Views
No views yet
| 测评集 | 测评指标 | openPangu-2.0-Pro-Thinking | openPangu-2.0-Pro-Non-Thinking |
|---|---|---|---|
| 通用能力 | |||
| CL-Bench | Acc | 19.7 | 17.8 |
| IFEval | Prompt Strict | 94.5 | 90.4 |
| IFBench | Prompt Strict | 79.9 | 59.5 |
| AgentIF | (CSR+ISR)/2 | 44.4 | 42.2 |
| SysBench | ISR | 90.8 | 82.6 |
| Multichallenge | Acc | 62.9 | 52.2 |
| SimpleQA Verified | Acc | 34.9 | 27.1 |
| Chinese-SimpleQA | Acc | 74.2 | 75.7 |
| 推理能力 | |||
| HMMT Feb 2026 | Avg@16 | 86.2 | 63.3 |
| - w/ Python | Avg@16 | 93.2 | - |
| AIME 2026 | Avg@16 | 95.4 | 87.3 |
| - w/ Python | Avg@16 | 97.9 | - |
| Apex | Avg@16 | 17.7 | 2.1 |
| IMO-AnswerBench | Acc | 84.3 | 60.8 |
| BBEH | Harmonic Mean | 69.7 | 55.0 |
| GPQA-Diamond | Avg@4 | 87.9 | 81.1 |
| HLE | Acc | 27.1 | 9.0 |
| Agent能力 | |||
| TAU2-Bench | Avg@3 | 81.1 | 71.6 |
| MCP-Atlas | Acc | 61.3 | 48.8 |
| SkillsBench | Avg@5 | 48.9 | 47.5 |
| PinchBench | Avg@3 | 84.0 | 88.9 |
| WildClawBench | Avg@3 | 46.5 | 43.1 |
| Claw-Eval | Pass^3 | 61.7 | 49.4 |
| GDPVal | Acc | 74.9 | 51.2 |
| BrowseComp | Acc | 65.7 | - |
| 代码能力 | |||
| LiveCodeBench V6 | Avg@3 | 85.7 | 74.5 |
| FeatBench | Avg@3 | 47.1 | 48.4 |
| DeepCodeBench | Avg@3 | 74.2 | 72.6 |
| SWE-bench Verified | Avg@3 | 68.5 | 66.8 |