arxiv
GPT-4o
5 benchmarks
ChartQA
#9 of 15
86.7%
MMMU
#15 of 27
70.7%
MathVista
#13 of 14
63.8%
WebArena
#17 of 18
31.5%
VisualWebArena
#8 of 9
26.3%
benchmark.space canonical public ledger of AI capability