Anthropic
Claude Opus 4.5
1 benchmarks
SWE-bench Verified
#3 of 40
80.9%
benchmark.space canonical public ledger of AI capability