Choose and tune a model
Select the AI model, reasoning effort, creativity, and response length used by Agent.
Open Settings → Model to choose the AI model that powers Agent. The available list is maintained in Settings so you can compare the current options there.
Choosing a model
Model descriptions summarize the intended balance between speed and capability. A faster model is often a good fit for simple everyday actions. A more capable model can be useful for long documents, ambiguous requests, or work that requires deeper planning.
Changing the model does not change the Shortcut or the tools you have enabled.
Reasoning effort
Reasoning effort controls how much additional work a supported model spends before responding or selecting tools.
- Automatic lets the selected model use its normal behavior.
- Lower levels favor quicker, simpler work.
- Higher levels can help with difficult or multi-step requests but may take longer.
Only the levels supported by the selected provider and model are shown.
Advanced settings
Advanced settings are optional. Leave them off unless you want to tune a specific behavior.
Temperature
Temperature adjusts variation in model responses. Lower values tend to be more consistent and precise. Higher values can introduce more variety but may make results less predictable.
Response length
Response length controls the maximum output available to the model. A shorter setting suits concise replies and quick actions. A longer setting leaves more room for detailed writing or analysis.
A practical default
For most users, choose a balanced everyday model, leave reasoning on Automatic, and keep Advanced Setup off. Adjust one setting at a time when you can name the behavior you want to change.