Anthropic releases Claude Fable 5, a Mythos-class model the public can finally use, days before a potential record IPO

Anthropic on Tuesday released Claude Fable 5, a model built on the same architecture as its restricted Mythos system, making Mythos-class intelligence publicly available for the first time. Fable 5 is available to enterprise customers and paid subscribers, but it comes with new safeguards that block responses in cybersecurity, biology, chemistry, and what Anthropic calls “distillation” scenarios, falling back to Claude Opus 4.8 to deliver a safe answer instead. Early data from Anthropic indicates the fallback triggers in fewer than 5% of sessions, meaning most users will interact with the full Fable 5 model for the vast majority of their queries.

“For us, it’s really around what we call ‘race to the top,’ being able to provide this technology in a valuable fashion, and at the same time providing the right safety guardrails so that it can do asymmetrically more benefits than harm,” Dianne Penn, Anthropic’s head of product management for research, told CNBC. The model represents what Penn described as a “significant jump” in capability over Opus 4.8, which is why the additional guardrails were necessary. On some benchmarks, Fable 5 scored more than 10% higher than Opus 4.8, according to Anthropic’s own blog post, though independent evaluations have not yet been published.

Anthropic also launched Claude Mythos 5, an updated version of the restricted Mythos model with the cybersecurity and biology safeguards lifted. Mythos 5 remains available only to the roughly 50 vetted partners in Project Glasswing, the company’s cybersecurity initiative that has uncovered more than 10,000 high- or critical-severity vulnerability candidates across major software projects since April. The two models share the same underlying architecture, with Fable 5 adding the safety restrictions that Anthropic deemed necessary for general deployment.

The pricing reflects the capability jump. Fable 5 costs $10 per million input tokens and $50 per million output tokens, twice as expensive as Opus 4.8, which launched late last month at $5 and $25 respectively. Through 22 June, Fable 5 is included at no extra cost for Pro, Max, Team, and seat-based Enterprise subscribers, but after that date access will require usage credits.

Penn said early customers have reported an improvement in spend per task, arguing that Fable 5’s higher intelligence means fewer iterations to complete work. “You just get a higher ROI by having more intelligent models,” she told CNBC. Those efficiency claims are based on early adopter feedback shared by Anthropic and have not been independently measured.

The launch arrives at a moment carefully calibrated for maximum commercial impact. Anthropic confidentially filed its IPO prospectus with the Securities and Exchange Commission last week, setting up a listing that could come as soon as this autumn. The company said in May that its revenue run rate had crossed $47 billion, up from roughly $10 billion in annual revenue in 2025, and it recently closed a $65 billion Series H round at a $965 billion post-money valuation.

That valuation placed Anthropic ahead of OpenAI for the first time. OpenAI closed its most recent round at $852 billion in late March and filed its own confidential IPO paperwork on Monday. Elon Musk’s SpaceX, which absorbed his AI startup xAI earlier this year, is expected to begin trading on Friday at a valuation approaching $1.8 trillion.

The convergence means three of the most valuable private technology companies are heading for public markets within months of each other, and the fall 2026 IPO window could see more than $200 billion in new public market value from those three listings alone. Claude Fable 5 is positioned as a demonstration that Anthropic can bring frontier-level intelligence to customers at scale without the catastrophic misuse scenarios that prompted it to restrict Mythos in the first place. The company has introduced a 30-day data retention policy on all Fable 5 and Mythos 5 traffic for safety monitoring, though it says the data will not be used for training.

Whether a model that blocks fewer than 5% of sessions can meaningfully prevent the kinds of attacks Mythos Preview was designed to find is a question Anthropic has not fully addressed. For investors evaluating Anthropic’s IPO prospects, the Fable 5 launch is a concrete answer to one of the company’s most persistent strategic questions: how to monetise Mythos-class capabilities without releasing the cybersecurity toolkit that makes the model both valuable and dangerous. The answer, at least for now, is a model that is broadly more intelligent everywhere else, at twice the price, and the coming quarters will determine whether that trade-off generates enough revenue to justify a near-trillion-dollar public valuation.