跳转到主要内容
AITier

Qwen3 235B A22B (Non-reasoning)

AlibabaQwenOpen WeightApache 2.0 · Commercial OK

描述

Qwen3 235B A22B is a large language model developed by Alibaba, featuring a Mixture-of-Experts (MoE) architecture with 235 billion total parameters and 22 billion activated parameters. It achieves competitive results in benchmark evaluations of coding, math, general capabilities, and more, compared to other top-tier models.

发布日期
2025-04-28
参数规模
235.0B
上下文长度
131K
支持模态
text

能力雷达图

33
general
23
coding
40
reasoning
39
science估算
70
agents
0
multimodal

Science 在缺少专门科学评测时使用推理能力代理估算。

排行榜排名

领域#排名分数来源
代码能力榜351
19.0
AA
通用能力榜286
38.0
AA
数学推理227
39.0
AA
推理能力32
79.0
LS
科学能力275
40.0
AA

基准测试分数 (LLM Stats)

Biology

GPQA47.5%自报

Chemistry

SuperGPQA44.1%自报

Code

EvalPlus0.78 / 100自报
LiveCodeBench70.7%自报
Aider61.8%自报

Creativity

Arena Hard95.6%自报

Finance

MMLU87.8%自报
MMLU-Pro68.2%自报

General

MMLU-Redux87.4%自报
MMMLU86.7%自报
MBPP0.81 / 100自报
LiveBench77.1%自报
Include73.5%自报
MultiLF71.9%自报
BFCL70.8%自报
MultiPL-E65.9%自报

Language

BBH88.9%自报

Math

GSM8k94.4%自报
AIME 202485.7%自报
MGSM83.5%自报
AIME 202581.5%自报
MATH71.8%自报

Reasoning

CRUX-O0.79 / 100自报

AA 评测指数

Math Index
23.7
Intelligence Index
17.0
Coding Index
14.0
Math 500
0.9
Mmlu Pro
0.8
Gpqa
0.6
Ifbench
0.4
Livecodebench
0.3
Aime
0.3
Scicode
0.3
Tau2
0.3
Aime 25
0.2
Terminalbench Hard
0.1
Hle
0.0
Lcr
0.0

LLM Stats 分类评分

Writing
100
Creativity
100
Language
80
Math
80
Reasoning
80
Tool Calling
70
Code
70
Finance
70
General
70
Healthcare
70
Legal
70
Biology
50
Chemistry
50
Physics
50
Economics
40

定价

输入价格$0.45 / 1M tokens
输出价格$1.8 / 1M tokens
混合价格(3:1)$0.787 / 1M tokens

速度

Tokens/秒64.1 tokens/s
首Token延迟1.24s
首回答延迟1.24s

可用提供商

(LS 内部计价单位)

暂无提供商数据

外部链接

LLM Stats