Skip to main content

​
Azure AI Foundry Quickstart

Get started with Azure AI Foundry in 15 minutes. This quickstart will have you deploying your first AI model and making API calls.

​
What you’ll accomplish

  • Create an Azure AI Foundry hub
  • Deploy a chat model
  • Make your first API call
  • Test the model in the playground

​
Prerequisites

​
Step 1: Create your hub (2 minutes)

  1. Go to Azure AI Foundry portal
  2. Sign in with your Azure account
  3. Click “Create new hub”
  4. Fill in:
    • Name: quickstart-hub
    • Subscription: Your Azure subscription
    • Resource group: Create new → rg-ai-quickstart
    • Location: East US 2
  5. Click “Create”
Your hub will be ready in 2-3 minutes.

​
Step 2: Deploy a model (3 minutes)

  1. From your hub, click “Explore models”
  2. Search for GPT-4o mini
  3. Click the model card, then “Deploy”
  4. Choose “Serverless API”
  5. Set:
    • Deployment name: gpt-4o-mini-quickstart
    • Content filter: Default
  6. Click “Deploy”
The model deploys in about 2 minutes.

​
Step 3: Get your API credentials (1 minute)

  1. Go to “Deployments”
  2. Click your gpt-4o-mini-quickstart deployment
  3. Copy the Endpoint URL and API Key

​
Step 4: Make your first API call (2 minutes)

Test your deployment with this curl command: 
curl -X POST "YOUR_ENDPOINT_URL/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "messages": [
      {
        "role": "user",
        "content": "Hello! What can you help me with?"
      }
    ],
    "max_tokens": 100
  }'
Replace YOUR_ENDPOINT_URL and YOUR_API_KEY with your actual values. You should get a response like: 
{
  "choices": [
    {
      "message": {
        "role": "assistant",
        "content": "Hello! I'm an AI assistant built on GPT-4. I can help you with a wide variety of tasks..."
      },
      "finish_reason": "stop"
    }
  ]
}

​
Step 5: Try the playground (2 minutes)

  1. In your project, go to “Playground” → “Chat”
  2. Select your deployment
  3. Type: Explain quantum computing in simple terms
  4. Press Enter
You’ll see the AI respond in real-time!

​
Step 6: Try different conversation styles (5 minutes)

In the playground, experiment with different prompts: Creative writing: 
Write a short poem about AI and humans working together
Problem solving: 
I have 100 apples. I eat 20, give away 30, and sell 25. How many do I have left?
Code generation: 
Write a Python function that calculates the fibonacci sequence

​
What’s next?

🎉 Congratulations! You’ve successfully:
  • Created an Azure AI Foundry hub
  • Deployed your first AI model
  • Made API calls to generate AI responses
  • Explored interactive chat in the playground

​
Ready to build something real?

Build a Chat App

Create a full chat application with authentication and memory

Explore Models

Try different AI models for various tasks

​
Need help with specific tasks?

Deploy Models

Learn different deployment options and configurations

Use the SDK

Integrate AI into your applications with SDKs

​
Troubleshooting

Can’t create a hub?
  • Verify your Azure subscription is active
  • Check that you have Contributor permissions
  • Try a different region if capacity is limited
Model deployment failed?
  • Ensure you have sufficient quota
  • Check the deployment logs for specific errors
  • Try deploying in a different region
API calls not working?
  • Verify your endpoint URL and API key are correct
  • Check that your model deployment status is “Succeeded”
  • Ensure you’re using the correct API format
Need more help? Visit our community forum or check the troubleshooting guide.
This quickstart gave you hands-on experience with the core Azure AI Foundry workflow. You learned by doing - deploying models, making API calls, and testing responses. Ready to go deeper? Choose your next step based on what you want to build.