economy-mode

AI model costs are reduced by shifting Layer 3 model selection to cost-optimized alternatives using task-aware auto-selection logic. This mechanism routes routine work to cheaper models while reserving premium models for complex, high-performance requirements.

Task-aware model selection evaluates task complexity to dynamically switch between premium and economy models. It applies to session-based or persistent agent workflows, ensuring token expenditure is managed across long-running projects.

Yes, you can set cost-optimized models as a default preference via configuration files. This persistent configuration allows developers to automatically apply economy mode to long-running projects without manual activation each session.

Yes, integrating with a squad configuration system is required to manage state and apply the economy model mapping table. This integration ensures the cost-optimized model selection functions correctly within your agent workflows.

Economy mode does not override explicit user intent or charter preferences. The system automatically shifts to cost-efficient model alternatives for routine tasks while strictly preserving the high-performance models needed for critical work.

The best way to activate economy mode is by simply telling the agent to use economy mode. This initiates cost-optimized model selection for the current session, dynamically switching to faster, cheaper models for routine documentation and logging.