The new Claude model arrived with a fact worth checking before anything else: on Anthropic's own pricing page, the Fable row reads No for the free plan. We loaded the page the morning of the announcement, November 9, 2026, and read the columns straight from the table. Fable 5.2, the new top of Anthropic's model lineup, is a paid-plan model.
What Anthropic released
Anthropic announced two models on November 9, 2026: Claude Fable 5.2 and Claude Mythos 5.2. They share one underlying model. Fable includes safeguards that block high-risk cybersecurity tasks and most research biology tasks. Mythos removes those safeguards, which is why access differs: Fable comes with paid plans, while Mythos goes only to organizations Anthropic has vetted through its Cyber and Life Sciences verification programs.
Anthropic positions Fable 5.2 for demanding technical work: long coding projects, hard reasoning problems, and reading complex charts and images. The company says it communicates more clearly than Fable 5.1 and generates output up to 30% faster. On the API, it costs $8 per million input tokens and $40 per million output tokens, with an optional fast mode that costs twice as much and runs up to 2.5 times faster.
Anthropic's advice about when to reach for it is worth reading twice: start most coding and knowledge tasks with Opus 5.5, the model the company recommends as a starting point, and move up to Fable when the work outgrows it.
Why a beginner might care
Three things about this release matter if you're new to AI. First, launch week: from November 9 through November 16, Pro plans and standard Team seats get Fable included in their weekly limits at no extra cost, so a $20-a-month Pro subscriber can try the top of the lineup without paying beyond the subscription.
Second, the plan you're on decides what Fable costs after that. On Max ($100 a month) and Ultra ($200 a month), Fable is included up to half your weekly limits as a standard part of the plan. On Pro ($20 a month, or $17 with annual billing), the launch-week inclusion ends November 16, and Fable switches to usage credits, which bill as you use them on top of the plan.
Third, the Mythos name will fill headlines today, and the setup is easy to misread: there is no individual sign-up for Mythos. Organizations working in cybersecurity or biomedical research apply through Anthropic's verification programs, and everyone else reads about it from the outside.
What we checked on launch day
The pricing table is direct. On claude.com/pricing, the model comparison shows the Fable row as No on Free, Usage credits on Pro, and 50% of weekly limits on Max and Ultra. The same table keeps Sonnet and Haiku on the free plan, with Opus for paid plans.
There's no model picker without an account. Claude's app at claude.ai redirects to a login page when you're not signed in, so the in-app steps depend on documentation rather than our own eyes.
The documents agree. Anthropic's announcement, two help center articles, and the developer docs all confirmed the same numbers: $8 and $40 per million tokens on the API, Fable 5.2 in Claude Code version 2.1.321 or later, and launch offers running November 9 through 16.
We did not run the model itself. Our test setup has no Claude account, we don't create paid ones, and the free plan doesn't include Fable. The guide labels the model-picker steps as documented by Anthropic rather than tested by us, and this post keeps that line visible too.
The honest catch
Fable 5.2 is a paid model with a spend curve. After launch week ends on November 16, Pro subscribers pay for Fable through usage credits, which are pay-as-you-go charges on top of the subscription. On Max and Ultra, Fable counts against the same weekly limits as your other models, capped at half your usage, and past that cap you can keep going with usage credits or switch models.
The safeguards also shape what the model will do. Fable 5.2 blocks high-risk cybersecurity tasks like penetration testing and exploit generation, along with most research biology tasks, and a blocked question gets a response from an older Claude model instead.
The performance numbers come from Anthropic's own benchmarks, including 69.9% against 62.3% for Opus 5.5 on their FrontierSWE v2 software-engineering test. Independent checks were not available at publish time.
If your days are chat, search, and documents, Anthropic's guidance points the other way: Opus 5.5 for most tasks, and the plan you already have is probably enough.
The full walkthrough
Our study guide covers the plan requirements first, then the model-picker steps, a starter prompt shaped for this kind of work, and a version stamp with everything we did and didn't test: Claude Fable 5.2, Anthropic's model for demanding technical work.
FAQ
Is Claude Fable 5.2 on the free plan?
No. Anthropic's pricing page shows No for Fable on the Free plan, and the help center's plan article names only paid plans. The free plan keeps access to Sonnet and Haiku, with Opus reserved for paid plans in the same table.
What's the difference between Fable and Mythos?
Same underlying model, different safeguards and access. Fable carries cybersecurity and biology safeguards and comes with paid plans. Mythos drops those safeguards and is restricted to vetted organizations through Anthropic's Cyber and Life Sciences verification programs.
How much does Fable 5.2 cost?
On the API, $8 per million input tokens and $40 per million output tokens, with fast mode at $16 and $80. In the apps, it depends on your plan: included up to half the weekly limits on Max and Ultra, usage credits on Pro once launch week ends November 16, 2026.
Can I try Claude Mythos 5.2?
Only through an organization. Anthropic gates Mythos to vetted organizations in cybersecurity and biomedical research, and there's no individual sign-up.
Which Claude model should I use day to day?
Anthropic's own guidance: start most coding and knowledge tasks with Opus 5.5, move up to Fable 5.2 when a job outgrows it, and use Sonnet 5.5 or Haiku 5.5 when you want speed.
Want the plan-by-plan version with the starter prompt? The guide walks it screen by screen: learn.greatwithtech.com/guides/claude-fable-52.
More on this: Why Does AI Usage Run Out So Fast? How Subscription Limits Work and How Much Does a Website Cost? Freelancer, Agency, and AI Prices.