Techstrong.ai reported that Anthropic launched Claude Fable 5.1 and Mythos 5.1 for complex enterprise workloads. The models handle extended autonomous tasks in coding, scientific research, and business workflows. Posts from the account and its linked source note major benchmark gains in coding and science. They also state a 75 percent cut in context caching costs to support lower operating expenses during long-running agentic operations. The updates focus on performance for sustained use cases rather than short queries.