Anthropic launches Claude Fable 5.2 and Mythos 5.2 for coding, cybersecurity
Synopsis
Anthropic has launched two new AI models — Claude Fable 5.2 and Mythos 5.2 — capable of working autonomously for days, formalising complex mathematical proofs, and tackling frontier scientific problems. The timing is striking: the launch coincides with Anthropic reportedly kicking off IPO marketing at a valuation that could reach $1.3 trillion, making these models as much a business signal as a technical one.
Key Takeaways
Anthropic launched Claude Fable 5.2 and Claude Mythos 5.2 on 10 November 2026 , targeting advanced coding, cybersecurity, mathematics, and scientific research.
Both models share the same underlying architecture; Mythos 5.2 carries tighter access controls for sensitive applications like vulnerability discovery.
The models reportedly formalised a proof of Fermat's Last Theorem using nearly 1,000 AI agents in under 30 hours .
Standard pricing is $8 per million input tokens and $40 per million output tokens ; a fast mode doubles those rates for up to 2.5x faster generation.
Anthropic is reportedly beginning IPO marketing on 10 November 2026 , targeting a Nasdaq listing under ticker 'ANT' at a valuation of $1.1–$1.3 trillion .
The company's revenue reached $16.3 billion in the six months ended 30 June 2026 .
Anthropic has unveiled two new artificial intelligence models — Claude Fable 5.2 and Claude Mythos 5.2 — purpose-built for advanced coding, mathematics, cybersecurity, robotics, and scientific research. The models are designed to handle complex technical tasks autonomously for extended periods, marking a significant step forward in agentic AI capabilities.
What Sets Fable and Mythos Apart
According to Anthropic, both models are built on the same underlying architecture but serve distinct user segments. Fable 5.2 is intended for general users and is available through cloud platforms including Amazon Bedrock, equipped with strong cybersecurity and biology safeguards to prevent misuse. Mythos 5.2, by contrast, offers deeper access to capabilities involving vulnerability discovery and sensitive scientific research — and consequently carries tighter access controls.
'Fable has strong cybersecurity and biology safeguards to guard against misuse, which allows us to make it available to all users. Organisations pursuing beneficial applications in cybersecurity or biomedical research can apply for access to Mythos through our recently expanded Life Sciences and Cyber Verification Programs,' the company said in a statement.
Key Technical Capabilities
Autonomous AI operations are a central feature, with both models reportedly capable of working independently for hours or even days, coordinating multiple AI agents, and managing complex software projects with limited human oversight. Anthropic highlighted notable progress in mathematical reasoning — including a case where the system formalised a proof of Fermat's Last Theorem using nearly 1,000 AI agents within 30 hours.
The models also reportedly designed three functioning electronic devices, demonstrated improved performance in simulated robotics tasks, and contributed to research addressing previously unresolved theoretical physics problems related to black holes.
Pricing and Fast Mode
Standard pricing for both models is set at $8 per million input tokens and $40 per million output tokens. An optional fast mode — which speeds up token generation by up to 2.5 times — is available at $16 per million input tokens and $80 per million output tokens, according to Anthropic's statement.
IPO Plans and Valuation
The launch comes as Anthropic is reportedly set to formally begin marketing its initial public offering (IPO) from 10 November 2026, with the offering potentially rivalling SpaceX's record debut in scale. Reports indicate potential investors are eyeing a valuation in the range of $1.1 trillion to $1.3 trillion — a step down from the $1.5 trillion to $2 trillion range discussed earlier in November. The San Francisco-based company has filed to list on the Nasdaq under the ticker symbol 'ANT'. Revenue for the six months ended 30 June 2026 reached $16.3 billion, according to reports. How the market receives the new model launches — and their enterprise uptake — could meaningfully influence Anthropic's IPO valuation narrative.
Point of View
One gated — reflects a maturing tension in frontier AI: how to commercialise maximum capability without handing dangerous tools to bad actors. Anthropic's verification-programme approach for Mythos is pragmatic, but its real-world robustness remains unproven. More tellingly, the launch is timed almost precisely with IPO marketing; the Fermat proof and black-hole physics claims read as much as investor-deck highlights as product announcements. At a $1.1–$1.3 trillion target valuation, Anthropic needs Fable and Mythos to do two jobs simultaneously — advance the science and justify the multiple.
NationPress
10 Nov 2026
Frequently Asked Questions
What are Claude Fable 5.2 and Claude Mythos 5.2?
They are two new AI models launched by Anthropic on 10 November 2026, built on the same underlying architecture and designed for advanced coding, mathematics, cybersecurity, robotics, and scientific research. Fable 5.2 is available to general users, while Mythos 5.2 offers deeper capabilities with restricted access for sensitive applications.
What is the difference between Claude Fable 5.2 and Claude Mythos 5.2?
Fable 5.2 targets general users and includes strong cybersecurity and biology safeguards, making it broadly accessible including via Amazon Bedrock. Mythos 5.2 provides greater access to capabilities such as vulnerability discovery and advanced scientific research, requiring organisations to apply through Anthropic's Life Sciences and Cyber Verification Programs.
What can the new Claude 5.2 models do autonomously?
The models can reportedly work independently for hours or days, coordinate multiple AI agents, and manage complex software projects with limited human oversight. Anthropic cited a notable example of the models formalising a proof of Fermat's Last Theorem using nearly 1,000 AI agents within 30 hours.
How is Anthropic's new Claude 5.2 priced?
Standard pricing is set at $8 per million input tokens and $40 per million output tokens. An optional fast mode — accelerating token generation by up to 2.5 times — is priced at $16 per million input tokens and $80 per million output tokens.
What is the status of Anthropic's IPO?
Anthropic reportedly began marketing its IPO on 10 November 2026, planning to list on the Nasdaq under the ticker symbol 'ANT'. Reports suggest potential investors are targeting a valuation of $1.1 trillion to $1.3 trillion, down from an earlier $1.5–$2 trillion range. The company reported $16.3 billion in revenue for the first half of 2026.
