Anthropic has launched two new artificial intelligence models, Claude Fable 5.2 and Claude Mythos 5.2, as the company steps up competition with OpenAI, Google and Meta. The November 9 announcement comes as the AI startup prepares for a planned initial public offering (IPO) and looks to strengthen its position in the rapidly expanding AI market.
The company says the new models offer faster output generation, improved communication and lower pricing compared with their previous versions. However, the two models are designed for different applications, with Mythos access restricted to verified organisations.
Claude Fable 5.2 and Mythos 5.2: What’s Different?
Claude Fable 5.2 is designed for coding, reasoning and AI agent tasks. It is intended for broader availability and targets developers and businesses looking to use AI for complex workflows and multi-step tasks.
Claude Mythos 5.2, meanwhile, focuses on specialised applications in cybersecurity and life sciences. Anthropic will restrict access to organisations approved through its verification programmes for these research areas.
The restricted availability reflects the specialised nature of the model and the need to control access to its capabilities in sensitive fields.
New Models Offer Up to 30% Faster Output
Anthropic says Claude Fable 5.2 and Mythos 5.2 can generate outputs up to 30% faster than their Claude 5.1 predecessors. The company has also highlighted improvements in how the models communicate with users.
Pricing has been set at $8 per million input tokens and $40 per million output tokens. These rates are lower than those charged for the previous Fable and Mythos releases.
The reduced pricing could make the models more attractive to businesses and developers using AI for coding, research and automated tasks.
The launch follows Anthropic’s introduction of Claude Haiku 5.5 in October, reflecting its continued expansion of the Claude product portfolio.
Anthropic Faces Growing Competition From OpenAI and Google
The latest releases come amid intensifying competition in the global AI industry. OpenAI, Google and Meta are also introducing models focused on coding, reasoning and AI agents to attract enterprise customers and developers.
AI agents are becoming an important area of competition because they can perform multi-step tasks with limited human intervention. Companies are increasingly looking to integrate these capabilities into software development, business operations and research workflows.
Anthropic’s strategy of offering specialised models alongside more broadly available products is aimed at strengthening its position in this competitive market.
Planned IPO Adds Pressure to Anthropic’s Growth Strategy
The product launches also come ahead of Anthropic’s planned stock market listing. The company is seeking to demonstrate continued innovation and commercial growth as it prepares to enter public markets.
According to reports about its IPO preparations, the company has been exploring a potentially large share offering. However, specific fundraising and pricing figures circulated in an earlier filing were subsequently described as inaccurate, so the final IPO terms remain uncertain.
The performance, adoption and commercial success of its latest models could become important factors in investor assessments of Anthropic’s long-term prospects.
Summary
Anthropic has introduced Claude Fable 5.2 for coding, reasoning and AI agents, alongside the restricted-access Mythos 5.2 for cybersecurity and life-sciences research. The models offer up to 30% faster output generation and are priced at $8 per million input tokens and $40 per million output tokens. The launch strengthens Anthropic’s product portfolio as it prepares for a planned IPO and competes with OpenAI, Google and Meta.
