On June 9, 2026, Anthropic released Claude Fable 5, its most capable AI model for the general public, marking a significant milestone in “Mythos-class” technology.
This release follows the limited “Project Glasswing” initiative which previously restricted such powerful models due to safety concerns.
Anthropic is positioning this model as a breakthrough for complex, long-running tasks. As explained by Mario Rodriguez, Anthropic’s Chief Product Officer:
“Claude Fable 5 is a clear step forward on agentic coding and prototyping.”
To enable public access while managing risks in sensitive areas like cybersecurity and biology, Anthropic implemented unique safety classifiers.
If a query is flagged, the system automatically redirects the request to the highly capable Claude Opus 4.8.
Alongside this, the company launched Claude Mythos 5, which is the same underlying model but with those specific safeguards removed.
Access to Mythos 5 remains strictly gated, provided only to vetted partners and security experts.
Anthropic’s head of product for research, Dianne Penn, stated, “For us, it’s really around what we call ‘race to the top,’ by being able to provide this technology in a valuable fashion and at the same time providing the right safety guardrails.”
