Anthropic launched Claude Fable 5.2 and Claude Mythos 5.2 on November 9, its first release of frontier models since CEO Dario Amodei called for a slower pace of AI development. The company says the new models generate outputs 30% faster, with pricing 20% below the 5.1 generation. The launch comes as Anthropic prepares for a possible IPO roadshow, with a target valuation of $1 trillion to $1.5 trillion.
The models’ commercial potential and Anthropic’s safety commitments are likely to be central to how investors assess the company. Anthropic says both products underwent safety and alignment testing, and that it introduced new evaluation processes and post-launch monitoring tools.
Fable 5.2 targets demanding technical work
Dianne Penn, Anthropic’s head of product management, said Fable 5.2 is designed for power users and complex applications. It is suited to open-ended problems in fields such as engineering, mathematics and robotics, where the solution cannot easily be mapped out in advance. Penn said there remains a performance gap between these frontier tasks and everyday use cases, and that Anthropic expects some of the capabilities to become available in its Opus series over time.
Anthropic said Fable 5.2 became generally available on Monday, priced at $8 per million input tokens and $40 per million output tokens. Claude Opus 5.5, released in September, is better suited to routine tasks. The choice between the models depends on the complexity of the work and the capabilities required, rather than simply opting for the most advanced model.
Mythos 5.2 is not open to all users. Anthropic said vetted organizations will be able to access it through its life sciences and cybersecurity validation programs. The company first introduced Mythos in April, limiting early access to a small number of organizations because of its cybersecurity capabilities. Fable, released two months later, was made available to the public with additional safety measures.
Model launch follows Amodei’s call to slow AI development
The release marks Anthropic’s first launch of frontier models since Amodei called for a slower pace of AI development in September. Penn said the company’s position is to manage the pace of progress, not to halt research. Anthropic said it carried out rigorous safety and alignment tests before the launch and added evaluation and post-release monitoring tools.
Balancing safety investment with commercial competition is also relevant to Anthropic’s IPO case. The company detailed related uncertainties in the risk factors section of its recent filing. Ahead of the roadshow, management will need to explain to prospective investors how its development pace and safety governance fit with its growth plans.
IPO valuation target below earlier discussions
Anthropic could begin its IPO process as soon as next week. It is now targeting a valuation of $1 trillion to $1.5 trillion, below the $2 trillion figure discussed in recent months. The revised range puts the commercial demand for its models, their ability to generate revenue and the cost of safety measures in focus for investors. The company has not disclosed final offering terms in the information presented here.
In an amended filing, Anthropic reported revenue of $33.24 billion for the nine months ended September 30, up about 1,200% from a year earlier. That growth is an important part of the valuation context. The different access policies for Fable 5.2 and Mythos 5.2 also show how Anthropic is distinguishing between expanding product revenue and limiting risks associated with model use.