GPT-4.5
来自 OpenAI · 发布于 2024-01-01
35.9
平均分
N/A
输入价格
N/A
输出价格
N/A
上下文窗口
text
类型
Tested on 13 benchmarks with 35.9% average. Top scores: MATH level 5 (78.6%), Lech Mazur Writing (75.6%), Fiction.LiveBench (63.9%).
基准测试分数
| 基准测试 | 类别 | 分数 | Bar |
|---|---|---|---|
| MATH level 5 | math | 78.6 | |
| Lech Mazur Writing | knowledge | 75.6 | |
| Fiction.LiveBench | knowledge | 63.9 | |
| GPQA diamond | knowledge | 58.3 | |
| Aider polyglot | coding | 44.9 | |
| WeirdML | coding | 39.4 | |
| OTIS Mock AIME 2024-2025 | math | 37.7 | |
| SimpleBench | reasoning | 21.4 | |
| Cybench | coding | 17.5 | |
| VPCT | knowledge | 17.5 | |
| ARC-AGI | reasoning | 10.3 | |
| ARC-AGI-2 | reasoning | 0.8 | |
| HLE | knowledge | 0.7 |