Anthropic has announced the release of its first Mythos-class model, Claude Fable 5. The company claims this new AI shows exceptional performance in areas like software engineering and knowledge work.

According to Anthropic, the new safeguards make it possible to release a model that was once deemed too risky due to its capability in sensitive fields like cybersecurity and biology.

Fable 5 is being released alongside Claude Mythos 5, which shares the same underlying model but with some safeguards lifted. Access is currently limited to organizations through Anthropic’s Project Glasswing initiative, with plans for broader access in the future.

Pricing has also been adjusted, with both models costing significantly more than previous flagship offerings.

Anthropic did not provide further details on how these new models relate to their earlier releases or exactly what those safeguards mean for users.