benchmark
.
space
benchmarks
rankings
compare
voices
transcripts
papers
articles
OpenAI
GPT-5.5
17 benchmarks
τ2-bench
#2 of 9
98.0%
GPQA Diamond
#6 of 101
93.6%
ARC-AGI 2
#2 of 9
85.0%
GDPval
#1 of 2
84.9%
BrowseComp
#7 of 33
84.4%
MMMU Pro
#1 of 22
83.2%
Terminal-Bench 2.0
#1 of 28
82.7%
CyberGym
#2 of 6
81.8%
MMMU Pro
#1 of 22
81.2%
OSWorld
#3 of 31
78.7%
MRCR v2
#6 of 13
74.0%
SWE-bench Pro
#4 of 27
58.6%
Toolathlon
#1 of 2
55.6%
Humanity's Last Exam
#9 of 50
52.2%
FrontierMath
#2 of 7
51.7%
Humanity's Last Exam
#9 of 50
41.4%
FrontierMath Tier 4
#3 of 5
35.4%