benchmark
.
space
benchmarks
rankings
compare
voices
transcripts
papers
articles
head to head
Gemini 3.1 Pro
vs
GLM-5.2
8 shared benchmarks
4
wins
0
ties
4
wins
98.13%
AIME
99.2%
12.0%
DeepSWE
44.0%
94.3%
GPQA Diamond
91.2%
44.4%
Humanity's Last Exam
40.5%
78.2%
MCP-Atlas
76.8%
54.2%
SWE-bench Pro
62.1%
80.6%
SWE-bench Verified
80.0%
70.3%
Terminal-Bench 2.1
81.0%