Anthropic ships Claude Fable 5.2 and Mythos 5.2
Anthropic released Claude Fable 5.2 and Claude Mythos 5.2 on November 9, 2026, marking the company's latest entry into the market for advanced AI models aimed at technical work Anthropic.
Two models, two paths to deployment
Fable 5.2 is now generally available for all users and organizations, positioned for demanding technical challenges in coding, reasoning, and vision work. Mythos 5.2, by contrast, operates under a restricted deployment model—available only to organizations approved through Anthropic's cybersecurity and life-sciences verification programs The Verge.
AnthropicDescription both models as built for "the hardest technical challenges," with Fable 5.2 incorporating safeguards against misuse in cybersecurity and biology contexts Anthropic. The restricted access for Mythos 5.2 reflects the company's approach to controlling deployment of more powerful capabilities in sensitive domains.
Performance and pricing
The new models deliver up to 30% faster output generation compared to their predecessors, Claude Fable 5.1 and Claude Mythos 5.1 CNBC. Anthropic priced both models at $8 per million input tokens and $40 per million output tokens, undercutting competitive offerings in the advanced model tier Anthropic.
Integration and availability paths
Fable 5.2 is accessible directly through Claude.com and via enterprise channels AWS. AWS has also integrated the model into its machine-learning services, offering customers direct API access. Mythos 5.2 requires pre-approval: organizations must apply through Anthropic's verification process and demonstrate legitimate use cases in cybersecurity research or life-sciences applications before gaining access Anthropic.
The release reflects a broader industry pattern in which frontier model developers segment their customer base by use case and risk profile. Fable 5.2's general availability signals confidence in its safety and usability across technical domains; Mythos 5.2's gating acknowledges the dual-use concerns inherent in models capable of assisting with sensitive cybersecurity or biological research.
For developers and enterprises, the speed improvement and aggressive pricing on Fable 5.2 directly compete with similarly positioned models from rivals. The tier-based access model for Mythos 5.2 adds operational friction for approval-seeking organizations but positions Anthropic as a managed provider of controlled, high-capability AI for regulated domains.
