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
8
general
17
coding
46
reasoning
22
science估算
50
agents
30
multimodal
Science 在缺少專門科學評測時使用推理能力代理估算。
排行榜排名
基準測試分數 (LLM Stats)
Agents
t2-bench
48.8%自報
BFCL-V4
43.6%自報
Biology
GPQA
51.6%自報
Chemistry
SuperGPQA
37.5%自報
Communication
Multi-Challenge
33.7%自報
Finance
MMLU-Pro
66.5%自報
MMLU-ProX
52.3%自報
General
MMLU-Redux
79.6%自報
IFEval
78.6%自報
C-Eval
73.2%自報
Global PIQA
69.3%自報
MMMLU
63.1%自報
MAXIFE
60.6%自報
Include
55.4%自報
NOVA-63
46.4%自報
IFBench
41.3%自報
LongBench v2
38.7%自報
Language
WMT24++
45.8%自報
Long Context
AA-LCR
25.6%自報
Math
PolyMATH
26.1%自報
AA 評測指數
Coding Index19.7
Intelligence Index10.2
Tau20.7
Gpqa0.5
Ifbench0.3
Terminalbench V2 10.3
Lcr0.2
Tau Banking0.1
Terminalbench Hard0.0
Scicode0.0
Hle0.0
LLM Stats 分類評分
Instruction Following60
Language60
Structured Output60
General60
Legal50
Math50
Physics50
Reasoning50
Finance50
Healthcare50
Agents50
Biology50
Tool Calling50
Chemistry40
Economics40
Long Context30
Multimodal30
Spatial Reasoning30
Communication30
Vision30
定價
輸入價格$0.02 / 1M tokens
輸出價格$0.1 / 1M tokens
混合價格(3:1)$0.04 / 1M tokens
速度
Tokens/秒35.6
首Token延遲0.48s
首回答延遲56.73s
供應商價格排行
供應商價格排行
1 個供應商
供應商輸入輸出
1Alibaba主要
$0.02
$0.1
比較該模型在不同 API 供應商之間的定價。