Mathematical Reasoning on AIME 2025 (avg@8)
85.6Avg@8Nanbeige4-3B-Thinking
Evaluation results
| Method | Links | |
|---|---|---|
| Nanbeige4-3B-ThinkingRelease Identifier=25112025.12 | 85.6 | |
| Qwen3-30A3-2507Release Identifier=25072025.12 | 85 | |
| Qwen3-4B-2507Release Identifier=25072025.12 | 81.3 | |
| Qwen3-32B-2504Release Identifier=25042025.12 | 72.9 | |
| Qwen3-14B-2504Release Identifier=25042025.12 | 70.4 | |
| Qwen3-8B-2504Release Identifier=25042025.12 | 67.3 |