benchmark
.
space
benchmarks
rankings
compare
voices
transcripts
papers
articles
head to head
Gemma 4 31B
vs
Qwen 3.6 27B
9 shared benchmarks
1
wins
0
ties
8
wins
89.2%
AIME
94.1%
84.3%
GPQA Diamond
87.8%
19.5%
Humanity's Last Exam
24.0%
85.2%
MMLU Pro
86.2%
76.9%
MMMU Pro
75.8%
51.7%
SWE-bench Multilingual
71.3%
35.7%
SWE-bench Pro
53.5%
52.0%
SWE-bench Verified
77.2%
42.9%
Terminal-Bench 2.0
59.3%