Artificial intelligence company Anthropic has launched two of its most powerful AI models to date — Claude Fable 5 and Claude Mythos 5 — marking what the company describes as a significant leap forward in AI capability, while simultaneously introducing new safety guardrails to manage the risks that come with that power.
The announcement, made on 9th June 2026, centres on Claude Fable 5: a Mythos-class model that Anthropic has made safe for general use.
What is a “Mythos-class” Model?
Mythos-class models represent a tier of Claude models that sit above the company’s existing Opus class in capability. The first such model — Claude Mythos Preview — was released in April through Project Glasswing, a restricted programme for cyber defenders and infrastructure providers. Fable 5 and Mythos 5 follow as the next generation.
The naming itself is telling. “Fable” is derived from the Latin fabula, meaning “that which is told,” akin to the Greek mythos — and the safeguards are precisely what distinguish the two models and justify their separate names.
Fable 5: Top of the Benchmarks
Fable 5’s capabilities exceed those of any model Anthropic has ever made generally available, achieving state-of-the-art performance on nearly all tested benchmarks, including software engineering, knowledge work, vision, scientific research, and many other areas. The longer and more complex the task, the larger Fable 5’s lead over previous models.
Early access customers have been effusive in their praise. Stripe reported that Fable 5 compressed months of engineering into days, performing a codebase-wide migration across a 50-million-line Ruby codebase in a single day — a task that would otherwise have taken an entire team more than two months.
On vision tasks, Fable 5 became the new state-of-the-art, capable of extracting precise numbers from detailed scientific figures and rebuilding a web application’s source code from screenshots alone. Notably, previous Claude models struggled to play the video game Pokémon FireRed even with helper tools, but Fable 5 completed the game using only raw visual input — no maps, navigation aids, or extra game-state information.
In finance, Fable 5 achieved the highest score of any model on Hebbia’s Finance Benchmark for senior-level reasoning, with substantial gains in document-based reasoning, chart and table interpretation, and problem solving.
The Restricted Twin: Claude Mythos 5
While Fable 5 is available to the public, its counterpart carries far greater power — and far greater restrictions.
Claude Mythos 5 is the same underlying model as Fable 5 but with safeguards lifted in certain areas. It will initially be deployed through Project Glasswing, in collaboration with the US government, as an upgrade to Claude Mythos Preview. Anthropic describes it as having the strongest cybersecurity capabilities of any AI model in the world.
The company’s internal teams put Mythos 5 to work in life sciences with striking results. Using Mythos 5, Anthropic’s protein design experts accelerated aspects of the drug design process by around ten times. In one example, the model — equipped with protein design and bioinformatics tools but no human assistance — matched or beat skilled human operators, completing all tasks normally performed by a scientist: choosing binding sites, selecting and running tools, and recovering from failures along the way. Nine of 14 protein targets yielded strong drug design candidates currently under investigation.
Mythos 5 also conducted novel genomics research over more than a week of largely autonomous work, assembling single-cell data for millions of cells across 138 animal species and training a custom machine learning model that outperformed a recently published model in the journal Science — despite being 100 times smaller.
Safety First: New Classifiers and a Cautious Rollout
Releasing a model this capable comes with risks. Without safeguards, Fable 5’s capabilities in areas like cybersecurity could be misused to cause serious damage. Anthropic has therefore launched the model with safeguards, meaning queries on some topics will instead receive a response from Claude Opus 4.8, the company’s next-most-capable model.
The safeguards have been tuned conservatively and will sometimes catch harmless requests, though they trigger, on average, in less than 5% of sessions. More than 95% of Fable sessions involve no fallback at all.
Three areas are covered by the new classifiers:
Cybersecurity — Mythos-class models excel at discovering and exploiting software vulnerabilities, making cyberattacks substantially easier and cheaper. To prevent misuse, cybersecurity classifiers cover both vulnerability exploitation and offensive cyber tasks more broadly. External red-teaming organisations engaged by Anthropic failed to find universal jailbreaks on long-horizon agentic tasks, though the UK AISI made progress toward one within a brief initial testing window.
Biology and chemistry — The company cited concern about well-resourced malicious actors attempting to exploit the models’ scientific capabilities. For the time being, Fable will fall back to Opus 4.8 on most biology and chemistry requests, with plans to narrow these restrictions as safeguards improve.
Distillation — Anthropic noted previous large-scale attempts to extract Claude’s capabilities to train competing models in authoritarian countries. Requests identified as part of such distillation attempts will also fall back to Opus 4.8.
A New Data Retention Policy
Alongside the model launches, Anthropic announced a change to how it handles business customer data for Fable 5, Mythos 5, and future models of similar or greater capability. The company will require 30-day data retention for all traffic on Mythos-class models. This data will not be used for model training or any non-safety-related purpose, and will be deleted after 30 days in almost all cases. The measure is intended to help defend against complex and novel attacks, including new jailbreaks.
Pricing and Availability
Both Fable 5 and Mythos 5 are priced at US$10 per million input tokens and US$50 per million output tokens — less than half the price of Claude Mythos Preview.
Claude Fable 5 is available everywhere today via the Claude API. Claude Mythos 5 remains restricted to Glasswing partners, with plans to expand access through a broader trusted access programme.
For subscription-plan users, access is being phased: from today through 22nd June, Fable 5 is included on Pro, Max, Team, and seat-based Enterprise plans at no extra cost. From 23rd June, using it will require usage credits. Anthropic has stated it intends to restore Fable 5 as a standard part of subscription plans as quickly as capacity allows.