跳转到主要内容

Qwen3.6 27B (Reasoning)

AlibabaQwen开源权重Apache 2.0 · 商用许可

描述

Qwen3.6-27B is a dense 27-billion-parameter multimodal model in the Qwen3.6 series, supporting both vision-language thinking and non-thinking modes in a single unified checkpoint. The 64-layer language model uses a hybrid layout of 16 repeats of (3 × Gated DeltaNet → FFN, 1 × Gated Attention → FFN) with hidden dim 5120 and FFN intermediate 17408 — Gated DeltaNet has 48/16 heads for V/QK (head dim 128) and Gated Attention has 24/4 heads for Q/KV (head dim 256). It supports a native 262,144-token context extensible to ~1,010,000 via YaRN and is trained with multi-token prediction. The release delivers flagship-level agentic coding, surpassing the previous-generation open-source flagship Qwen3.5-397B-A17B (397B total / 17B active) on every major coding benchmark including SWE-bench Verified (77.2), SWE-bench Pro (53.5), Terminal-Bench 2.0 (59.3), and SkillsBench (48.2), and reaches 87.8 on GPQA Diamond. Released as open weights under Apache 2.0; accessible via Qwen Studio with the Alibaba Cloud Model Studio API coming soon.

发布日期
2026-04-22
参数规模
27.8B
上下文长度
262K
支持模态
image, text, video

能力雷达图

80
general
60
coding
80
reasoning
77
science估算
60
agents
80
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137
35.0
LS
多模态榜16
65.0
LS

基准测试分数 (LLM Stats)

(LLM Stats (zeroeval))

Agents

AndroidWorld70.3%自报

Code

QwenWebBench1487.00 / 2000自报
Claw-Eval60.6%自报
ZClawBench53.4%自报
SkillsBench48.2%自报
NL2Repo36.2%自报

General

C-Eval91.4%自报

Language

MMLU-Redux93.5%自报
MMLU-Pro86.2%自报

Math

AIME 202694.1%自报
HMMT 202593.8%自报
HMMT2590.7%自报
MathVista-Mini87.4%自报
DynaMath85.6%自报
HMMT Feb 2684.3%自报
IMO-AnswerBench80.8%自报

Multimodal

VideoMMMU84.4%自报
MMMU82.9%自报

Reasoning

GPQANYU + Cohere + Anthropic (2023)87.8%自报
LiveCodeBench v683.9%自报
CharXiv-R78.4%自报
SWE-Bench Verified77.2%自报
SWE-bench Multilingual71.3%自报
SuperGPQA66.0%自报
Terminal-Bench 2.0Stanford × Laude Institute (2026)59.3%自报
SWE-Bench ProPrinceton NLP (2024)53.5%自报
Humanity's Last Exam24.0%自报

Video

MLVU86.6%自报

Vision

CountBench0.98 / 100自报
VLMsAreBlind97.0%自报
V*94.7%自报
RefCOCO-avg0.93 / 100自报
MMBench-V1.192.3%自报
OCRBench89.4%自报
VideoMME w sub.87.7%自报
EmbSpatialBench0.85 / 100自报
RealWorldQA84.1%自报
MMStar81.4%自报
CC-OCR81.2%自报
MMMU-Pro75.8%自报
MVBench75.5%自报
RefSpatialBench0.70 / 100自报
ERQA62.5%自报
SimpleVQA0.56 / 100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))
Language
90
Long Context
90
Biology
90
Legal
80
Math
80
Multimodal
80
Physics
80
Reasoning
80
Spatial Reasoning
80
Structured Output
80
Embodied
80
Finance
80
Frontend Development
80
General
80
Grounding
80
Healthcare
80
Chemistry
80
Text-to-image
80
Video
80
Vision
80
Image To Text
70
Economics
70
Code
60
Tool Calling
60
Agents
50

定价

输入价格$0.6 / 1M tokens
输出价格$3.6 / 1M tokens
混合价格(3:1)$1.04 / 1M tokens

速度

Tokens/秒55.8
首Token延迟3.53s
首回答延迟105.35s

供应商价格排行

供应商价格排行

8 个供应商

最便宜: DevPass (LLM Gateway)最贵: Novita
供应商输入输出
1DevPass (LLM Gateway)最便宜
$0.12
$1.08
2DeepInfra
$0.32
$3.2
3Venice AI
$0.325
$3.25
4EmpirioLabs AI
$0.41256
$2.47538
5Cortecs
$0.446
$3.008
6OVHcloud AI Endpoints
$0.47
$3.19
7Alibaba主要
$0.6
$3.6
8Novita
$0.6
$3.6

比较该模型在不同 API 供应商之间的定价。

外部链接