Anthropic released Claude Fable 5.2 and Mythos 5.2 on Nov. 9. Fable 5.2 costs $8 per million input tokens and $40 per million output tokens, 20% less than Fable 5.1 and Mythos 5.1, which Anthropic's pricing page lists at $10 and $50. Fable is open to all users.

Who can use what

Anthropic says Fable and Mythos are the same underlying model. The difference is safeguards.

  • Fable 5.2: strong cyber and biology safeguards, so it is available to all users. It is on all platforms, including AWS, Google Cloud and Azure. The API ID is claude-fable-5-2.
  • Mythos 5.2: organizations apply through the Cyber Verification Program or the Life Sciences Verification Program.
  • Enterprise: Claude Security is powered by the model family, including Mythos 5.2.

What it costs

  • Standard: $8 input and $40 output per million tokens, down from $10 and $50.
  • Fast mode: $16 input and $80 output, twice the standard 5.2 rate. Anthropic says it generates tokens up to 2.5x faster. It runs on Claude Code, Claude Platform, Claude Models on Google Cloud and Microsoft Foundry. Mythos fast mode is only for organizations in the two verification programs.
  • Prompt caching: a 5-minute write is $10, a 1-hour write is $16 and a read is $0.20, per million tokens. On 5.1 these were $12.50, $20 and $0.25.

Subscriber promotion

Subscribers on eligible paid plans get extra Fable usage this week: one to three Fable resets, depending on how long they have subscribed. As a launch promotion, Fable is included for up to 50% of weekly limits on Pro and standard Team seats. See our look at the $200 Ultra plan for another recent Claude plan change.

What Anthropic claims

Anthropic says 5.2 improves coding, reasoning and vision, and communicates more clearly. It says output is up to 30% faster than Fable and Mythos 5.1.

Anthropic also says that in prerelease use, engineers completed software migrations that were otherwise infeasible, designed three working electronic devices, solved two open black-hole physics problems and formalized the proof of Fermat's Last Theorem in 30 hours. These are Anthropic's own claims. We have not verified them.

A footnote says Mythos 5.2 finds cyber vulnerabilities "roughly on par with Opus 5.5" on Anthropic's own evaluations.

Safety and data retention

Anthropic says high-stakes misalignment risks remain low and that Fable 5.2's cyber and biology safeguards are as robust as 5.1's. Eligible customers can keep using the models with zero data retention until they put safety monitoring with Enterprise Frontier Safeguards (EFS) in place. Enterprise customers can request EFS through their account team. Other customers get 30-day retention for safety monitoring. For earlier context, see METR's review of Anthropic's risk report.

Why it matters

The cut applies to both token rates, so heavy API users save the same 20% on input and output. Fast mode reverses that for anyone who turns it on: it costs double the standard rate, so speed is a paid choice. Mythos is the same model, but only verified organizations can use it, and Mythos fast mode is limited the same way. Most people will use Fable.

What we checked

We read Anthropic's announcement and pricing page on Nov. 9. The Verge, TechCrunch and 9to5Mac also covered the launch. Performance claims are Anthropic's. We did not run the models ourselves.