Anthropic Launches Claude Fable 5.2 and Mythos 5.2

What happened

Anthropic announced the two models on its website on Nov 9. Fable 5.2 is available from today on Claude paid plans, in Claude Code, through the Claude Platform and on Amazon Web Services, Google Cloud and Microsoft Azure, with the API model ID claude-fable-5-2; Mythos 5.2 is open only to organizations approved through the company's Cyber Verification Program and Life Sciences Verification Program.

Fable and Mythos share the same underlying model and differ in safeguards: Fable carries cybersecurity and biology safeguards that let Anthropic offer it to all users, while Mythos drops those blocks and goes only to vetted organizations. Anthropic said both models communicate more clearly than the 5.1 versions and generate outputs up to 30% faster; an optional fast mode speeds token generation by up to 2.5 times, at $16 per million input tokens and $80 per million output tokens.

In prerelease testing, Anthropic said the models completed software migrations that would otherwise have been infeasible, designed three working electronic devices from scratch, solved two open problems in black-hole physics, and had 1,000 agents formalize the proof of Fermat's Last Theorem in 30 hours; one engineer also had Fable 5.2 build a bootable prototype operating system from an empty folder in eight days. These results are the company's own claims and have not been independently reproduced.

Key facts

1. Publisher and timing: Anthropic released both models on Nov 9; Fable 5.2 went live the same day and Mythos 5.2 opened to vetted organizations
2. How the two relate: The same underlying model; Fable carries cybersecurity and biology safeguards for all users, while Mythos drops those blocks for vetted organizations only
3. Speed: Outputs up to 30% faster than the 5.1 generation; an optional fast mode speeds token generation by up to 2.5 times
4. Price: $8 per million input tokens and $40 per million output tokens, down from $10 and $50 for the 5.1 generation — about 20% lower; fast mode costs $16 and $80
5. Vendor benchmarks: Anthropic's published scores put Fable 5.2 at 49.1% on FrontierCode 2.0, above 32.4% for Fable 5.1 and 45.5% for Opus 5.5; the company said the two new models sit above Opus 5.5 on nearly every capability evaluated — all vendor-reported
6. Safety claims: Mythos 5.2's system card said it shows stronger covert capabilities than prior models on several evaluations; Anthropic said it cannot act covertly against its user; Fable 5.2's safeguards are as robust as the 5.1 generation's, with fewer unnecessary blocks
7. IPO context: The first frontier-model release since CEO Dario Amodei called in September for a slower pace of AI development; CNBC reported the company could list as early as next week, with valuation discussions between $1 trillion and $1.5 trillion