Qwen 3.6 27B
14 benchmarks
AIME
#24 of 108
94.1%
GPQA Diamond
#24 of 101
87.8%
MMLU Pro
#12 of 53
86.2%
RealWorldQA
#4 of 9
84.1%
LiveCodeBench v6
#2 of 3
83.9%
MMMU
#5 of 27
82.9%
CharXiv
#8 of 11
78.4%
SWE-bench Verified
#22 of 79
77.2%
MMMU Pro
#12 of 22
75.8%
SWE-bench Multilingual
#10 of 15
71.3%
AndroidWorld
#5 of 8
70.3%
Terminal-Bench 2.0
#18 of 28
59.3%
SWE-bench Pro
#15 of 27
53.5%
Humanity's Last Exam
#40 of 50
24.0%