跳转到主要内容
AITier

Qwen3.5 2B (Reasoning)

AlibabaQwen开源权重Apache 2.0 · 商用许可

描述

Qwen3.5-2B is a 2 billion parameter vision-language model using Gated DeltaNet hybrid architecture with a 3:1 ratio of linear attention to full softmax attention. It supports 262K native context length and features both thinking and non-thinking modes.

发布日期
2026-03-02
参数规模
2.0B
上下文长度
—
支持模态
—

能力雷达图

6
general
3
coding
46
reasoning
32
science
50
agents
30
multimodal

排行榜排名

领域#排名分数来源
智能体能力模型榜104
31.0
LS
代码能力榜577
10.0
AA
通用能力榜495
28.0
AA
科学能力544
22.0
AA

基准测试分数 (LLM Stats)

(LLM Stats (zeroeval))

Agents

t2-bench48.8%自报

Chat

IFEvalGoogle Research (2023)78.6%自报
Multi-Challenge33.7%自报

General

C-Eval73.2%自报
MAXIFE60.6%自报
Include55.4%自报
NOVA-6346.4%自报

Instruction Following

IFBench41.3%自报

Language

MMLU-Redux79.6%自报
MMLU-Pro66.5%自报
MMMLU63.1%自报
MMLU-ProX52.3%自报
WMT24++45.8%自报

Long Context

LongBench v238.7%自报

Math

PolyMATH26.1%自报

Reasoning

Global PIQA69.3%自报
GPQANYU + Cohere + Anthropic (2023)51.6%自报
SuperGPQA37.5%自报
AA-LCR25.6%自报

Tool Calling

BFCL-V443.6%自报

AA 评测指数

(Artificial Analysis)
Tau2(Sierra + U Toronto + Vector Institute (2025))
69.0
Gpqa(NYU + Cohere + Anthropic (2023))
45.6
Ifbench(Google Research (2023))
31.5
Lcr(Artificial Analysis)
21.0
Intelligence Index(Artificial Analysis)
6.9
Terminalbench Hard(Stanford × Laude Institute (2026))
3.8
Terminalbench V2 1
3.0
Coding Index(Artificial Analysis)
2.9
Hle(Center for AI Safety + Scale AI (2025))
2.6

LLM Stats 分类评分

(LLM Stats (zeroeval))
Chat
60
Instruction Following
60
Language
60
Structured Output
60
Legal
50
Math
50
Physics
50
Reasoning
50
Finance
50
General
50
Healthcare
50
Agents
50
Biology
50
Tool Calling
50
Chemistry
40
Economics
40
Long Context
30
Multimodal
30
Spatial Reasoning
30
Communication
30
Vision
30

定价

输入价格免费
输出价格免费
混合价格(3:1)免费

速度

Tokens/秒0.0
首Token延迟0.00s
首回答延迟0.00s

供应商价格排行

暂无提供商数据

外部链接

LLM StatsArtificial Analysis