Qwen3 235B A22B

Reasoning

Creator

Alibaba

Released

2025-04-28

Intelligence

19.8

Artificial Analysis Index

Coding

17.4

Artificial Analysis Index

In $/1M

$0.70

input tokens

Out $/1M

$8.40

output tokens

Blended $/1M

$2.63

3:1 blended

Speed

61

tokens / sec

Benchmark breakdown

Independent evaluation scores, normalized to 0–100.

GPQA Diamond
70
Humanity’s Last Exam
12
MMLU-Pro
83
SciCode
40
LiveCodeBench
62
MATH-500
93
AIME 2025
82
τ²-Bench (agentic)
24
Terminal-Bench Hard
6
IFBench
39
Via Artificial Analysis