Setting up Azure Open AI Large Language Model

K can be integrated with your own deployment of Azure Open AI models

You must request access to the Open AI services prior to completing the below steps.

Microsoft request form for Open AI services

Setting up your Azure Open AI deployment

• Open you Azure Console

• Search for Open AI

• Create the Open AI service by

  • Selecting (or creating) the resource group to deploy to

  • Select the Region

  • Add a Name

  • Select a Pricing Tier

• Review the network options and click Create

Deploying the model

• Select the deployment created above

• Select Go to Azure OpenAI Studio
  

  

• Go to Deployment tab. Click Create new deployment
  

  

• Add the details for deploying the model and click Create

  • Model: GPT-4

  • Model version: Auto-update to default

  • Name: Provide a name for the model deployment
    

    

• Once the model is deployed go to the Chat tab.
  
  Make sure the deployment is configured to the model you deployed.
  
  Click on View code
  

• Change to CURL. Note down the URL highlight below. Note down the KEY that is hidden at the bottom of the modal.
  

  

  
  

• Using the URL and Key you can now setup K.ai to use your deployed models by following the steps here - Setting up K.ai