Anthropic's new Mythos-class model, Claude Fable 5, is now generally available on Databricks. It sets a new performance bar for complex, long-horizon tasks, but it comes with clear trade-offs in speed and cost.

Anthropic’s new AI model brings long-horizon autonomy, first-shot accuracy, and stronger code review to GitLab.

Anthropic launches Mythos with Fable 5. Six capabilities for agents: 80.3% code reasoning, extended thinking, autonomy. Vision, science acceleration, enterprise pricing