Back to Features & Tools

AI Chatbot

6 min readUpdated 15 March 2026
The AI Chatbot feature lets you add a smart, conversational assistant to your website that answers visitor questions based on your site content. It requires a business plan and an OpenAI API key.
This feature is only available on business plans. You will also need an OpenAI account with billing enabled to generate an API key.

Creating an OpenAI API Key

Get Your API Key

  1. 1Create an account at OpenAI's sign-up page if you don't already have one
  2. 2Add credit card details to enable API access
  3. 3Navigate to the API Keys page in your OpenAI dashboard
  4. 4Generate a new API key and copy it to your clipboard

Integrating with Limecube

Connect Your API Key

  1. 1In the Limecube editor, navigate to Administration > AI Chatbot
  2. 2Paste your OpenAI API key in the Setup tab
  3. 3Click Save to store the key

Training Your Chatbot

Once your API key is connected, you need to train the chatbot on your website content. The training process scans your pages so the chatbot can provide accurate, relevant answers to visitor questions.

Train Your Chatbot

  1. 1Open the AI Chatbot settings and navigate to the Training tab
  2. 2Click 'Scan' to automatically detect your website pages, or add individual URLs manually
  3. 3Select the pages you want the chatbot to learn from
  4. 4Click 'Train' to process the selected pages
  5. 5Review trained pages — you can activate, deactivate, edit, retrain, or delete them at any time
Retrain your chatbot after making content updates to your website. The AI does not automatically detect new or changed content — you need to trigger retraining manually.

Activating Your Chatbot

  1. 1Navigate to the Settings tab within the AI Chatbot section
  2. 2Toggle the activation button to enable the chatbot on your live site
  3. 3Click Save and test on your published website

Additional Training

You can customise the chatbot's behaviour by modifying the system prompt under 'Additional Training'. This controls the tone, focus areas, and response style of your chatbot. Always back up your changes locally before updating the system prompt.

Chat History

The Chatbot History tab lets you review all visitor interactions. Filter conversations by date to identify patterns, spot gaps where the chatbot provided inaccurate information, and improve your training data. This is invaluable for ensuring your chatbot delivers a quality experience.

Chatbot Greeting

Customise the initial greeting message that visitors see when the chatbot widget appears. Navigate to the Chatbot Greeting option under Settings to update this message to something that matches your brand voice.
Was this helpful?

Related Articles

Still need help?

Submit a Support Ticket