arxiv
Qwen 3 Coder + CodeStruct
1 benchmarks
SWE-bench Verified
#56 of 86
66.2%
benchmark.space canonical public ledger of AI capability