#AnthropicReleasesFable5Model Anthropic Unveils Claude Fable 5: The First Public "Mythos-Class" Model Balances Unprecedented Power with Cutting-Edge Safeguards

SAN FRANCISCO – June 9, 2026 – Just 11 days after unveiling Claude Opus 4.8, Anthropic has announced its most significant release to date: Claude Fable 5, the company's first publicly available Mythos-class AI model . The launch represents a carefully calibrated balance between democratizing access to frontier AI and mitigating the serious risks associated with models of this capability level .

Alongside Fable 5, Anthropic is releasing Claude Mythos 5—the same underlying model but with fewer restrictions—exclusively to a small group of trusted partners, including US government agencies and critical infrastructure providers, through Project Glasswing .

A Model That Outperforms Everything Before It

Claude Fable 5 represents a "major-version-bump-deserving step change forward," according to former Tesla AI director Andrej Karpathy . The model is state-of-the-art on nearly all tested benchmarks of AI capability, showing exceptional performance in :

· Software Engineering: Complete complex code migrations at unprecedented speed
· Knowledge Work: Senior-level reasoning on finance, legal, and analytical tasks
· Vision: Extract data from scientific figures and rebuild applications from screenshots
· Scientific Research: Autonomous work on complex multi-day research projects

"The longer and more complex the task, the larger Fable 5's lead over our other models," Anthropic stated in its announcement .

Real-World Impact: Stripe's Two-Month Migration Completed in One Day

Early testing partner Stripe reported extraordinary results: Fable 5 completed a codebase-wide migration across 50 million lines of Ruby code in a single day—a task previously estimated to require a team of engineers more than two months to complete manually .

IMC, a global trading firm, found that Fable 5 "aced their trading-analysis evaluations nearly across the board, including factual lookup, conceptual reasoning, root-cause analysis, and expected-value analysis" .

Safety First: The Fable 5 Safeguard System

Given that Anthropic previously described unrestricted Mythos-class models as posing "significant risk" in cybersecurity and biology, Fable 5 launches with a novel multi-layered safety system .

When the model detects requests related to cybersecurity, biology, chemistry, or model distillation, responses are automatically handled by Claude Opus 4.8 instead—a less capable but safer model . Users are informed whenever this routing occurs.

"Without safeguards, Fable 5's capabilities in areas like cybersecurity could be misused to cause serious damage," Anthropic explained . The safeguards are initially tuned "conservatively," triggering in less than 5% of sessions on average .

30-Day Data Retention: A New Standard for Safety Monitoring

Anthropic is implementing a mandatory 30-day retention policy for all Mythos-class model traffic—even for customers who previously used no-logs mode . The company stated:

"We won't use this data to train new Claude models, or for any non-safety-related purpose. This data will help us defend against complex and novel attacks, including new jailbreaks and attacks that operate across many requests" .

Pricing and Availability

Both Fable 5 and Mythos 5 are priced at $10 per million input tokens** and **$50 per million output tokens—less than half the cost of the previous Claude Mythos Preview . For context, this is double the price of Claude Opus 4.8 but significantly below OpenAI's GPT-5.5 Pro tier .

For Subscription Users:

· From June 9 through June 22, Fable 5 is included at no extra cost on Pro, Max, Team, and seat-based Enterprise plans
· Starting June 23, access will temporarily shift to a usage-credit system
· Anthropic aims to restore Fable 5 as a standard subscription feature "as quickly as capacity allows"

API Access:
Fable 5 is available immediately through the Claude API, Claude Platform on AWS, Amazon Bedrock, Vertex AI, and Microsoft Foundry .

Mythos 5: The Unrestricted Version

For trusted partners only, Claude Mythos 5 represents the full power of the Mythos architecture without the safety restrictions applied to Fable . Early results are remarkable:

· Drug Discovery: Protein design specialists accelerated parts of the drug discovery process by approximately 10x
· Hypothesis Generation: Mythos 5 generated novel molecular biology hypotheses preferred ~80% of the time over Opus-class models
· Autonomous Research: The model conducted genomics research autonomously over more than a week, assembling single-cell data from 138 animal species and training custom machine learning models

"It has the strongest cybersecurity capabilities of any model in the world," Anthropic stated. "Soon, we intend to expand access to Mythos 5 through a broader trusted access program" .

Industry Reaction

Andrej Karpathy, former Tesla AI Director, wrote on X: "This is a super exciting release. You can give it a lot more ambitious tasks than what you're used to—the model 'gets it' and it will just go" .

Boris Cherny, Anthropic's Head of Developer Relations, called Fable 5 "the biggest step up I've felt in our models since Opus 4.5," noting that Claude has evolved from a coding tool to a full-fledged product design partner .

Strategic Context: IPO on the Horizon

The launch comes at a pivotal moment for Anthropic, which recently confidentially filed its draft S-1 with the SEC, officially beginning the march toward a blockbuster IPO . The company completed a $650 billion Series H financing round on May 28, reaching a post-money valuation of $965 billion—surpassing OpenAI's $852 billion valuation .

This rapid release cadence—just 11 days after Opus 4.8, compared to 41 days between previous versions—signals Anthropic's aggressive push to maintain its technological edge while preparing for public markets .

About Anthropic
Founded in 2021, Anthropic is a public benefit corporation dedicated to building reliable, interpretable, and steerable AI systems. Its Claude family of models is trusted by enterprises worldwide, including Stripe, IMC, and numerous Fortune 500 companies. As of April 2026, Anthropic's annualized operating revenue exceeded $470 billion .

#AnthropicReleasesFable5Model #ClaudeFable5 #Anthropic #MythosClass