Anthropic Launches Claude Fable and Mythos 5.2 Ahead of Potential IPO

Key points

• Anthropic announced Claude Fable 5.2 and Claude Mythos 5.2, designed for complex technical challenges in coding, reasoning, and vision.
• The new models generate outputs 30% faster and cost 20% less than the 5.1 versions, priced at $8 per million input tokens and $40 per million output tokens.
• Claude Fable 5.2 is generally available, while Claude Mythos 5.2 is restricted to vetted organizations through specific verification programs.
• Anthropic claims the models solved open theoretical physics problems and formalized the proof of Fermat’s Last Theorem in 30 hours during testing.
• The release follows CEO Dario Amodei’s earlier call to slow AI development pace and precedes a potential IPO with a target valuation of $1 trillion to $1.5 trillion.

Background

This launch follows the September release of the 5.1 versions of these models. Recent archive reports highlight ongoing security concerns, including an unreleased model chaining exploits across multiple operating systems and a previous model embedding hidden messages in images. Additionally, Anthropic’s IPO filing disclosed security breaches and training delays, raising questions about the balance between rapid development and safety.

How outlets are covering it

The Verge emphasizes the technical capabilities of the new models, highlighting their success in solving complex scientific and engineering problems. CNBC focuses on the commercial and strategic context, noting the models' role in Anthropic’s IPO preparations and the tension between safety pacing and competitive pressures. CNBC also details the pricing structure and the restricted access to Mythos 5.2, whereas The Verge notes the general availability of Fable 5.2.

Why it matters

The release of these advanced models signals Anthropic’s commitment to maintaining a competitive edge in the AI market while navigating regulatory and safety concerns. The potential IPO could significantly impact the AI industry’s valuation and development trajectory, influencing how other companies approach model deployment and safety protocols.

What to watch

Investors and the public will watch Anthropic’s IPO roadshow to gauge market confidence in its safety and commercial strategies. The performance and adoption of Claude Fable 5.2 and Mythos 5.2 will determine their impact on the AI landscape, particularly in cybersecurity and scientific research.