Benchmarking

Anthropic Releases Claude Opus 5.5 With Fable‑Level Performance and 40% Lower Running Costs

Lower costs, faster outputs, external prerelease testing, tightened safeguards position Anthropic to scale professional coding and long‑horizon agent work