Advanced Settings for AI Agent - Full Guide

⚙️ Advanced Settings for AI Agent

The Advanced Settings are located in the "Update" section of each AI chatbot.

By default, optimal settings are applied when a chatbot is created. However, this section offers full control if you want to customize your AI agent further.


📌 Model Version

Choose the AI LLM model powering your chatbot.

What is a Model?

A model is the AI engine running your chatbot. Different models have unique costs and capabilities.

Model Description Use Case
GPT-3.5 Best for simple use cases Affordable and fast
GPT-4o Most powerful, supports function calling Best for all advanced use cases
GPT-4 Turbo Highest cost model Use for low-volume complex tasks

📌 Prioritize Question/Answer Results

When enabled: Prioritizes knowledge base Q&A content for more precise responses.
⚠️ Currently in experimental phase.


📌 Chatbot Visibility

Control who can see and interact with the chatbot.


📌 Temperature Setting

Adjust the AI's creativity/randomness.

  • Lower values (0.05 - 0.20): More predictable, controlled responses.
  • Higher values: More creative, varied outputs.

Recommended: Start testing between 0.05 and 0.20 for most businesses.


📌 Message Template (Advanced Users Only)

Modify the core template that controls how the AI agent processes knowledge base data and system prompts.

⚠️ Warning: Only change this if you fully understand AI prompt engineering.

Use this to build specialized agents (e.g., classify text, write emails).


📌 Support Email

Set a specific support email for this chatbot.
Displayed to users if errors or issues occur.


📌 Domains

Specify domains where your chatbot will work.

  • Separate multiple domains with commas
  • Format: https://yourdomain.com
  • If visibility is public, this setting is not required.

📌 Debug Mode

Enable this to see what information is pulled from the knowledge base and what’s sent to OpenAI.

✅ Great for testing and troubleshooting AI behavior.


Pro Tip: Mastering these advanced settings allows you to fine-tune performance, control costs, and expand your AI agent's capabilities.