跳轉到主要內容
AITier

Qwen3.8 Max (0902)

AlibabaQwen開源權重Qwen3.8-Max License · 商用許可

描述

Qwen3.8 Max, released as the Qwen3.8-2.4T-A95B open-weight checkpoint, is Qwen's flagship mixture-of-experts model for coding, research, professional workflows, and long-horizon agents. It has 2.4 trillion total parameters with 95 billion active parameters, a native 262,144-token context window extensible to about 1 million tokens, and always-on thinking. QwenCloud serves the same model in the Qwen3.8 Max API family with text and image input.

發布日期
2026-09-02
參數規模
2.4T
上下文長度
1.0M
支援模態
image, pdf, text, video

能力雷達圖

45
general
72
coding
93
reasoning
69
science
60
agents
90
multimodal

排行榜排名

領域#排名分數來源
智慧體能力模型榜18
61.0
LS
程式碼能力榜43
92.0
AA
通用能力榜36
78.0
AA
數學推理31
91.0
LB
多模態榜1
84.0
LS
推理能力24
88.0
LB
科學能力62
80.0
AA

基準測試分數 (LLM Stats)

(LLM Stats (zeroeval))

Agents

QwenSVG1713.00 / 2000自報
AndroidWorld85.3%自報
MobileWorld77.8%自報
AndroidBench75.1%自報
CoWorkBench74.8%自報
Toolathlon72.5%自報
Workspace Bench67.7%自報
Job Bench53.4%自報
OneMillion Bench52.5%自報
Agents' Last Exam52.4%自報
MLS-Bench Lite41.0%自報
AutomationBench27.3%自報

Code

QwenReactBench1724.00 / 2000自報
PaperBench93.0%自報
QwenSWEBench80.7%自報
FrontierSWE73.5%自報
SkillsBench70.2%自報
Vision2Web69.0%自報
QwenQoderBench58.4%自報
DeepSWE 1.156.6%自報
NL2Repo55.9%自報

Healthcare

HealthBench60.2%自報

Instruction Following

IFBench82.8%自報

Knowledge

PLawBench73.2%自報
PRBench-Finance58.3%自報
PRBench-Legal57.6%自報

Long Context

MRCR v2 (8-needle)92.9%自報
LongBench v266.3%自報

Multimodal

OSWorld-Verified86.1%自報

Reasoning

GPQANYU + Cohere + Anthropic (2023)92.6%自報
Terminal-Bench 2.186.6%自報
SWE-Bench ProPrinceton NLP (2024)67.7%自報
Humanity's Last Exam (with tools, text-only)56.2%自報
Humanity's Last Exam43.6%自報

Search

WideSearch81.9%自報

Vision

VideoMME w sub.90.4%自報
RealWorldQA88.0%自報
ScreenSpot Pro84.5%自報
MMMU-Pro82.3%自報
LVBench81.8%自報
ERQA77.8%自報
PerceptionBench63.5%自報

AA 評測指數

(Artificial Analysis)
Gpqa(NYU + Cohere + Anthropic (2023))
92.8
Terminalbench V2 1
88.8
Lcr(Artificial Analysis)
80.3
Coding Index(Artificial Analysis)
76.2
Scicode(UIUC + Argonne National Lab (2024))
52.1
Tau Banking
47.8
Intelligence Index(Artificial Analysis)
45.4
Hle(Center for AI Safety + Scale AI (2025))
43.1
Terminalbench V4 0
38.9

LLM Stats 分類評分

(LLM Stats (zeroeval))
Physics
90
Biology
90
Chemistry
90
Video
90
Instruction Following
80
Multimodal
80
Search
80
Spatial Reasoning
80
Grounding
80
Vision
80
Long Context
70
Reasoning
70
Structured Output
70
General
70
Agents
70
Code
70
Productivity
60
Healthcare
60
Tool Calling
60
Math
50

定價

輸入價格$2 / 1M tokens
輸出價格$6 / 1M tokens
混合價格(3:1)$3 / 1M tokens
快取讀取價格$0.25 / 1M tokens
快取寫入價格$2.5 / 1M tokens

速度

Tokens/秒39.8
首Token延遲1.81s
首回答延遲52.03s

供應商價格排行

供應商價格排行

6 個供應商

最便宜: DeepInfra最貴: EmpirioLabs AI
供應商輸入輸出
1DeepInfra最便宜
$0
$0
2Novita
$0
$0.00001
3Fireworks
$0
$0.00001
4Together
$0
$0.00001
5Alibaba主要
$2
$6
6EmpirioLabs AI
$2
$6

比較該模型在不同 API 供應商之間的定價。

外部連結

LLM StatsArtificial Analysis