Integrate Azure AI services with Service Connector

This page shows supported authentication methods and clients, and shows sample code you can use to connect Azure AI Services to other cloud services using Service Connector. You might still be able to connect to Azure AI Services using other methods. This page also shows default environment variable names and values you get when you create the service connection.

Supported compute services

Service Connector can be used to connect the following compute services to Azure AI Services:

  • Azure App Service
  • Azure Container Apps
  • Azure Functions
  • Azure Kubernetes Service (AKS)
  • Azure Spring Apps

Supported authentication types and client types

The table below shows which combinations of authentication methods and clients are supported for connecting your compute service to Azure AI Services using Service Connector. A “Yes” indicates that the combination is supported, while a “No” indicates that it is not supported.

Client type System-assigned managed identity User-assigned managed identity Secret/connection string Service principal
.NET Yes Yes Yes Yes
Java Yes Yes Yes Yes
Node.js Yes Yes Yes Yes
Python Yes Yes Yes Yes
None Yes Yes Yes Yes

This table indicates that all combinations of client types and authentication methods in the table are supported. All client types can use any of the authentication methods to connect to Azure AI Services using Service Connector.

Default environment variable names or application properties and sample code

Use the connection details below to connect compute services to Azure AI Services. For more information about naming conventions, check the Service Connector internals article.

System-assigned managed identity

Default environment variable name Description Sample value
AZURE_AISERVICES_OPENAI_BASE Azure OpenAI endpoint https://<your-Azure-AI-Services-endpoint>.openai.azure.com/
AZURE_AISERVICES_COGNITIVESERVICES_ENDPOINT Azure Cognitive Services token provider service https://<your-Azure-AI-Services-endpoint>.cognitiveservices.azure.com/
AZURE_AISERVICES_SPEECH_ENDPOINT Speech to Text (Standard) API endpoint https://<location>.stt.speech.microsoft.com

Sample code

Refer to the steps and code below to connect to Azure AI Services using a system-assigned managed identity.

You can use the Azure client library to access various cognitive APIs that Azure AI Services support. We use Azure AI Text Analytics as an example in this sample. Refer to Authenticate requests to Azure AI services to call the cognitive APIs directly.

  1. Install the following dependencies. We use Azure.AI.TextAnalytics as an example.

    dotnet add package Azure.AI.TextAnalytics
    dotnet add package Azure.Identity
    
  2. Authenticate using the Azure Identity library and get the Azure AI Services endpoint from the environment variables added by Service Connector. When using the code below, uncomment the part of the code snippet for the authentication type you want to use.

    using Azure.AI.TextAnalytics;
    using Azure.Identity;
    
    string endpoint = Environment.GetEnvironmentVariable("AZURE_AISERVICES_COGNITIVESERVICES_ENDPOINT");
    
    // Uncomment the following lines corresponding to the authentication type you want to use.
    // system-assigned managed identity
    // var credential = new DefaultAzureCredential();
    
    // user-assigned managed identity
    // var credential = new DefaultAzureCredential(
    //     new DefaultAzureCredentialOptions
    //     {
    //         ManagedIdentityClientId = Environment.GetEnvironmentVariable("AZURE_AISERVICES_CLIENTID");
    //     });
    
    // service principal 
    // var tenantId = Environment.GetEnvironmentVariable("AZURE_AISERVICES_TENANTID");
    // var clientId = Environment.GetEnvironmentVariable("AZURE_AISERVICES_CLIENTID");
    // var clientSecret = Environment.GetEnvironmentVariable("AZURE_AISERVICES_CLIENTSECRET");
    // var credential = new ClientSecretCredential(tenantId, clientId, clientSecret);
    
    TextAnalyticsClient languageServiceClient = new(
      new Uri(endpoint),
      credential);
    

User-assigned managed identity

Default environment variable name Description Sample value
AZURE_AISERVICES_OPENAI_BASE Azure OpenAI endpoint https://<your-Azure-AI-Services-endpoint>.openai.azure.com/
AZURE_AISERVICES_COGNITIVESERVICES_ENDPOINT Azure Cognitive Services token provider service https://<your-Azure-AI-Services-endpoint>.cognitiveservices.azure.com/
AZURE_AISERVICES_SPEECH_ENDPOINT Speech to Text (Standard) API endpoint https://<location>.stt.speech.microsoft.com
AZURE_AISERVICES_CLIENTID Your client ID <client-ID>

Sample code

Refer to the steps and code below to connect to Azure AI Services using a user-assigned managed identity.

You can use the Azure client library to access various cognitive APIs that Azure AI Services support. We use Azure AI Text Analytics as an example in this sample. Refer to Authenticate requests to Azure AI services to call the cognitive APIs directly.

