GPT-4.1 mini
OpenAIGPTProprietary
描述
GPT-4.1 mini provides a balance between intelligence, speed, and cost. It's a significant leap in small model performance, even beating GPT-4o in many benchmarks while reducing latency and cost.
發布日期
2025-04-14
參數規模
—
上下文長度
1.0M
支援模態
file, image, text
能力雷達圖
37
general
31
coding
54
reasoning
45
science估算
50
agents
85
multimodal
Science 在缺少專門科學評測時使用推理能力代理估算。
排行榜排名
基準測試分數 (LLM Stats)
Biology
GPQA
65.0%自報
Code
Aider-Polyglot
34.7%自報
Aider-Polyglot Edit
31.6%自報
SWE-Bench Verified
23.6%自報
Communication
Multi-IF
67.0%自報
TAU-bench Retail
55.8%自報
TAU-bench Airline
36.0%自報
Multi-Challenge
35.8%自報
Finance
MMLU
87.5%自報
General
IFEval
84.1%自報
MMMLU
78.5%自報
MMMU
72.7%自報
Internal API instruction following (hard)
45.1%自報
Language
COLLIE
54.6%自報
Long Context
ComplexFuncBench
49.3%自報
OpenAI-MRCR: 2 needle 128k
47.2%自報
OpenAI-MRCR: 2 needle 1M
33.3%自報
Graphwalks BFS >128k
15.0%自報
Graphwalks parents >128k
11.0%自報
Math
MathVista
73.1%自報
AIME 2024
49.6%自報
AIME 2025
40.2%自報
HMMT 2025
35.0%自報
Humanity's Last Exam
3.7%自報
Multimodal
CharXiv-D
88.4%自報
CharXiv-R
56.8%自報
Reasoning
Graphwalks BFS <128k
61.7%自報
Graphwalks parents <128k
60.5%自報
AA 評測指數
Math Index46.3
Intelligence Index22.9
Coding Index18.5
Math 5000.9
Mmlu Pro0.8
Gpqa0.7
Tau20.5
Livecodebench0.5
Aime 250.5
Aime0.4
Lcr0.4
Scicode0.4
Ifbench0.4
Terminalbench Hard0.1
Hle0.0
LLM Stats 分類評分
Finance90
Legal90
Healthcare80
Instruction Following80
Structured Output70
Biology70
Chemistry70
Language70
Multimodal70
Physics70
Vision60
General60
Tool Calling50
Writing50
Communication50
Math50
Reasoning50
Spatial Reasoning40
Code30
Long Context30
Frontend Development20
定價
輸入價格$0.4 / 1M tokens
輸出價格$1.6 / 1M tokens
混合價格(3:1)$0.7 / 1M tokens
速度
Tokens/秒78.8 tokens/s
首Token延遲0.52s
首回答延遲0.52s
可用提供商
(LS 內部計價單位)| 提供商 | 輸入價格 | 輸出價格 |
|---|---|---|
| OpenAI | 400K | 1.6M |