benchmark
.
space
benchmarks
rankings
compare
voices
transcripts
papers
articles
head to head
GPT-5.5
vs
Kimi K3
25 shared benchmarks
1
wins
1
ties
23
wins
38.5%
APEX-Agents
41.0%
22.7%
AutomationBench
30.8%
83.6%
BabyVision
85.7%
84.4%
BrowseComp
90.4%
84.1%
CharXiv
84.8%
67.0%
DeepSWE
67.3%
64.9%
FrontierSWE
81.2%
93.5%
GPQA Diamond
93.5%
41.4%
Humanity's Last Exam
43.5%
38.3%
JobBench
52.9%
92.2%
MATH-Vision
94.3%
82.8%
MCP-Atlas
84.2%
35.5%
MLS Bench
48.3%
81.2%
MMMU Pro
81.6%
60.9%
OfficeQA Pro
63.3%
89.4%
OmniDocBench
91.1%
55.8%
PerceptionBench
58.5%
28.4%
PostTrain Bench
36.6%
70.8%
Program Bench
77.8%
14.0%
SWE Marathon
42.0%
29.05%
SpreadsheetBench 2
34.8%
83.4%
Terminal-Bench 2.1
88.3%
73.5%
Toolathlon-Verified
73.2%
38.5%
WorldVQA
51.0%
22.0%
ZeroBench
23.0%