arxiv
Kimi-VL-A3B-Instruct
2 benchmarks
MathVista
#8 of 9
67.1%
MMMU
#14 of 16
52.0%
benchmark.space canonical public ledger of AI capability