ToplineAnthropic released Claude Fable 5 on Tuesday, the first public version of its highly anticipated and highly feared Claude Mythos model—equipped with guardrails to prevent users from using it to exploit cybersecurity vulnerabilities or develop bioweapons.Anthropic said the new version, called Claude Fable 5, will have safeguards to prevent misuse in fields like cybersecurity.Getty ImagesKey FactsThe company also said it would release another more powerful model, Claude Mythos 5, for a “small group of cyberdefenders and infrastructure providers” that already had access to Claude Mythos Preview.Anthropic claims the two new models are superior at software engineering, knowledge work like financial analysis, and tasks that require “vision” and “long-context.”The AI company said it was working with the U.S. government to “steadily expand access” to the more advanced model, which it claims has the “strongest cybersecurity capabilities of any model in the world.”Claude Fable 5 and Mythos 5 will be priced at $10 per million input tokens and $25 per million output tokens—roughly twice the price of Anthropic’s most advanced Claude Opus models currently available, but still less than half the price Claude Mythos Preview was available for.Rumors about the expected release began circulating after the Sources newsletter first reported Anthropic was getting ready to release a public version with “substantial guardrails” on Tuesday, which was then also reported by The Information.How Will Claude Fable’s Safeguards Work?Anthropic said Claude Fable will launch with new systems to detect misuse such as attempts to “jailbreak” the model to access potentially harmful information. These systems, which Anthropic calls “classifiers,” work by rerouting queries about certain topics to Claude Opus instead after detection. These classifiers will include queries about cybersecurity, including attempts to use Claude for “agentic hacking,” as well as certain questions about biology and chemistry that could lead to development of bioweapons. The third classifier is what Anthropic calls “distillation,” or attempts to use Claude to build competing AI models.Key BackgroundAnthropic has been offering details on its upcoming model for months, announcing it in April but declining to release it to the public due to security concerns. Instead, it opted to instead release a version to a group of tech companies like Apple, Microsoft and Google through Project Glasswing in order to identify cybersecurity vulnerabilities. Anthropic publicly released Mythos just over one week after confidentially filing for an initial public offering. The company is reportedly aiming to go public later this year, and is currently valued at $965 billion after a fundraising round in May.