Anthropic details access, pricing and security features for Claude 5.2 models

Anthropic says Fable 5.2 is available for general use and has safeguards for certain high-risk biology and cybersecurity tasks.[3] The company lists access for Pro, Max, Ultra, Team and Enterprise users, as well as developers through its platform, marketplaces and major cloud services.[4] Mythos 5.2 is more restricted: Anthropic says direct access is limited to vetted people and organizations through trusted-access programs, including vetted cyberdefenders and life scientists.[3][5] The distinction matters for teams evaluating whether they can adopt Fable directly or need to seek trusted access to Mythos.

Anthropic lists standard prices of $8 per million input tokens and $40 per million output tokens for the models.[1] The company says the rates are 20% lower than for the 5.1 generation.[6] Anthropic also says the models generate outputs up to 30% faster than Fable 5.1 and Mythos 5.1.[1] An optional fast mode can increase token-generation speed by as much as 2.5 times, at $16 per million input tokens and $80 per million output tokens.[1] Teams comparing costs should distinguish this faster mode’s rates from the standard prices.

Anthropic says the models improve coding, reasoning and vision.[2] On FrontierCode 2.0, the company reports that Fable 5.2 scored 49.1% at Xhigh effort, the highest result among released models it tested.[1] Anthropic says Fable 5.2 at High effort scored about as well as Opus 5.5’s best FrontierCode result for roughly half the per-task cost.[1] The company nevertheless recommends starting with Opus 5.5 for most coding and knowledge tasks.[1] It also says an agent team using an earlier Fable 5.2 snapshot completed a formal proof of Fermat’s Last Theorem within 30 hours.[3]

Anthropic says Mythos 5.2 found 10.2% more vulnerabilities than Opus 5.5 and 14.4% more than Mythos 5.1.[1] The company describes it as the strongest cybersecurity-capable model it has released.[3] Anthropic says Claude Security, available to Enterprise customers, now includes Mythos 5.2.[1] For Fable 5.2, Anthropic says cyber and biology safeguards are as robust as Fable 5.1’s but cause fewer unnecessary blocks.[1] The company also says it believes the risk of high-stakes misalignment remains low.[1] These are company-reported assessments, not evidence here of outcomes in a particular organization’s environment.