Anthropic launches Claude Fable 5.1 and Mythos 5.1, claims major leap in coding and AI research
&imwidth=600&imheight=450&format=webp&quality=medium "Anthropic launches Claude Fable 5.1 and Mythos 5.1, claims major leap in coding and AI research")
Claude Fable 5.1 arrives with big coding upgrade and lower AI costs
Anthropic has launched Claude Fable 5.1 and Mythos 5.1. Fable 5.1 targets coding, research and knowledge work with lower costs and stronger performance, while Mythos 5.1 offers higher-risk capabilities only to vetted cybersecurity and life-science organisations.
Anthropic has launched Claude Fable 5.1 and Claude Mythos 5.1, its latest AI models aimed at coding, knowledge work and scientific research. The company calls them its most advanced models for these tasks, while claiming Fable 5.1 can deliver stronger performance at a lower cost than its predecessor.
The two models are actually the same underlying AI system. The key difference is in their safeguards. Fable 5.1 is available generally, while Mythos 5.1 has fewer restrictions for approved work in areas such as cybersecurity and life sciences.
Claude Fable 5.1 is cheaper and more capable
Anthropic says Fable 5.1 will cost about 25 per cent less than Fable 5 for typical token-based workloads. For highly agentic tasks, where AI carries out longer sequences of actions, savings can reach about 45 per cent because of lower cache-read pricing.
The company also says the model is better at long-running tasks and less likely to take shortcuts that reduce the quality of its work. One striking example came from investment firm Millennium. Anthropic says Fable 5.1 identified the cause of an extremely rare software crash that engineers and other AI models had failed to explain for several years.
Anthropic is positioning Fable 5.1 as more than a coding assistant. Its tests cover scientific research, business workflows, computer use, multidisciplinary reasoning and software engineering. On Anthropic's reported benchmarks, Fable 5.1 scored 52.6 per cent on Terminal-Bench-Science, compared with 24.7 per cent for Fable 5.
Anthropic says early users have also seen improvements in research and long-running projects, suggesting the model is increasingly designed to act as an AI worker rather than simply answer individual prompts.
Trending Stories
&im=FitAndFill=(700,400))
)
)
)
)
)
)
)
)
)
)
)
)
)
)
&im=FitAndFill=(700,400))
)
)
)
)
)
)
)
)
)
)
)
)
)
&im=FitAndFill=(700,400))
What is Claude Mythos 5.1?
Mythos 5.1 is the restricted version. Anthropic says it is designed for vetted cyberdefenders and life-science researchers through trusted access programmes. Access is currently limited mainly to selected US organisations.
The reason is capability. Anthropic says advanced models in cybersecurity and biology could be useful for legitimate research but could also be misused. For general users, Fable 5.1 is the model that matters. It is available through Claude and Anthropic's API ecosystem, while Anthropic continues to refine safeguards designed to make powerful capabilities usable without opening the same risks as the restricted Mythos version. The launch shows where the AI race is heading: models are no longer being judged only by how well they answer questions. Increasingly, the battle is about whether they can work for hours, solve difficult problems, write and debug software, conduct research and deliver useful results at a lower cost.
