Claude Fable 5 and Claude Mythos 5 is a Large Language Models (LLMs) tool. Powerful AI models for software engineering, research, and knowledge work with safeguards. Key features include Mythos-Class Capabilities, Built-In Safety Guardrails, and Extended Autonomous Work. Best for software developers and engineers, scientists and researchers and data scientists and analysts.
About Claude Fable 5 and Claude Mythos 5
Key Features
<strong>Mythos-Class Capabilities.</strong> Both models deliver exceptional performance on software engineering, knowledge work, vision tasks, and scientific research. They outperform previous Claude models on complex, long-running tasks.
<strong>Built-In Safety Guardrails.</strong> Fable 5 includes classifiers that block high-risk requests in cybersecurity, biology, and chemistry, automatically routing them to Claude Opus 4.8. This happens in less than 5% of sessions on average.
<strong>Extended Autonomous Work.</strong> The models can work autonomously for longer periods than any previous Claude model. They handle complex tasks that require sustained reasoning and execution without human intervention.
<strong>1M Token Context Window.</strong> Both models support a 1 million token context window by default and up to 128k output tokens per request. This allows processing of extremely large documents and codebases.
<strong>Adaptive Thinking Mode.</strong> The models use adaptive thinking as their only mode, with an effort parameter to control thinking depth. Thinking output can be summarized or omitted based on your needs.
<strong>Restricted Mythos 5 Access.</strong> Mythos 5 offers the same underlying model as Fable 5 but with safeguards lifted for approved users. It's available only through Project Glasswing for vetted cybersecurity and life sciences teams.
Frequently Asked Questions
Both models share the same base capabilities, but Fable 5 includes safety guardrails that block responses in high-risk areas like cybersecurity and biology. Mythos 5 has these safeguards lifted and is only available to vetted organizations through Project Glasswing. For most ordinary tasks, they perform the same.
Both Claude Fable 5 and Mythos 5 are priced at $10 per million input tokens and $50 per million output tokens. This is less than half the price of the previous Claude Mythos Preview model. When Fable 5 routes a request to Opus 4.8 for safety reasons, you only pay Opus prices.
Claude Fable 5 is available through the Claude API, Claude Platform on AWS, Amazon Bedrock, Vertex AI, and Microsoft Foundry. It's also included in Pro, Max, Team, and Enterprise subscription plans, though availability on subscriptions will shift to usage credits after June 22, 2026.
The models excel at software engineering, knowledge work, scientific research, and vision tasks. Early testing showed Fable 5 compressed months of engineering work into days. Mythos 5 accelerated drug design processes by around ten times and has the strongest cybersecurity capabilities of any model in the world.
