Anthropic Says a Mythos-Class AI Model Will Be Available Soon

Anthropic says Claude Opus 4.8 is a modest upgrade over Opus 4.7 and is now rolling out.

The company says it is making progress on safeguards for a broader release of Claude Mythos Preview in the coming weeks.

Mythos is currently restricted to selected Project Glasswing partners because of its strong cybersecurity capabilities.

Researchers say the model can find exploits quickly, though its high operating cost may help limit misuse.

Anthropic also added response-effort controls that can affect Claude’s accuracy, speed, and usage limits.