  1. Install the following dependencies. We use Azure.AI.TextAnalytics as an example.

    dotnet add package Azure.AI.TextAnalytics
    dotnet add package Azure.Identity
    
  2. Authenticate using the Azure Identity library and get the Azure AI Services endpoint from the environment variables added by Service Connector. When using the code below, uncomment the part of the code snippet for the authentication type you want to use.

    using Azure.AI.TextAnalytics;
    using Azure.Identity;
    
    string endpoint = Environment.GetEnvironmentVariable("AZURE_AISERVICES_COGNITIVESERVICES_ENDPOINT");
    
    // Uncomment the following lines corresponding to the authentication type you want to use.
    // system-assigned managed identity
    // var credential = new DefaultAzureCredential();
    
    // user-assigned managed identity
    // var credential = new DefaultAzureCredential(
    //     new DefaultAzureCredentialOptions
    //     {
    //         ManagedIdentityClientId = Environment.GetEnvironmentVariable("AZURE_AISERVICES_CLIENTID");
    //     });
    
    // service principal 
    // var tenantId = Environment.GetEnvironmentVariable("AZURE_AISERVICES_TENANTID");
    // var clientId = Environment.GetEnvironmentVariable("AZURE_AISERVICES_CLIENTID");
    // var clientSecret = Environment.GetEnvironmentVariable("AZURE_AISERVICES_CLIENTSECRET");
    // var credential = new ClientSecretCredential(tenantId, clientId, clientSecret);
    
    TextAnalyticsClient languageServiceClient = new(
      new Uri(endpoint),
      credential);
    

Connection string

Default environment variable name Description Sample value
AZURE_AISERVICES_OPENAI_BASE Azure OpenAI endpoint https://<your-Azure-AI-Services-endpoint>.openai.azure.com/
AZURE_AISERVICES_COGNITIVESERVICES_ENDPOINT Azure Cognitive Services token provider service https://<your-Azure-AI-Services-endpoint>.cognitiveservices.azure.com/
AZURE_AISERVICES_SPEECH_ENDPOINT Speech to Text (Standard) API endpoint https://<location>.stt.speech.microsoft.com
AZURE_AISERVICES_KEY Azure AI Services API key <api-key>

Sample Code

Refer to the steps and code below to connect to Azure AI Services using a connection string.

You can use the Azure client library to access various cognitive APIs that Azure AI Services support. We use Azure AI Text Analytics as an example in this sample. Refer to Authenticate requests to Azure AI services to call the cognitive APIs directly.

  1. Install the following dependencies. We use Azure.AI.TextAnalytics as an example.

    dotnet add package Azure.AI.TextAnalytics
    dotnet add package Azure.Core --version 1.40.0
    
  2. Get the Azure AI Services endpoint and key from the environment variables added by Service Connector.

    using Azure.AI.TextAnalytics;
    
    string endpoint = Environment.GetEnvironmentVariable("AZURE_AISERVICES_COGNITIVESERVICES_ENDPOINT")
    string key = Environment.GetEnvironmentVariable("AZURE_AISERVICES_KEY");
    
    TextAnalyticsClient languageServiceClient = new(
      new Uri(endpoint),
      new AzureKeyCredential(key));
    

Service principal

Default environment variable name Description Sample value
AZURE_AISERVICES_OPENAI_BASE Azure OpenAI endpoint https://<your-Azure-AI-Services-endpoint>.openai.azure.com/
AZURE_AISERVICES_COGNITIVESERVICES_ENDPOINT Azure Cognitive Services token provider service https://<your-Azure-AI-Services-endpoint>.cognitiveservices.azure.com/
AZURE_AISERVICES_SPEECH_ENDPOINT Speech to Text (Standard) API endpoint https://<location>.stt.speech.microsoft.com
AZURE_AISERVICES_CLIENTID Your client ID <client-ID>
AZURE_AISERVICES_CLIENTSECRET Your client secret <client-secret>
AZURE_AISERVICES_TENANTID Your tenant ID <tenant-ID>

Sample code

Refer to the steps and code below to connect to Azure AI Services using a service principaL.

You can use the Azure client library to access various cognitive APIs that Azure AI Services support. We use Azure AI Text Analytics as an example in this sample. Refer to Authenticate requests to Azure AI services to call the cognitive APIs directly.

  1. Install the following dependencies. We use Azure.AI.TextAnalytics as an example.

    dotnet add package Azure.AI.TextAnalytics
    dotnet add package Azure.Identity
    
  2. Authenticate using the Azure Identity library and get the Azure AI Services endpoint from the environment variables added by Service Connector. When using the code below, uncomment the part of the code snippet for the authentication type you want to use.

    using Azure.AI.TextAnalytics;
    using Azure.Identity;
    
    string endpoint = Environment.GetEnvironmentVariable("AZURE_AISERVICES_COGNITIVESERVICES_ENDPOINT");
    
    // Uncomment the following lines corresponding to the authentication type you want to use.
    // system-assigned managed identity
    // var credential = new DefaultAzureCredential();
    
    // user-assigned managed identity
    // var credential = new DefaultAzureCredential(
    //     new DefaultAzureCredentialOptions
    //     {
    //         ManagedIdentityClientId = Environment.GetEnvironmentVariable("AZURE_AISERVICES_CLIENTID");
    //     });
    
    // service principal 
    // var tenantId = Environment.GetEnvironmentVariable("AZURE_AISERVICES_TENANTID");
    // var clientId = Environment.GetEnvironmentVariable("AZURE_AISERVICES_CLIENTID");
    // var clientSecret = Environment.GetEnvironmentVariable("AZURE_AISERVICES_CLIENTSECRET");
    // var credential = new ClientSecretCredential(tenantId, clientId, clientSecret);
    
    TextAnalyticsClient languageServiceClient = new(
      new Uri(endpoint),
      credential);
    

Next steps

Follow the tutorial listed below to learn more about Service Connector.