Advanced Reasoning on ruMATH-500
97.2AccuracyDeepSeek-R1
Evaluation Results
| Method | Links | |
|---|---|---|
| DeepSeek-R1Model Category=Open Source Larger Scale & Proprietary Models2025.12 | 97.2 | |
| o4-mini (medium)Model Category=Open Source Larger Scale & Proprietary Models2025.12 | 95.8 | |
| T-pro 2.0Model Category=Open Source Models (27B-32B class)2025.12 | 94 | |
| Qwen3-32BModel Category=Open Source Models (27B-32B class)2025.12 | 93.8 | |
| DeepSeek-R1-Distill-Qwen-32BModel Category=Open Source Models (27B-32B class)2025.12 | 89.8 | |
| DeepSeek-V3Model Category=Open Source Larger Scale & Proprietary Models2025.12 | 88.2 | |
| Gemma 3 27BModel Category=Open Source Models (27B-32B class)2025.12 | 86 | |
| GPT-4oModel Category=Open Source Larger Scale & Proprietary Models2025.12 | 76.6 | |
| GigaChat 2 MaxModel Category=Open Source Larger Scale & Proprietary Models2025.12 | 70.2 | |
| YandexGPT5-ProModel Category=Open Source Larger Scale & Proprietary Models2025.12 | 68.2 | |
| RuadaptQwen3-32B-InstructModel Category=Open Source Models (27B-32B class)2025.12 | 45 |