Anthropic Releases Claude Fable 5: First Mythos-Class Model Available to the Public

Anthropic released Claude Fable 5, its most powerful model made broadly available, achieving Mythos-class performance in software engineering, knowledge work, and vision tasks with performance advantages that grow on longer, more complex tasks. The model ships with cybersecurity guardrails that block responses in high-risk domains including offensive cybersecurity and biology; Project Glasswing partners receive access to the upgraded Mythos 5 tier separately. Claude Fable 5 is available via the API, Claude.ai, and on Amazon Bedrock as part of a concurrent AWS launch. The release comes days after Anthropic raised concerns about AI capabilities growing faster than safety practices — the guardrails represent Anthropic’s current answer to that tension.