The reasoning version of the latest iteration of the Opus family of Claude LLM models. Improving in every aspect over its predecesor, 4.5 Opus, with an extended context length.