Anthropic Launches Claude Fable 5 and Mythos 5: A New Model Class Above Opus

Anthropic released Claude Fable 5 and Mythos 5 on June 9, 2026 — a new model class above Opus, split by safety tier. What launched, pricing, and why it matters.

Amon Taboi

10 Jun 2026 — 4 min read

TL;DR

On June 9, 2026, Anthropic publicly released Claude Fable 5 — its first general-availability model in a new Mythos class that sits above the Opus class.
A sibling model, Claude Mythos 5, is the same underlying system with safeguards lifted, released only to a small group of cyberdefenders and infrastructure providers through Project Glasswing.
Fable 5 is state-of-the-art on nearly all tested benchmarks and scored more than 10% higher than Claude Opus 4.8 on some, according to CNBC.
It is priced at $10 per million input tokens and $50 per million output — roughly double Opus 4.8 — and the launch lands days after Anthropic confidentially filed for an IPO.

Anthropic has opened up its most capable public model yet. On June 9, 2026 the company released Claude Fable 5, the first Mythos-class model available to the general public — a tier that, by Anthropic’s own description, exceeds the capability of anything it has previously shipped broadly. Alongside it, a less-restricted sibling called Claude Mythos 5 went to a narrow set of security and infrastructure partners. The two models are the same underlying system; what separates them is safety.

A new model class that sits above Opus

Mythos-class models are positioned as a tier above the Opus class in raw capability. The naming is deliberate: Anthropic notes that “Fable” comes from the Latin fabula — “that which is told” — which is close in meaning to the Greek mythos. This is not a point release on top of Opus 4.8; it is a new rung on the ladder, and the first Mythos-class model, Claude Mythos Preview, already shipped quietly in April through Project Glasswing.

What Fable 5 can actually do

Anthropic describes Fable 5 as state-of-the-art on nearly every benchmark it tested, with standout results in software engineering, knowledge work, vision, and scientific research. The pattern it emphasizes is telling: the longer and more complex the task, the larger Fable 5’s lead over other models. According to CNBC, on some benchmarks it scored more than 10% higher than Claude Opus 4.8, the model Anthropic announced just weeks earlier.

For anyone building on these models, the “long-horizon” framing matters more than the headline benchmark number. A 10% benchmark gap is easy to dismiss; a model that holds its advantage as tasks get longer is a different kind of tool for agents, multi-step research, and codebases.

The real story: a safety split

The key difference between Fable 5 and Mythos 5 is not capability — it is the safeguards. Anthropic shipped Fable 5 with conservative guardrails: in high-risk domains like cybersecurity, biology, chemistry, and distillation, the model declines and falls back to a response from Claude Opus 4.8. The company says those safeguards were tuned conservatively and trigger, on average, in less than 5% of sessions — meaning they will sometimes catch harmless requests.

Before release, Anthropic says it stress-tested the model’s classifiers against jailbreaks: an internal bug bounty produced no universal jailbreaks across more than 1,000 hours of testing, and external red-teaming organizations also failed to find one. Claude Mythos 5 is the same model with those safeguards lifted in some areas, released only to a small group of cyberdefenders and infrastructure providers — initially through Project Glasswing, in collaboration with the U.S. government. Anthropic describes it as having the strongest cybersecurity capabilities of any model in the world.

That is the part worth sitting with: capability and access are now being split by safety tier. The same weights are public-with-guardrails for everyone and unrestricted for a vetted few.

Pricing and rollout

Both models are priced at $10 per million input tokens and $50 per million output tokens — less than half what Mythos Preview cost, but roughly double the price of Opus 4.8. Anthropic says it expects demand for Fable 5 to be “very high, and difficult to predict,” and for that reason subscription plans will gain access in stages rather than all at once.

Why this lands now

The timing is not incidental. Fable 5 arrives days after Anthropic said it confidentially filed its IPO prospectus with the SEC. The company has reported a revenue run rate of roughly $47 billion, up from about $10 billion a year earlier, and recently closed a round at a $965 billion valuation — ahead of OpenAI’s $852 billion mark from late March. Releasing a model tier above Opus, at a premium price, with an explicit safety story, into a pre-IPO moment, is a statement as much as a product.

Frequently asked questions

What is Claude Fable 5?

Claude Fable 5 is Anthropic’s first publicly available Mythos-class model, released June 9, 2026. The Mythos class sits above the Opus class in capability, and Anthropic describes Fable 5 as state-of-the-art on nearly all tested benchmarks.

How is Fable 5 different from Claude Mythos 5?

They are the same underlying model. Fable 5 ships with conservative safeguards and is generally available; Mythos 5 has those safeguards lifted in some areas and is released only to a small group of vetted cyberdefenders and infrastructure providers through Project Glasswing.

How much does Claude Fable 5 cost?

$10 per million input tokens and $50 per million output tokens — less than half the price of Mythos Preview, and roughly double Claude Opus 4.8.

Is Fable 5 better than Claude Opus 4.8?

On capability, Anthropic positions it above Opus, and CNBC reports it scored more than 10% higher on some benchmarks. But in high-risk domains Fable 5 deliberately falls back to Opus 4.8, so for restricted topics you are effectively still getting an Opus-level response.