# Azure OpenAI

### Step-by-Step Token Generation

#### 1. **Create Azure OpenAI Resource**

- Log into [Azure Portal](https://portal.azure.com)
- Search for **"Azure OpenAI"** in the top search bar
- Click **"Create"** → **"Azure OpenAI"  
    [![Screenshot 2025-09-08 at 1.48.55 PM.png](https://docs.roost.ai/uploads/images/gallery/2025-09/scaled-1680-/AGYpc2rfJJgUKIrE-screenshot-2025-09-08-at-1-48-55-pm.png)](https://docs.roost.ai/uploads/images/gallery/2025-09/AGYpc2rfJJgUKIrE-screenshot-2025-09-08-at-1-48-55-pm.png)**
- Configure: 
    - **Subscription**: Select your Azure subscription
    - **Resource Group**: Create new or select existing
    - **Region**: Choose supported region (e.g., East US, West Europe)
    - **Name**: Enter unique name (e.g., "roostgpt-openai-prod")
    - **Pricing Tier**: Select Standard S0

#### 2. **Deploy Models to Resource**

- Navigate to your created Azure OpenAI resource
- Go to **"Model deployments"** in the left sidebar
- Click **"Create new deployment"**
- Configure deployment: 
    - **Model**: Select model (e.g., gpt-4o)
    - **Deployment name**: Enter name (e.g., "gpt-4o-deployment")
    - **Model version**: Select latest version
    - **Deployment type**: Standard
- Click **"Create"**

#### 3. **Access API Keys and Endpoint**

- In your Azure OpenAI resource, navigate to **"Keys and Endpoint"**
- Copy **KEY 1** or **KEY 2**
- Copy the **Endpoint URL** (format: `https://your-resource.openai.azure.com/`)
- **Important**: Store both key and endpoint securely

#### 4. **Secure Key Storage**

- **Copy immediately**: Keys are always visible but treat as sensitive
- Store endpoint URL and API key together

## Azure OpenAI Recommended Models for RoostGPT Integration

### **GPT-4o** (Primary Recommended)

- **Model Name**: `gpt-4o` (deployment name as configured)
- **Context Window**: 128K tokens
- **Pricing**: ~$2.50/1M input tokens, ~$10/1M output tokens
- **Deployment Required**: Yes - must deploy to your Azure resource
- **Best For**: **Primary choice** for all RoostGPT tasks - reliable, proven performance
- **Key Features**: 
    - Multimodal capabilities (text, images, code)
    - Enterprise-grade security and compliance
    - Data residency control
    - Excellent code understanding and generation
    - Integration with Azure ecosystem

### **GPT-4o-mini** (Cost-Effective Option)

- **Model Name**: `gpt-4o-mini` (deployment name as configured)
- **Context Window**: 128K tokens
- **Pricing**: ~$0.15/1M input tokens, ~$0.60/1M output tokens
- **Deployment Required**: Yes
- **Best For**: High-volume, cost-sensitive tasks
- **Key Features**: 
    - Significantly lower cost than GPT-4o
    - Good performance for standard tasks
    - Fast response times
    - Suitable for basic test generation