Anthropic launches Claude Opus 4.8 with dynamic workflows
anthropic released claude opus 4.8, topping gpt-5.5 in benchmarks with dynamic workflows and effort controls.
claude opus 4.8 improves coding accuracy, scoring 69.2% on swe-bench pro and 57.9% on humanity's last exam. new dynamic workflows allow planning and launching hundreds of parallel sub-agents for codebase migrations, while effort controls let users tune response speed and depth. standard api pricing remains $5 input and $25 output, but fast mode costs drop to $10 and $50.