跳轉到主要內容

GPT-5.4 (low)

OpenAIGPTProprietary

描述

GPT-5.4 is OpenAI's most capable and efficient frontier model for professional work. It combines industry-leading coding capabilities with native computer-use, up to 1M tokens of context, full-resolution vision processing, tool search for large tool ecosystems, and improved reasoning across spreadsheets, presentations, and documents. It is the most token-efficient reasoning model in the GPT-5 series.

發布日期
2026-03-05
參數規模
上下文長度
1.1M
支援模態
file, image, text

能力雷達圖

44
general
46
coding
87
reasoning
63
science估算
70
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

領域#排名分數來源
智能体与工具18
68.0
LS
代码能力榜27
79.0
AA
通用能力榜57
77.0
AA
多模态榜10
89.0
LS
推理能力15
87.0
LS
科学能力23
80.0
AA

基準測試分數 (LLM Stats)

Agents

BrowseComp82.7%自報
Terminal-Bench 2.075.1%自報
OSWorld-Verified75.0%自報
MCP Atlas67.2%自報
SWE-Bench Pro57.7%自報
Finance Agent56.0%自報
Toolathlon54.6%自報

Biology

GPQA92.8%自報

Communication

Tau2 Telecom98.9%自報

General

MMMU-Pro81.2%自報

Long Context

Graphwalks parents >128k32.4%自報
Graphwalks BFS >128k21.4%自報

Math

FrontierMath47.6%自報
Humanity's Last Exam39.8%自報

Multimodal

OmniDocBench 1.589.1%自報

Reasoning

ARC-AGI93.7%自報
Graphwalks BFS <128k93.0%自報
Graphwalks parents <128k89.8%自報
ARC-AGI v273.3%自報

AA 評測指數

Intelligence Index
47.9
Coding Index
45.6
Gpqa
0.9
Tau2
0.7
Lcr
0.7
Ifbench
0.7
Scicode
0.5
Terminalbench Hard
0.4
Hle
0.3

LLM Stats 分類評分

Communication
100
Structured Output
90
Biology
90
Chemistry
90
Physics
90
Vision
80
General
80
Multimodal
80
Search
80
Spatial Reasoning
70
Tool Calling
70
Agents
70
Code
70
Reasoning
70
Finance
60
Long Context
40
Math
40
Healthcare
30

定價

輸入價格$2.5 / 1M tokens
輸出價格$15 / 1M tokens
混合價格(3:1)$5.625 / 1M tokens

速度

Tokens/秒67.1 tokens/s
首Token延遲1.85s
首回答延遲1.85s

可用提供商

(LS 內部計價單位)
提供商輸入價格輸出價格
OpenAI2.5M15.0M

外部連結