Anthropic has added two new models to its Claude family, continuing a steady run of releases as the company works toward a planned initial public offering.
The models, Claude Fable 5.2 and Claude Mythos 5.2, were unveiled on Monday, according to a report by The Economic Times. One is aimed at a broad base of developers and businesses, while the other is reserved for approved organizations in sensitive research fields.
Two Models Built for Different Audiences
Claude Fable 5.2 is designed for coding, reasoning and tasks performed by AI agents, the kinds of workloads that have become central to how companies use large language models.
Claude Mythos 5.2 has a narrower focus. Anthropic is positioning it for cybersecurity and life-sciences research, two areas where powerful AI tools can deliver significant value but also raise concerns about misuse.
Restricted Access for Claude Mythos 5.2
Anthropic said Mythos 5.2 will only be available to organizations that pass its verification programs for cybersecurity and life-sciences research. Fable 5.2, by contrast, is being offered more widely.
The split reflects a broader pattern in the AI industry, where developers are increasingly gating their most capable or sensitive tools behind vetting processes rather than releasing everything to the general public.
Faster Responses and Lower Prices
According to Anthropic, both new models produce outputs up to 30% faster than the Claude 5.1 versions they replace. The company also said the models communicate more clearly with users.
Pricing is set at $8 per million input tokens and $40 per million output tokens. Anthropic said that is cheaper than what it charged for earlier Fable and Mythos releases.
For businesses running large volumes of requests, lower per-token costs combined with quicker responses can meaningfully change the economics of building AI-powered products.
A Busy Release Calendar Ahead of the IPO
The launch comes shortly after Anthropic rolled out Claude Haiku 5.5 in October. Together, the releases show the company widening its product lineup as it moves toward going public.
A broader catalogue of models at different capability levels and price points gives Anthropic more ways to win enterprise customers, which is likely to matter to prospective investors evaluating its growth.
Rivals Are Moving Just as Quickly
Anthropic is operating in an increasingly crowded field. OpenAI, Alphabet's Google and Meta Platforms are all shipping new models at a rapid pace, many of them focused on the same coding, reasoning and agent capabilities.
Each company is competing for the attention of developers and enterprise buyers, who are often willing to switch providers when a model offers better performance or lower costs.
What the Launch Means for Builders and Buyers
For startups and enterprise teams, the release adds another option at a lower price, with the speed improvements likely to be most noticeable in agent-based and coding workflows.
The gated approach to Mythos 5.2 also signals that access to specialized AI capabilities may increasingly depend on passing provider checks. As Anthropic heads toward its IPO, how well it balances openness, safety and pricing against fast-moving rivals will be closely watched.
Source: The Economic Times
