Claude Fable and Mythos 5.2 target the hardest technical work
Anthropic released Claude Fable 5.2 and Claude Mythos 5.2, the same underlying model with different access safeguards. Anthropic reports clearer output, up to 30% faster generation than 5.1, and lower API pricing of $8 per million input tokens and $40 per million output tokens.
Source checked Nov 10, 2026
Listen to this articleAbout 2 min
Preparing audio…
AI narration. Your voice and speed choices stay in this browser.
EDITORIAL TAKEAWAY
Start with Opus 5.5 for everyday work, as Anthropic recommends. Move to Fable only for unusually difficult open-ended technical tasks, and compare the result with the added cost and safeguards.
Before you try it
- Fable 5.2 is available broadly with Anthropic's cyber and biology safeguards; qualifying organizations can apply for Mythos through expanded verification programs.
- Optional fast mode is priced at $16 per million input tokens and $80 per million output tokens.
- Eligible paid subscriptions that were in place by November 9 at 10:00 a.m. PT receive limited launch usage or resets; exclusions and expiry rules apply.
A small next step
Our suggested way to explore the change.
- Choose one demanding technical task that has resisted a daily-driver model.
- Run it first with Opus 5.5, then compare Fable 5.2 on correctness, clarity, latency and cost.
- If using a launch reset, review the Help Center rules before applying it because a reset cannot be undone.
Go to the source
Anthropic · Claude Fable 5.2 and Mythos 5.2 ↗Claude Help · Fable 5.2 launch offers ↗Availability can change. Check the maker's current requirements before making a decision.