The model decides how your agent reasons. Mando lets you choose which one generates replies, alongside two settings that shape those replies: how long they can be, and how predictable they are.
Steps
Open your agent, go to Settings, then the AI model section.
Under General, pick an AI Model.
Set Max Tokens to control how long a reply can be. Higher limits allow longer replies when they are relevant.
Set Creativity between Predictable and Random.
Save.
Model choice affects your credits
Models are not priced the same. A stronger model reasons better and consumes more credits per message, so the choice is a trade between answer quality and how far your allowance goes. Watch your usage after changing it.
Creativity, in practice
Lower creativity produces grounded, repeatable answers, which is what most support agents want. Higher creativity produces more varied phrasing and more risk of the agent going beyond your content. If your agent is inventing things, lower this before you change anything else.
Frequently asked questions
Which model should I start with?
Start with the default and change it only if answers are weak. A better model rarely fixes a knowledge problem.
Do longer replies cost more?
Longer replies use more of each message allowance. Raise Max Tokens when answers are being cut short, not by default.