By Annie / 10 November 2026
For most UK small-business work, no - not as your everyday model. On 9 November 2026 Anthropic released Claude Fable 5.2 and Claude Mythos 5.2. The company itself says Fable and Mythos are for "the most ambitious, open-ended technical work", and that "for most coding and knowledge tasks" you should start with Claude Opus 5.5, which it calls its "strongest daily driver". The official API list price for Fable 5.2 and Mythos 5.2 is $8 per million input tokens and $40 per million output tokens, which is lower than the Fable 5.1 rates still shown on our Claude pricing UK page (checked 29 October 2026). Mythos 5.2 is not a model you can simply turn on: access is through Anthropic's verification programmes for cybersecurity and life-sciences work.
This is a documentation review, not a hands-on test. I have not run Fable 5.2, compared it with Opus 5.5, or checked a UK Claude or Console account.
What Anthropic actually launched
Fable 5.2 and Mythos 5.2 are, Anthropic says, the same underlying model. Fable is the generally available version, with "strong cybersecurity and biology safeguards" that let Anthropic offer it to all users. Organisations that need those dual-use capabilities can apply for Mythos through the Life Sciences and Cyber Verification Programmes. Claude Enterprise customers can also use Claude Security, which Anthropic says is now powered in part by Mythos 5.2.
Anthropic's announcement (9 November 2026) also says the new models:
- show improvements in coding, reasoning and vision,
- "generate outputs up to 30% faster" than Fable 5.1 and Mythos 5.1,
- communicate more clearly than earlier Fable and Mythos models, in testers' feedback,
- and can be used in an optional fast mode, which Anthropic says speeds token generation by up to about 2.5 times, at $16 per million input tokens and $80 per million output tokens.
Those speed, quality and benchmark figures are Anthropic's claims and early-tester quotes, not measurements from this site. See How we test.
The model ID given for developers is claude-fable-5-2. Anthropic says Fable 5.2 is available "today on all platforms, including Amazon Web Services, Google Cloud, and Microsoft Azure". I have not confirmed that a UK login sees it.
Update, 13 November 2026: Microsoft said on 11 November 2026 that Claude Fable 5.2 is rolling out to eligible users in Copilot Cowork and Copilot Studio. Admins manage access in the Microsoft 365 admin center. Microsoft says availability may vary by region and organisation. I have not tested a UK tenant. Source: Available today: Anthropic Claude Fable 5.2 in Microsoft Copilot.
Official prices, and what that means in pounds
On claude.com/pricing, checked on 10 November 2026 (Europe/London), the API tab lists Fable 5.2 at $8 per million input tokens and $40 per million output tokens. The same page says fast mode for Opus 5.5 and Fable 5.2 is available at twice the standard price, which matches the $16 / $80 fast-mode figures in the 9 November announcement.
I am quoting those official US-dollar figures as they appear. I am not converting them into new pound prices here, because our usual conversions on the Claude pricing UK page were last calculated on 29 October 2026 at a stated $1 = £0.75 and have not been refreshed for this article. When that page is updated, the Fable 5.1 row ($10 / $50) should be replaced or supplemented with Fable 5.2.
VAT still depends on how you buy. Anthropic's UK pricing page, as recorded on our 29 October check, says consumer plan prices "include 20% VAT". API credit purchases are a different matter: Anthropic's billing documentation, as cited on that same page, says they are charged "plus applicable taxes". Our VAT on AI subscriptions guide covers the general UK position. This is general information, not tax advice.
Who can use Fable on a Claude plan?
The plan table on claude.com/pricing, checked 10 November 2026, still describes Fable on a Pro plan as paid with usage credits, and on a Max plan as included for up to half of your weekly limit. That matches the shape of Anthropic's earlier "Claude Fable models on your plan" help article, which our pricing and usage-limits pages recorded on 29 October 2026 (Fable is not on the Free plan).
