Anthropic Launches Claude Fable 5, Its Most Powerful Public AI Model Yet
Anthropic released Claude Fable 5, a Mythos-class AI model now available to the general public, marking the company's most capable model ever offered for public use. The launch includes safety measures and token-based billing after a June 23 transition.
Latest Frontier Model Enters Public Market
Anthropic launched Claude Fable 5, a Mythos-class model made safe for general use, representing a significant milestone in making cutting-edge AI accessible beyond select researchers and enterprises. The model's capabilities exceed those of any model Anthropic has ever made generally available.
Safety-First Deployment
Claude Fable 5 uses conservative safety classifiers that trigger a fallback to Claude Opus 4.8 in less than 5% of sessions, in areas like cybersecurity. This architecture demonstrates Anthropic's commitment to deploying increasingly powerful models while maintaining guardrails. The company has restricted the model from engaging directly with certain sensitive queries, channeling them to safer alternatives instead.
Pricing Shift Signals Market Evolution
Fable 5 will be available on subscriptions until June 23, at which point it will transition to token-based billing, signaling the era of subsidized AI is coming to an end. This move follows GitHub Copilot's similar shift earlier in June, reflecting a broader industry trend toward usage-based pricing that directly reflects computational costs.
Industry Context
The Fable 5 launch arrives amid intense competition from OpenAI, Google, and Microsoft for dominance in frontier AI capabilities. Fable 5 is far better than any model Anthropic has previously released on most tasks, positioning the company as a serious contender in the race to commercialize next-generation AI systems. The public availability of such a powerful model could accelerate AI adoption across enterprises and individual developers worldwide.