跳轉到主要內容

Qwen3.5 0.8B (Non-reasoning)

AlibabaQwen開源權重Apache 2.0 · 商用許可

描述

Qwen3.5-0.8B is a 0.8 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
800M
上下文長度
支援模態

能力雷達圖

3
general
1
coding
24
reasoning
13
science
20
agents
10
multimodal

排行榜排名

領域#排名分數來源
智慧體能力模型榜144
27.0
LS
程式碼能力榜539
4.0
AA
通用能力榜505
21.0
AA
科學能力567
9.0
AA

基準測試分數 (LLM Stats)

(LLM Stats (zeroeval))

Agents

BFCL-V425.3%自報
t2-bench11.6%自報

Biology

GPQANYU + Cohere + Anthropic (2023)11.9%自報

Chemistry

SuperGPQA21.3%自報

Communication

Multi-Challenge18.9%自報

Finance

MMLU-Pro42.3%自報
MMLU-ProX34.6%自報

General

MMLU-Redux59.5%自報
Global PIQA59.4%自報
C-Eval50.5%自報
MMMLU44.3%自報
IFEvalGoogle Research (2023)44.0%自報
NOVA-6342.4%自報
Include40.6%自報
MAXIFE39.2%自報
LongBench v226.1%自報
IFBench21.0%自報

Language

WMT24++27.2%自報

Long Context

AA-LCR4.7%自報

Math

PolyMATH8.2%自報

AA 評測指數

(Artificial Analysis)
Intelligence Index(Artificial Analysis)
2.9
Coding Index(Artificial Analysis)
1.2
Tau2(Sierra + U Toronto + Vector Institute (2025))
0.7
Gpqa(NYU + Cohere + Anthropic (2023))
0.2
Ifbench(Google Research (2023))
0.2
Lcr(Artificial Analysis)
0.1
Hle(Center for AI Safety + Scale AI (2025))
0.1
Scicode(UIUC + Argonne National Lab (2024))
0.0
Terminalbench V2 1
0.0
Terminalbench Hard(Stanford × Laude Institute (2026))
0.0

LLM Stats 分類評分

(LLM Stats (zeroeval))
Math
40
Structured Output
40
Language
40
Legal
30
Physics
30
Reasoning
30
Instruction Following
30
Finance
30
General
30
Healthcare
30
Long Context
20
Agents
20
Chemistry
20
Communication
20
Economics
20
Tool Calling
20
Multimodal
10
Spatial Reasoning
10
Biology
10
Vision
10

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