Anthropic’s advanced AI model Claude Fable 5 is back now

Anthropic has announced updated access guidelines for its advanced artificial intelligence model, Fable 5, allowing eligible Claude subscribers to use the system under a temporary usage allocation through July 7. During this period, users can consume up to half of their weekly plan allowance on the model before additional usage requires credits. After the deadline, access to Fable 5 will be provided exclusively through a credit-based system. Introduced on June 9, Fable 5 is the first publicly available model in Anthropic’s Mythos-class series. The company says the model delivers stronger performance across a range of tasks, including software development, scientific research, visual analysis, and knowledge-based work. It is also designed to handle longer and more complex assignments with greater autonomy than previous Claude models. Anthropic said the model was initially released with strict safety protections, but later faced U.S. export restrictions that required the company to limit access for certain users because it could not reliably verify users’ nationalities. The restrictions also affected access to Mythos 5, an AI model developed under the company’s Project Glasswing initiative for organizations working to protect critical infrastructure. Following security research that identified a method of bypassing some safeguards, Anthropic introduced an updated classifier that it says blocks the technique in the vast majority of cases. The company has also restored access to Mythos 5 for eligible U.S. organizations participating in Project Glasswing, while Fable 5 is currently available to Pro, Max, Team, and selected Enterprise subscribers.