Qwen3 235B A22B 2507 Instruct
AlibabaQwenOpen WeightApache 2.0 · Commercial OK
描述
Qwen3-235B-A22B-Instruct-2507 is the updated instruct version of Qwen3-235B-A22B featuring significant improvements in general capabilities including instruction following, logical reasoning, text comprehension, mathematics, science, coding and tool usage. It provides substantial gains in long-tail knowledge coverage across multiple languages and markedly better alignment with user preferences in subjective and open-ended tasks.
發布日期
2025-07-21
參數規模
235.0B
上下文長度
262K
支援模態
text
能力雷達圖
40
general
34
coding
76
reasoning
49
science估算
60
agents
0
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
BFCL-v3
70.9%自報
Biology
GPQA
77.5%自報
Chemistry
SuperGPQA
62.6%自報
Code
Aider-Polyglot
57.3%自報
Communication
WritingBench
85.2%自報
Multi-IF
77.5%自報
Tau2 Retail
71.3%自報
Tau2 Airline
44.0%自報
Creativity
Creative Writing v3
87.5%自報
Arena-Hard v2
79.2%自報
Factuality
SimpleQA
54.3%自報
Finance
MMLU-Pro
83.0%自報
MMLU-ProX
79.4%自報
General
MMLU-Redux
93.1%自報
IFEval
88.7%自報
MultiPL-E
87.9%自報
CSimpleQA
84.3%自報
Include
79.5%自報
LiveBench 20241125
75.4%自報
LiveCodeBench v6
51.8%自報
Math
AIME 2025
70.3%自報
HMMT25
55.4%自報
PolyMATH
50.2%自報
Reasoning
ZebraLogic
95.0%自報
ARC-AGI
41.8%自報
AA 評測指數
Math Index71.7
Intelligence Index25.0
Coding Index22.1
Math 5001.0
Mmlu Pro0.8
Gpqa0.8
Aime0.7
Aime 250.7
Livecodebench0.5
Ifbench0.5
Scicode0.4
Tau20.3
Lcr0.3
Terminalbench Hard0.2
Hle0.1
LLM Stats 分類評分
Structured Output80
Writing80
Biology80
Creativity80
Finance80
Healthcare80
Instruction Following80
Language80
Legal80
Agents70
Chemistry70
Communication70
General70
Math70
Physics70
Reasoning70
Tool Calling60
Code60
Economics60
Spatial Reasoning50
Vision50
Factuality50
Multimodal50
定價
輸入價格$0.2 / 1M tokens
輸出價格$0.825 / 1M tokens
混合價格(3:1)$0.356 / 1M tokens
速度
Tokens/秒67.8 tokens/s
首Token延遲1.17s
首回答延遲1.17s
可用提供商
(LS 內部計價單位)| 提供商 | 輸入價格 | 輸出價格 |
|---|---|---|
| Fireworks | 150K | 800K |