Separately, the 9 November announcement describes a launch-week promotion: extra Fable usage for subscribers on eligible paid plans, with one to three Fable resets based on how long they have subscribed, and Fable included for up to 50 per cent of weekly usage limits for Pro plans and standard Team seats. The launch post, rechecked on 14 November 2026, describes this as running "this week" and gives no end date, so it may already have ended. Treat it as a time-limited launch offer, not as the standing Pro entitlement on the pricing page, and check your own plan before you rely on it.
For API users, remember that a list price is not a monthly bill. Output tokens cost five times input on this model's standard tariff, and long agent jobs can use a lot of both. Set a spend cap in the Console before you point production work at a more expensive model.
When a small UK team should stay on Opus 5.5

If your week looks like drafting client emails, summarising meetings, writing first-pass website copy, or everyday Claude Code work, Anthropic is telling you to start with Opus 5.5. That is also the practical default for a small UK team that already budgets around a Pro or Max subscription: Fable on Pro, outside any live promotion, is usage credits on top of the plan.
Consider stepping up to Fable 5.2 only when all of these are true:
- You have a clearly harder job than your current model is finishing well - Anthropic's examples are long, open-ended technical work, not routine office tasks.
- You can read and approve the output. A more capable model is still capable of being wrong.
- You have a spend limit in place if you are paying per token.
- You know whether the work is allowed under Fable's safeguards. Anthropic says Fable will block high-risk cybersecurity and most research-biology tasks, and may fall back to a different Claude model when it does.
Do not choose Mythos 5.2 as a "better Fable" for general business use. It is aimed at vetted cybersecurity and biomedical-research organisations.
What this article does not do
- It does not rank Fable 5.2 against Opus 5.5, GPT-6 Astra or Gemini 4 Carbon. The benchmark table in Anthropic's post is the company's own evaluation, with its own effort settings and footnotes.
- It does not confirm UK in-app availability, the exact remaining life of the launch promotion, or what you will see in a particular Plan or Console view.
- It does not refresh the pound figures on our pricing hub or the month on the UK AI price tracker, which still reads "October 2026". Those pages carry their own dated notes on Fable 5.2.
If you need data residency rather than a new model, start with our UK data-residency guide, not with a model swap.
Frequently asked questions
Is Claude Fable 5.2 better than Claude Opus 5.5?
Anthropic presents Fable 5.2 as a step up for the hardest technical work, and still recommends Opus 5.5 as the daily driver for most coding and knowledge tasks. We have not tested either model for this article, so we are not adding our own ranking.
How much does Claude Fable 5.2 cost?
The official API list price on 10 November 2026 is $8 per million input tokens and $40 per million output tokens, or $16 / $80 in optional fast mode. What you pay on a Pro or Max plan depends on that plan's Fable rules and on whether any launch promotion is still running. Check Claude pricing UK and claude.com/pricing.
Can a UK small business use Claude Mythos 5.2?
Only if it is accepted onto Anthropic's Cyber or Life Sciences Verification Programmes, or, for some Enterprise customers, through Claude Security. It is not described as a general-availability business model.
Does faster output mean a lower bill?
Not necessarily. Fast mode is priced at twice the standard API rate, according to both the announcement and the pricing page. Standard-mode "up to 30% faster" is a speed claim, not a promise that you will use fewer tokens.
Sources
- Anthropic, "Claude Fable 5.2 and Mythos 5.2", 9 November 2026, checked 10 November 2026: https://www.anthropic.com/claude-fable-and-mythos-5-2
- Anthropic, Claude pricing (claude.com/pricing), checked 10 November 2026: https://claude.com/pricing
- Our Claude pricing UK and Claude usage limits pages, last checked against Anthropic's UK plan tables on 29 October 2026
Last checked: 10 November 2026 (Europe/London); launch post and API prices rechecked 14 November 2026. Evidence: Documentation review of Anthropic's 9 November 2026 announcement and claude.com/pricing. I have not used Fable 5.2 or Mythos 5.2, and I have not signed in to a UK Claude or Console account. General information, not financial, tax or technical advice. How we test. Editorial policy.



