Anthropic has released a new version of its AI model, Fable 5.1, which demonstrates superior performance in programming and scientific tasks. In terms of programming, it excels at handling long-term complex tasks; for scientific capabilities, it can cover scenarios such as experimental design. Regarding costs, the token pricing for enterprise users remains the same as Fable 5, while the cost of reusing previously processed information by the model is reduced by 75%, significantly lowering overall operational costs. In terms of safety, the classifier has been upgraded to more accurately identify risky instructions. Additionally, Anthropic plans to allow enterprise customers to store data from their most powerful models in their own cloud infrastructure while continuing to perform security checks.
