​
Overview

Azure OpenAI provides a secure, enterprise-grade way to use OpenAI’s models in your camelAI deployment. This guide walks you through the process of setting up Azure OpenAI and configuring the necessary models.

​
Prerequisites

  • An active Microsoft Azure account
  • Subscription with Azure OpenAI access (may require approval)

​
Step-by-Step Setup Process

​
1. Create an Azure OpenAI Resource

  1. Go to the Azure Portal
  2. Click Create a Resource
  3. Search for “openai”
  4. Click Azure OpenAI
  5. Click Create
  6. Fill in the required details:
    • Subscription: Select your Azure subscription
    • Resource group: Create new or select existing
    • Region: Choose East US2 (Global Standard). o3 is not available in other regions.
    • Name: Give your resource a unique name
    • Pricing tier: Select a pricing tier (Standard S0 is recommended)
  7. Click Review + create
  8. Click Create

​
2. Deploy Required Models

After your Azure OpenAI resource is created:

  1. Click Go to resource
  2. Click Azure AI Foundry portal
  3. Click Model catalog in the sidebar
  4. Deploy the following models (required for camelAI):
    • o3
    • gpt-4.1-mini
    • o3-mini

For each model:

  1. Click on the model name
  2. Click Deploy
  3. Use the default deployment name as provided by Azure
  4. Click Deploy
  5. Repeat for each required model

​
3. Get API Information

After deploying the models:

  1. Go to Home in the sidebar
  2. Copy the following information:
    • Azure OpenAI Service endpoint: This will be your AZURE_OPENAI_ENDPOINT
    • API Key 1 or API Key 2: This will be your AZURE_OPENAI_API_KEY

​
4. Configure camelAI

Update your .env.docker file with the Azure OpenAI settings:

# Azure OpenAI Settings
AZURE_OPENAI_ENDPOINT=your_azure_endpoint
AZURE_OPENAI_API_KEY=your_azure_api_key

​
Troubleshooting

Common issues:

  1. Models Not Available: Ensure the models are fully deployed and check deployment names match exactly what camelAI expects (default names)

  2. Authentication Errors: Double-check your API key and endpoint URL

  3. Rate Limit Errors: Azure OpenAI has rate limits. Consider adjusting your quota or upgrading your pricing tier

  4. Regional Availability: Some models may only be available in specific regions

​
Next Steps

After setting up Azure OpenAI:

  1. Configure Other Environment Variables
  2. Start Services
  3. Create a Superuser