Skip to main content
The LLM is the brain of the agent. It generates responses and controls how the agent handles the conversation.
Recommended model: Use callkaro/arjuna-2.5 for your agent’s LLM. It is optimized for real-time voice conversations.

​
Temperature

To fine-tune the brain (make it sharper or more creative), use the Temperature setting in the LLM Model section. You can set it between 0 and 1 — the higher the temperature, the broader and more creative the agent becomes.
Tip: Start with callkaro/arjuna-2.5, build your use case, and then tune the temperature based on how the agent responds.

​
Secondary LLM

But what if the main brain fails? That is where the Secondary LLM comes in — it is a backup brain. If the primary LLM goes down, the secondary one quietly takes over, and the customer on the other end won’t even notice.