arxiv
Qwen 3 Coder + CodeStruct
1 benchmarks
SWE-bench Verified
#48 of 77
66.2%
benchmark.space canonical public ledger of AI capability