Documentation

Azure OpenAI

Integrate Azure OpenAI deployments with ModelRed

☁️

Azure OpenAI Service

Test enterprise-grade OpenAI and custom models deployed on Azure with enhanced security, compliance, and regional deployment options.

Quick Setup

Get Started in 4 Steps

Connect your Azure OpenAI deployment for security testing.

1

Get Azure Resource Info

From your Azure Portal, collect your OpenAI resource details.

Endpoint: https://your-resource.openai.azure.com/API Key: from Keys and EndpointDeployment: your-model-deployment
2

Set Environment Variables

BASH
export AZURE_OPENAI_API_KEY="your-azure-api-key"
export AZURE_OPENAI_ENDPOINT="https://your-resource.openai.azure.com/"
export AZURE_OPENAI_DEPLOYMENT="gpt-4-deployment"
3

Register Your Model

PYTHON
from modelred import ModelRed

async with ModelRed() as client:
    await client.register_azure_model(
        model_id="my-azure-gpt4",
        api_key="your-azure-api-key",  # or use env var
        endpoint="https://your-resource.openai.azure.com/",
        deployment_name="gpt-4-deployment"
    )
4

Run Security Test

PYTHON
# Test your Azure deployment
result = await client.run_assessment(
    model_id="my-azure-gpt4",
    test_suites=["basic_security"]
)
print(f"🔍 Assessment: {result.assessment_id}")
⚡

GPT-3.5 Turbo

Cost-effective for development and testing

Most Popular4K Context
🧠

GPT-4

Best reasoning for production systems

Enterprise8K Context
🚀

GPT-4 Turbo

Large context for document analysis

Latest128K Context

Configuration Options

Advanced Setup

Additional configuration options for Azure deployments.

🔧

Multiple Deployments

PYTHON
# Register different Azure deployments
deployments = [
    ("azure-gpt35", "gpt-35-turbo"),
    ("azure-gpt4", "gpt-4"),
    ("azure-gpt4-turbo", "gpt-4-turbo")
]

for model_id, deployment in deployments:
    await client.register_azure_model(
        model_id=model_id,
        endpoint="https://your-resource.openai.azure.com/",
        deployment_name=deployment
    )
⚙️

Custom API Version

PYTHON
await client.register_azure_model(
    model_id="azure-gpt4-custom",
    endpoint="https://your-resource.openai.azure.com/",
    deployment_name="gpt-4",
    api_version="2024-02-15-preview",
    metadata={
        "region": "eastus",
        "environment": "production"
    }
)

Enterprise Features

🏢 Azure Advantages

Security & Compliance

•

Private network deployment

•

Customer-managed encryption keys

•

SOC 2, HIPAA, ISO compliance

•

Regional data residency

Enterprise Management

•

Azure AD integration

•

Role-based access control

•

Usage analytics & billing

•Dedicated capacity

Common Issues

⚠️ Troubleshooting

Invalid Endpoint
ConnectionError: Invalid endpoint URL
Solutions:
Deployment Not Found
DeploymentNotFound: The model deployment 'gpt-4' not found
Solutions:
  • • Check deployment name in Azure OpenAI Studio
  • • Ensure deployment is successfully created
  • • Verify deployment status is 'Succeeded'
Access Denied
Unauthorized: Access denied
Solutions:
  • • Verify API key from Keys and Endpoint section
  • • Check network restrictions and firewall rules
  • • Ensure correct Azure subscription

Quick Test

✅ Verify Your Setup

Run this test to confirm your Azure OpenAI integration is working:

PYTHON
import asyncio
from modelred import ModelRed

async def test_azure():
    async with ModelRed() as client:
        # Register Azure model
        await client.register_azure_model(
            model_id="test-azure",
            endpoint="https://your-resource.openai.azure.com/",
            deployment_name="gpt-35-turbo"
        )
        print("✅ Azure OpenAI model registered!")

        # Run security test
        result = await client.run_assessment(
            model_id="test-azure",
            test_suites=["basic_security"]
        )
        print(f"🔍 Assessment started: {result.assessment_id}")

asyncio.run(test_azure())

Next Steps

🧪

Run Security Tests

Start testing your Azure OpenAI deployments

🤖

Compare with OpenAI

See differences between Azure and standard OpenAI