趋近智
技术规格
注意力结构
Multi-Head Attention
隐藏维度大小
-
层数
-
注意力头
-
键值头
-
激活函数
-
归一化
-
位置嵌入
Absolute Position Embedding
GPT-5 Pro
OpenAI's most capable GPT-5 model with advanced reasoning and comprehensive knowledge. Delivers exceptional performance across all domains including coding, mathematics, language understanding, and complex reasoning tasks. Features 400K context window and sophisticated thinking capabilities. Excels at web development (Elo 1398), mathematics (75.81), reasoning (81.69), and coding (72.11). Top-tier performance on MMLU (92.5%) and strong results across LiveBench categories.
关于 GPT-5
OpenAI's latest generation of language models featuring advanced reasoning capabilities, extended context windows up to 400K tokens, and specialized variants for coding, general intelligence, and efficiency. GPT-5 series introduces improved thinking modes, superior performance across benchmarks, and variants optimized for different use cases from high-capacity Pro models to efficient Nano models. Features native multimodal understanding, enhanced mathematical reasoning, and state-of-the-art coding abilities through Codex variants.
其他 GPT-5 模型
评估基准
排名
#1
| 基准 | 分数 | 排名 |
|---|---|---|
Summarization ProLLM Summarization | 0.98 | 🥇 1 |
General Knowledge MMLU | 0.93 | 🥇 1 |
StackUnseen ProLLM Stack Unseen | 0.88 | 🥉 3 |
Reasoning LiveBench Reasoning | 0.82 | ⭐ 5 |
Web Development WebDev Arena | 1398 | 10 |
Mathematics LiveBench Mathematics | 0.76 | 15 |
Coding LiveBench Coding | 0.72 | 26 |
排名
排名
#1 🥇
编程排名
#8