Anthropic releases Claude Fable 5.2 and Mythos 5.2 ahead of IPO
On Monday 9 November 2026, Anthropic released two models. Claude Fable 5.2 is generally available on its own platform and through major cloud providers. Claude Mythos 5.2 is limited to organisations approved through the company's Life Sciences and Cyber Verification Programs. According to Anthropic, both models generate outputs about 30% faster than their predecessors. Fable 5.2 costs $8 per million input tokens and $40 per million output tokens, down from $10 and $50 for Fable 5.1, according to Bloomberg as reported by The Next Web. For businesses that automate work, the practical change is a lower per-token cost for Fable and a faster model. The 30% speed figure is Anthropic's own claim, so teams should test it on their own tasks before changing anything.
In short
- Claude Fable 5.2 became generally available on 9 November 2026; Claude Mythos 5.2 is limited to vetted organisations.
- Fable 5.2 is priced at $8 per million input tokens and $40 per million output tokens, down from $10 and $50 for Fable 5.1.
- Anthropic says both models are about 30% faster, a claim to test on your own workflows before relying on it.
- Some Fable requests are routed to an earlier Opus model, so test the routing for security and biology tasks.
- The company's listing is expected in November 2026, but the exact date is reported differently by different outlets.
What happened
Anthropic released Claude Fable 5.2 and Claude Mythos 5.2 on Monday 9 November 2026. According to CNBC, Fable 5.2 is generally available that day, while Mythos 5.2 is available to vetted organisations through the company's Life Sciences and Cyber Verification Programs. The launch came as the company prepares for a stock market listing that CNBC described as potentially historic and possibly as soon as the following week.
Anthropic describes both models as its most advanced offerings. The company said they generate outputs 30% faster than their predecessors, and that they went through safety and alignment testing before release. According to Fortune, Fable 5.2 is available on all platforms, including Amazon Web Services, Google Cloud and Microsoft Azure. The UK AI Security Institute, Frontier Design and METR also tested the model before release, according to Anthropic as reported by The Next Web.
The price change
The price cut is the detail most relevant to budgets. Fable 5.2 now costs $8 per million input tokens and $40 per million output tokens. Fable 5.1 cost $10 per million input tokens and $50 per million output tokens, according to Bloomberg as reported by The Next Web, which puts the reduction at 20% on both. Yahoo Finance, citing the same pricing, says the company expects Fable 5.2 to be less expensive to run than its predecessor.
| Model | Access on 9 November 2026 | Input price per million tokens | Output price per million tokens |
|---|---|---|---|
| Claude Fable 5.2 | Generally available on Anthropic's platform and major clouds | $8 | $40 |
| Claude Fable 5.1 (previous) | Generally available | $10 | $50 |
| Claude Mythos 5.2 | Vetted organisations only, through verification programs | Not stated in the sources reviewed | Not stated in the sources reviewed |
Mythos 5.2 has no published general price in the sources used here. Anthropic's Mythos page states that use of Mythos models requires 30-day data retention for safety monitoring by default. Organisations with zero data retention arrangements can keep them until Anthropic offers that option for Mythos, according to the company's page.
Why it matters for businesses that automate work
Most business automation runs on model calls: reading an invoice, classifying an email, drafting a follow-up, or checking a bank line against a ledger. For these jobs, the cost per call and the speed of each reply shape the monthly bill and the wait time for staff. A lower input price and faster output can change which tasks are worth automating at all.
Fable 5.2 is the model that most general business workflows would use. Mythos 5.2 is aimed at cybersecurity and life sciences teams. Most businesses will not qualify for it, and most do not need it for document work.
What changes in practice for a business that automates work
The first change is a cost review. A team that already calls Fable 5.1 through an API or a partner product can compare its last three months of usage against the new rates. The saving depends on how many input and output tokens each workflow consumes. A workflow that sends long documents in and short answers out gains most from the lower input price. A workflow that writes long reports gains less from the lower input price, because output is still priced at $40 per million tokens.
The second change is speed. Fable 5.2 and Mythos 5.2 produce outputs about 30% faster, according to Anthropic. A fast mode that speeds up token generation by up to 2.5 times is available on Claude Code, the Claude Platform, Claude Models on Google Cloud and Microsoft Foundry, according to MacRumors. Fast mode costs more, and Mythos fast mode is limited to organisations in the verification programs, according to Help Net Security.
The third change is task routing. Anthropic says Fable's safety classifiers block high-risk cybersecurity tasks such as penetration testing and most biology research tasks. When Fable blocks a request in those areas, the system serves a response from an earlier Opus model, according to Help Net Security. A business that automates security checks or lab documentation should test which requests get routed, because the answer may come from a different model than the one it expected.
A practical test is to take ten real tasks from the last month, run them through Fable 5.1 and Fable 5.2, and compare cost, time and error rate. A person who knows the work should check the outputs. Vendor benchmarks, including those Anthropic has published, measure general tests, not a company's own invoices, emails or records.
What to watch next
- The listing timetable. The company has filed a price range of $70 to $78 per share, according to PYMNTS. UPI reports a Nasdaq debut on 19 November, but Yahoo Finance notes that the official date had not been confirmed as of 10 November.
- Independent testing of the 30% speed claim and the new benchmarks, from the UK AI Security Institute or other testers.
- Whether Mythos access widens beyond vetted organisations, and whether a general price for Mythos is published.
- Changes to Fable's safeguards, since those affect which requests return from the full model and which return from an older one.
Automation does not remove the need to check outputs. Models that are faster and cheaper still make errors, and a person still needs to decide what happens with a wrong answer, especially in finance, payroll or regulated records.
How AiStaffo would automate this
AiStaffo would start by connecting the model to the tasks a business already runs, such as invoice reading, payment follow-ups or document checks, through the tools the business already uses. The system would run the routine steps automatically, log each model call with its token count and time, and send a summary of cost and speed changes after the switch. A person would still approve any output that reaches a customer, a bank record or a payroll file, and would decide which tasks stay on the older model. The owner would keep control of the data and the final decision. Book a free automation audit to measure your own token use and task volumes.
Questions people ask
Is Claude Fable 5.2 available to all businesses now?
How much cheaper is Fable 5.2 than Fable 5.1?
Who can use Claude Mythos 5.2?
Is the 30% speed improvement independently verified?
When is the Anthropic IPO expected?
Book a free automation audit
Thirty minutes. We look at one process you run every week and tell you exactly what an AI worker would take off your desk, and what it would not.
