benchmark
.
space
benchmarks
rankings
compare
voices
transcripts
papers
articles
head to head
Claude Opus 4.8
vs
GLM-4.7
3 shared benchmarks
2
wins
1
ties
0
wins
95.7%
AIME
95.7%
91.0%
GPQA Diamond
85.7%
49.8%
Humanity's Last Exam
42.8%