模型对比
选择模型进行全方位并排比较。
| 指标 | GPT-4o OpenAI | Claude 3.5 Sonnet Anthropic | Gemini 2.5 Pro Google |
|---|---|---|---|
| Composite Score | 89.2 | 88.4 | 97.5 |
| Arena ELO | 1287 | 1272 | 1398 |
| MMLU | 88.7% | 88.7% | 92% |
| GPQA | 53.6% | 65% | 84% |
| HumanEval | 90.2% | 92% | 93% |
| SWE-Bench | 38.4% | 49% | 63.8% |
| 速度 (t/s) | 109.5 | 82.3 | 85 |
| 首字时间 (ms) | 320ms | 380ms | 450ms |
| 输出价格 ($/M) | $10.00 | $15.00 | $10.00 |
| 许可证 | Proprietary | Proprietary | Proprietary |