Qwen3 Max Thinking
AlibabaQwenProprietary
描述
The latest flagship reasoning model in the Qwen3 family. Further enhanced by multiple innovations like adaptive tool-use and advanced test-time scaling techniques
发布日期
2026-01-26
参数规模
1.0T
上下文长度
262K
支持模态
text
能力雷达图
60
general
50
coding
70
reasoning
77
science估算
40
agents
0
multimodal
缺少专门科学评测时,Science 由 LLM Stats 科学得分或推理能力估算。
排行榜排名
| 领域 | #排名 | 分数 | 来源 |
|---|---|---|---|
| 智能体能力模型榜 | 109 | 41.0 | LS |
基准测试分数 (LLM Stats)
(LLM Stats (zeroeval))Agents
VITA-Bench
40.9%自报
MCP-Mark
33.5%自报
DeepPlanning
28.7%自报
Toolathlon
18.8%自报
Chat
IFEvalGoogle Research (2023)
93.4%自报
Multi-Challenge
63.3%自报
Code
SecCodeBench
57.5%自报
General
C-Eval
93.7%自报
MAXIFE
84.0%自报
NOVA-63
54.2%自报
Instruction Following
IFBench
70.9%自报
Knowledge
HLE-Verified
37.6%自报
Language
MMLU-Redux
92.8%自报
MMLU-Pro
85.7%自报
MMMLU
84.4%自报
WMT24++
77.6%自报
Long Context
LongBench v2
60.6%自报
Math
AIME 2026
93.3%自报
IMO-AnswerBench
83.9%自报
Reasoning
GPQANYU + Cohere + Anthropic (2023)
87.4%自报
Global PIQA
86.0%自报
LiveCodeBench v6
85.9%自报
SWE-Bench Verified
75.3%自报
AA-LCR
68.7%自报
SuperGPQA
67.3%自报
SWE-bench Multilingual
66.7%自报
BrowseComp-zh
60.9%自报
BrowseCompOpenAI (2025)
53.9%自报
Humanity's Last Exam (with tools, text-only)
49.8%自报
Seal-0
46.9%自报
Humanity's Last Exam (no tools, text-only)
30.2%自报
Terminal-Bench 2.0Stanford × Laude Institute (2026)
22.5%自报
Search
WideSearch
57.9%自报
Tool Calling
BFCL-V4
67.7%自报
AA 评测指数
(Artificial Analysis)暂无 AA 评测数据
LLM Stats 分类评分
(LLM Stats (zeroeval))Language90
Biology90
Chat80
Instruction Following80
Legal80
Physics80
Structured Output80
Finance80
Frontend Development80
Healthcare80
Chemistry80
Math70
Economics70
Long Context60
Reasoning60
General60
Communication60
Search50
Code50
Knowledge40
Agents40
Tool Calling40
定价
输入价格$1.2 / 1M tokens
输出价格$6 / 1M tokens
混合价格(3:1)$2.4 / 1M tokens
速度
暂无速度数据
供应商价格排行
供应商价格排行
2 个供应商
最便宜: Alibaba最贵: DeepInfra
供应商输入输出
1Alibaba主要
$1.2
$6
2DeepInfra
$1.2
$6
比较该模型在不同 API 供应商之间的定价。