Claude Fable 5: The first Mythos model is powerful, expensive, and heavily filtered

Anthropic has unveiled Claude Fable 5, the inaugural model in its new Mythos class, which is designed to enhance AI capabilities while maintaining stringent safety measures. This launch is particularly significant as it positions Anthropic to compete more aggressively in the rapidly evolving AI landscape, where performance and safety are paramount.

Claude Fable 5 has achieved an impressive benchmark score of 95 percent on the SWE-bench, indicating its high performance relative to existing models. However, this power comes at a steep price, with costs ranging from $10 to $50 per million tokens, which is notably higher than its competitor, Opus 4.8. This pricing strategy may limit its accessibility for smaller developers and startups, potentially consolidating market power among larger enterprises that can afford these costs. Additionally, the model has a request blockage rate of 9 percent due to its extensive safety filters, which aim to prevent harmful outputs but could also hinder its usability in certain applications.

The introduction of Claude Fable 5 comes at a time when AI companies are increasingly focused on balancing performance with ethical considerations. As noted by The Decoder, the model’s data retention policy is set at 30 days, reflecting a commitment to user privacy and data security. This approach may resonate well with users concerned about data misuse, but it also raises questions about the long-term training capabilities of the model, especially in a field where continuous learning from user interactions can enhance performance.

In the competitive context, Claude Fable 5’s entry into the market could prompt other AI developers to reevaluate their own pricing and safety strategies. Companies like OpenAI and Google, which are also investing heavily in AI safety, may need to respond to Anthropic’s positioning by enhancing their offerings or adjusting their pricing structures to maintain market share. The emphasis on safety and ethical AI is becoming a key differentiator, and those who can navigate this landscape effectively will likely gain a competitive edge.

Looking ahead, it will be crucial to monitor how the market responds to Claude Fable 5’s pricing and performance, as well as how competitors adapt to Anthropic’s new model and its implications for AI safety standards.