Skip to main content
Version: 2026 R1

Azure AI Foundry

Azure AI Foundry is Microsoft's platform providing access to multiple AI models - from proven GPT-4o, through economical GPT-3.5-turbo, to the newest models like DeepSeek. It offers enterprise-grade SLA, Microsoft technical support, and easy integration with the Azure ecosystem.

When to choose Azure AI Foundry​

Already using Azure:

  • Infrastructure in Azure
  • Centralized cost and subscription management
  • Integration with Azure AD, Key Vault, Application Insights

Enterprise requirements:

  • 99.9% SLA for models
  • Microsoft technical support
  • Compliance with corporate policies

Flexibility in model selection:

  • Access to OpenAI models (GPT-4o, GPT-4o-mini)
  • Open-source models (DeepSeek, Llama, Mistral)
  • Model Router - automatic selection of best model

Cost control:

  • Shared TPM limits for multiple deployments
  • Detailed usage reports in Azure Cost Management
  • Ability to set budgets and alerts

Security and compliance:

  • Data remains in selected Azure region
  • GDPR and other regulations compliance
  • Private endpoints and VNET integration

Requirements​

  • Active Azure subscription
  • Azure AI Foundry workspace with deployed model
  • API Key and Endpoint URL

Step 1: Azure preparation​

Get access credentials​

  1. Log in to Azure Portal
  2. Go to your Azure AI Foundry workspace
  3. In the menu select Keys and Endpoint
  4. Copy:
    • Key (API Key)
    • Endpoint OpenAI URL

Deploy model​

  1. In workspace go to Deployments
  2. Click Create new deployment
  3. Select a model (e.g., GPT-4o, GPT-4o-mini, or other model supporting responses)
  4. Assign a deployment name (e.g., gpt-4o-mini)
  5. Click Create
Supported models

You can use various models available in Azure AI Foundry:

  • GPT models (gpt-4o, gpt-4o-mini, gpt-35-turbo)
  • Model-router (automatic model selection)
  • DeepSeek and other models supporting responses

Step 2: AI Proxy Configuration​

Example aiconfiguration.json​

{
  "ProviderConnections": {
    "AzureFoundry": {
      "Description": "Azure AI Foundry Connection",
      "Type": "AzureAi",
      "ProviderConfiguration": {
        "ApiKey": "your-azure-api-key-here",
        "Endpoint": "https://your-workspace.openai.azure.com/"
      }
    }
  },
  "ProviderModels": [
    {
      "ConnectionName": "AzureFoundry",
      "Priority": 100,
      "Name": "Azure GPT-4o-mini",
      "Description": "",
      "TextModel": {
        "ModelName": "gpt-4o-mini"
      },
      "ImageModel": {
        "ModelName": "gpt-4o-mini"
      },
      "EmbeddingModel": {
        "ModelName": "text-embedding-3-small"
      }
    }
  ],
  "MethodTypesConfiguration": {
    "ConciergePrompt": [ "Azure GPT-4o-mini" ],
    "ConciergeExecuteTool": [ "Azure GPT-4o-mini" ]
  }
}
Important

In the ModelName field enter the deployment name you created in Azure (not the model name). If you named the deployment my-gpt4, use "ModelName": "my-gpt4".

Example docker-compose.yml​

name: aiproxy_containers
services:
  ai-proxy:
    image: webconbps/aiproxy:1.0.0.235
    container_name: ai-proxy
    restart: unless-stopped
    ports:
      - "5298:8080"
      - "7033:8081"
    environment:
      - ASPNETCORE_ENVIRONMENT=Production
      - AppConfiguration__SelfHosted__Certificate__Path=/app/https/certificate.pem
      - Logging__LogLevel__Default=Information
      - Logging__LogLevel__Microsoft=Warning
    volumes:
      - ./certificates/certificate.pem:/app/https/certificate.pem:ro
      - ./aiconfiguration.json:/app/aiconfiguration.json:ro

Step 3: Startup​

# Make sure you have prepared files
# - ./certificates/certificate.pem
# - ./aiconfiguration.json (with filled data)

# Run container
docker-compose up -d

# Check logs
docker-compose logs -f ai-proxy

Troubleshooting​

Error: 401 Unauthorized​

Causes:

  • Invalid API Key
  • Invalid Endpoint URL

Solution:

# Check if key and endpoint are correct in aiconfiguration.json
# Verify in Azure Portal > Keys and Endpoint
# Restart container
docker-compose restart ai-proxy

Error: 404 Not Found / Model not found​

Causes:

  • Deployment name in ModelName is incorrect
  • Deployment doesn't exist or is not active

Solution:

# Check deployment name in Azure Portal > Deployments
# Make sure deployment has "Succeeded" status
# Update ModelName in aiconfiguration.json
# Restart container
docker-compose restart ai-proxy

Error: 429 Too Many Requests​

Cause:

  • Exceeded TPM (Tokens Per Minute) limit for deployment

Solution:

  • Wait a moment before next requests
  • Consider increasing TPM limits in Azure for your deployment

When creating deployment you can choose:

  • gpt-4o - newest GPT-4 Optimized model
  • gpt-4o-mini - faster and cheaper version
  • gpt-4-turbo - GPT-4 with larger context
  • gpt-35-turbo - GPT-3.5 model (in Azure called 35 instead of 3.5)
  • text-embedding-3-small - for creating embeddings
  • model-router - automatic selection of best model
  • deepseek and other models supporting responses
info

All models must support responses API to work correctly with AI Proxy.