| # | 模型 | 编程 | 知识 | 综合 | 稳定性 |
|---|---|---|---|---|---|
| 🥇 | DeepSeek V3 DeepSeek | 75.6 | 80.0 | 83.1 | |
| 🥈 | Claude Opus 4.6 Anthropic | 93.3 | 100.0 | 81.1 | |
| 🥉 | GPT-4o OpenAI | 87.8 | 93.3 | 84.0 | |
| 4 | GPT-o3 OpenAI | 86.7 | 86.7 | 75.0 | |
| 5 | Qwen Max Alibaba | 93.3 | 93.3 | 86.9 | |
| 6 | Claude Sonnet 4.6 Anthropic | 86.7 | 91.7 | 81.7 | |
| 7 | DeepSeek R1 DeepSeek | 87.8 | 93.3 | 87.6 | |
| 8 | Gemini 2.5 Pro Google | 100.0 | 63.3 | 74.7 |