arxiv
Qwen 3 VL 32B Instruct
2 benchmarks
MathVista
#3 of 14
81.8%
MMMU
#16 of 27
70.6%
benchmark.space canonical public ledger of AI capability