跳轉到主要內容
AITier

Qwen3.5 2B (Reasoning)

AlibabaQwen開源權重Apache 2.0 · 商用許可

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.

發布日期
2026-03-02
參數規模
2.0B
上下文長度
—
支援模態
—

能力雷達圖

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

排行榜排名

領域#排名分數來源
智慧體能力模型榜104
31.0
LS
程式碼能力榜577
10.0
AA
通用能力榜495
28.0
AA
科學能力544
22.0
AA

基準測試分數 (LLM Stats)

(LLM Stats (zeroeval))

Agents

t2-bench48.8%自報

Chat

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

General

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

Instruction Following

IFBench41.3%自報

Language

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

Long Context

LongBench v238.7%自報

Math

PolyMATH26.1%自報

Reasoning

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

Tool Calling

BFCL-V443.6%自報

AA 評測指數

(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 分類評分

(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

定價

輸入價格免費
輸出價格免費
混合價格(3:1)免費

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

暫無提供商資料

外部連結

LLM StatsArtificial Analysis