跳轉到主要內容

Qwen3.6 Plus

AlibabaQwenProprietary

描述

Qwen3.6 Plus is Alibaba's next-generation flagship model featuring a 1 million token native context window, up to 65,536 output tokens, and always-on chain-of-thought reasoning. It uses a next-generation hybrid architecture optimized for efficiency and scalability. It leads on Terminal-Bench 2.0 agentic coding (61.6), surpassing Claude 4.5 Opus, and achieves strong results on document understanding (OmniDocBench 91.2) and multimodal reasoning (MMMU 86.0). Compared to Qwen 3.5, it is significantly more decisive in reasoning, using fewer tokens on straightforward tasks with better agent stability.

發布日期
2026-04-02
參數規模
上下文長度
1.0M
支援模態
image, text, video

能力雷達圖

38
general
52
coding
88
reasoning
59
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97
40.0
LS
程式碼能力榜89
74.0
AA
通用能力榜55
79.0
AA
數學推理29
84.0
LB
多模態榜2
75.0
LS
推理能力29
76.0
LB
科學能力103
70.0
AA

基準測試分數 (LLM Stats)

(LLM Stats (zeroeval))

Agents

WideSearch74.3%自報
MCP Atlas74.1%自報
TAU3-Bench70.7%自報
OSWorld-Verified62.5%自報
TIR-Bench61.6%自報
Terminal-Bench 2.0Stanford × Laude Institute (2026)61.6%自報
Claw-Eval58.7%自報
SWE-Bench ProPrinceton NLP (2024)56.6%自報
MCP-Mark48.2%自報
SkillsBench45.7%自報
VITA-Bench44.3%自報
DeepPlanning41.5%自報
Finance Agent v240.8%
Toolathlon39.8%自報
NL2Repo37.9%自報
FrontierSWE22.0%

Biology

GPQANYU + Cohere + Anthropic (2023)90.4%自報

Chemistry

SuperGPQA71.6%自報

Code

SWE-Bench Verified78.8%自報
SWE-bench Multilingual73.8%自報

Finance

MMLU-Pro88.5%自報
MMLU-ProX84.7%自報

General

MMLU-Redux94.5%自報
IFEvalGoogle Research (2023)94.3%自報
C-Eval93.3%自報
Global PIQA89.8%自報
MMMLU89.5%自報
MAXIFE88.2%自報
LiveCodeBench v687.1%自報
MMMU86.0%自報
Include85.1%自報
MMStar83.3%自報
MMMU-Pro78.8%自報
IFBench74.2%自報
LiveBench70.9%
SimpleVQA0.67 / 100自報
LongBench v262.0%自報
NOVA-6357.9%自報

Grounding

RefCOCO-avg0.94 / 100自報
ScreenSpot Pro68.2%自報

Healthcare

VideoMMMU84.0%自報

Language

WMT24++84.3%自報

Long Context

MLVU86.7%自報
AA-LCR68.3%自報
MMLongBench-Doc0.62 / 100自報

Math

HMMT 202596.7%自報
AIME 202695.3%自報
HMMT2594.6%自報
We-Math89.0%自報
DynaMath88.0%自報
MathVision88.0%自報
HMMT Feb 2687.8%自報
IMO-AnswerBench83.8%自報
PolyMATH77.4%自報
Humanity's Last Exam28.8%自報

Multimodal

V*96.9%自報
AI2D94.4%自報
OmniDocBench 1.591.2%自報
Video-MME84.2%自報
CC-OCR83.4%自報
CharXiv-R81.5%自報

Reasoning

CountBench0.98 / 100自報
ERQA65.7%自報

Spatial Reasoning

RealWorldQA85.4%自報

Vision

ODinW51.8%自報

AA 評測指數

(Artificial Analysis)
Coding Index(Artificial Analysis)
54.5
Intelligence Index(Artificial Analysis)
40.5
Tau2(Sierra + U Toronto + Vector Institute (2025))
1.0
Gpqa(NYU + Cohere + Anthropic (2023))
0.9
Ifbench(Google Research (2023))
0.8
Lcr(Artificial Analysis)
0.7
Terminalbench V2 1
0.6
Terminalbench Hard(Stanford × Laude Institute (2026))
0.4
Scicode(UIUC + Argonne National Lab (2024))
0.4
Hle(Center for AI Safety + Scale AI (2025))
0.3
Tau Banking
0.2

LLM Stats 分類評分

(LLM Stats (zeroeval))
Language
90
Biology
90
Video
90
Legal
80
Math
80
Multimodal
80
Physics
80
Reasoning
80
Spatial Reasoning
80
Structured Output
80
Instruction Following
80
Frontend Development
80
Grounding
80
Healthcare
80
Chemistry
80
Text-to-image
80
Vision
80
Long Context
70
Search
70
Image To Text
70
Finance
70
General
70
Economics
70
Code
60
Tool Calling
60
Agents
50

定價

輸入價格$0.5 / 1M tokens
輸出價格$3 / 1M tokens
混合價格(3:1)$1.125 / 1M tokens
快取讀取價格$0.05 / 1M tokens
快取寫入價格$0.625 / 1M tokens

速度

Tokens/秒0.0
首Token延遲0.00s
首回答延遲0.00s

供應商價格排行

供應商價格排行

17 個供應商

最便宜: Together最貴: Charm Hyper
供應商輸入輸出
1Together最便宜
$0
$0
2Novita
$0
$0
3Merge Gateway
$0.276
$1.651
4AIHubMix
$0.28
$1.69
5CrossModel
$0.32
$1.88
6OpenRouter
$0.325
$1.95
7Kilo Gateway
$0.325
$1.95
8Pioneer
$0.325
$1.95
9Alibaba主要
$0.5
$3
10OpenCode Go
$0.5
$3
11Alibaba (China)
$0.5
$3
12ZenMux
$0.5
$3
13OpenCode Zen
$0.5
$3
14LLM Gateway
$0.5
$3
15OrcaRouter
$0.5
$3
16EmpirioLabs AI
$0.5
$3
17Charm Hyper
$2
$6

比較該模型在不同 API 供應商之間的定價。

外部連結