Connecting Open AI

1. To connect Open AI GPT technology to Virbe AI Editor, you'll need an API key from your own Open AI's account. If you don't have it already create an account in Open AI's developer platform.

2. Then, head out to the Open AI dashboard and make sure you configure all necessary settings to have your account up and running (your virtual being will be using credits from this account so make sure you have billing and limits all settled in OpenAI first).

3. Next, head straight to the "API keys" tab on the left-hand sidebar.

OpenAI dashboard: API keys tab is in the sidebar

1. Once there, click on "Create new secret key" and make sure you save that key (for example in your password manager or just a text file) for later use. Open AI notifies that they will not display this key again after they are generated.

Create API key in Open AI dashboard

1. Copy the key and go to Virbe Dashboard of your virtual being. There, head straight to the "Conversation Engine" tab in the side navigation pane, and click on "Go to the Editor".

   
2. There, head straight to "Settings" and "LLM configuration". Select "Open AI" from the AI service dropdown list, paste your API key and select the model version you'd like to use (we highly recommend going for GPT-4, as it's better supported, easier to prompt and results in better and more natural conversational experience).

   

   AI Editor: settings for LLM configuration
3. Remember to also fill in the field with Open AI's "Embeddings", which is necessary for the AI Editor to work, based on the information you'll provide in the Knowledge Base on later stages. Use the same API key that you already copied & pasted in the previous step.

   

   Embeddings are required for the AI Editor to work

Now you're ready to start generating your prompts and filling Knowledge Base content to give your virtual being tasks and personality!