MATH-MC Level 4 — leaderboard
Metric: Accuracy (%). Source: github.com. 69 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 99.58 |
| 2 | Gemma 4 31B (IT) | 99.58 |
| 3 | Kimi K2.5 | 99.58 |
| 4 | Qwen 3.5 122B A10B | 99.58 |
| 5 | MiMo-V2-Flash | 99.58 |
| 6 | GLM-5.1 FP8 | 99.58 |
| 7 | Qwen3-VL-235B-A22B-Thinking-FP8 | 99.58 |
| 8 | GPT-5.5 | 99.5 |
| 9 | Claude Opus 4.6 | 99.5 |
| 10 | NVIDIA-Nemotron-3-Super-120B-A12B-FP8 | 99.5 |
| 11 | Qwen 3 Next 80B A3B (Thinking) | 99.42 |
| 12 | Qwen 3.5 35B A3B | 99.42 |
| 13 | Step 3.5 Flash | 99.42 |
| 14 | DeepSeek V3.2 Speciale | 99.42 |
| 15 | GLM-5 FP8 | 99.42 |
Interactive version: theaggregate.ai/benchmark?slug=math-mc-level-4 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.