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
基準測試分數 (LLM Stats)
(LLM Stats (zeroeval))Agents
BFCL-V4
25.3%自報
t2-bench
11.6%自報
Biology
GPQANYU + Cohere + Anthropic (2023)
11.9%自報
Chemistry
SuperGPQA
21.3%自報
Communication
Multi-Challenge
18.9%自報
Finance
MMLU-Pro
42.3%自報
MMLU-ProX
34.6%自報
General
MMLU-Redux
59.5%自報
Global PIQA
59.4%自報
C-Eval
50.5%自報
MMMLU
44.3%自報
IFEvalGoogle Research (2023)
44.0%自報
NOVA-63
42.4%自報
Include
40.6%自報
MAXIFE
39.2%自報
LongBench v2
26.1%自報
IFBench
21.0%自報
Language
WMT24++
27.2%自報
Long Context
AA-LCR
4.7%自報
Math
PolyMATH
8.2%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 10.0
Terminalbench Hard(Stanford × Laude Institute (2026))0.0
LLM Stats 分類評分
(LLM Stats (zeroeval))Math40
Structured Output40
Language40
Legal30
Physics30
Reasoning30
Instruction Following30
Finance30
General30
Healthcare30
Long Context20
Agents20
Chemistry20
Communication20
Economics20
Tool Calling20
Multimodal10
Spatial Reasoning10
Biology10
Vision10
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