Anthropic's Claude Opus 5 Beats Rivals at Half the Cost

Anthropic has released Claude Opus 5, a large language model available on its chatbot service and developer platform. The model outperformed its existing top public model, Fable 5, on eight of 13 benchmarks while costing about 50% less. It also excels at output verification and biology research, including automated protein analysis. Anthropic calls Opus 5 its safest model yet, scoring zero on an internal benchmark measuring vulnerability exploitation. It features more relaxed guardrails than Fable 5, with safety filters expected to trigger 85% less often. Opus 5 is now the default model in Claude's Max tier and available in the Pro plan.