Multilingual Mathematical Reasoning on MGSM (native CoT avg)
87.73Native CoT Avg ScoreNemotron-3-Ultra 550B-A55B-Base
Evaluation Results
| Method | Links | |
|---|---|---|
| Nemotron-3-Ultra 550B-A55B-Baseshots=8-shot, reasoning=native CoT2026.06 | 87.73 | |
| Kimi-K2 Baseshots=8-shot, reasoning=native CoT2026.06 | 85.2 | |
| Mistral-Large-3 675B-Base-2512shots=8-shot, reasoning=native CoT2026.06 | 82.93 | |
| DeepSeek-V3.2 Exp-Baseshots=8-shot, reasoning=native CoT2026.06 | 82.33 | |
| GLM-4.5 Baseshots=8-shot, reasoning=native CoT2026.06 | 81.27 |