How to generate text responses with Microsoft Foundry Models

Learn how to generate text responses from Foundry Models, such as Microsoft AI and DeepSeek models, by using the Responses API.

How to generate text responses with Microsoft Foundry Models

This article refers to the Microsoft Foundry (new) portal.

This article explains how to generate text responses for Foundry Models, such as Microsoft AI, DeepSeek, and Grok models, by using the Responses API. For a full list of the Foundry Models that support use of the Responses API, see Supported Foundry Models.

Prerequisites

To use the Responses API with deployed models in your application, you need:

• An Azure subscription. If you’re using GitHub Models, you can upgrade your experience and create an Azure subscription in the process. Read Upgrade from GitHub Models to Microsoft Foundry Models if that’s your case.

• A Foundry project. This kind of project is managed under a Foundry resource. If you don’t have a Foundry project, see Create a project for Microsoft Foundry.

• Your Foundry project’s endpoint URL, which is of the form https://YOUR-RESOURCE-NAME.services.ai.azure.com/api/projects/YOUR_PROJECT_NAME.

• A deployment of a Foundry Model, such as the MAI-DS-R1 model used in this article. If you don’t have a deployment already, see Add and configure Foundry Models to a model deployment to your resource.

Use the Responses API to generate text

Use the code in this section to make Responses API calls for Foundry Models. In the code samples, you create the client to consume the model and then send it a basic request.

Some of the SDK packages used in these samples are currently in preview. API surface and behavior might change before general availability.

When you deploy a model in the Foundry portal, you assign it a deployment name. Use this deployment name (not the model catalog ID) in the model parameter of your API calls.

Use keyless authentication with Microsoft Entra ID. To learn more about keyless authentication, see What is Microsoft Entra authentication? and DefaultAzureCredential.

Python

1. Install libraries, including the Azure Identity client library:

   pip install azure-identity
   pip install openai
   pip install --pre azure-ai-projects>=2.0.0b1 
   

2. Use the following code to configure the OpenAI client object in the project route, specify your deployment, and generate responses.

   import os
   from azure.identity import DefaultAzureCredential
   from azure.ai.projects import AIProjectClient
   
   project_client = AIProjectClient(
       endpoint="https://YOUR-RESOURCE-NAME.services.ai.azure.com/api/projects/YOUR_PROJECT_NAME",
       credential=DefaultAzureCredential(),
   )
   
   openai_client = project_client.get_openai_client()
   
   response = openai_client.responses.create(
       model="MAI-DS-R1", # Replace with your deployment name, not the model ID 
       input="What is the capital/major city of France?",
   )
   
   print(response.model_dump_json(indent=2)) 
   

C#

1. Install the Azure Identity client library:

   dotnet add package Azure.Identity
   

2. Use the following code to configure the OpenAI client object in the project route, specify your deployment, and generate responses.

   using Azure.Identity;
   using Azure.AI.Projects; 
   using Azure.AI.Projects.OpenAI;
   using OpenAI.Responses;
   
   #pragma warning disable OPENAI001
   
   const string deploymentName = "MAIDSR1"; // Replace with your deployment name, not the model ID 
   const string endpoint = "https://YOUR-RESOURCE-NAME.services.ai.azure.com/api/projects/YOUR_PROJECT_NAME";
   
   AIProjectClient projectClient = new(new Uri(endpoint), new DefaultAzureCredential());
   
   ProjectResponsesClient responseClient = projectClient.OpenAI.GetProjectResponsesClientForModel(deploymentName);
   
   ResponseResult response = responseClient.CreateResponse("What is the capital/major city of France?");
   
   Console.WriteLine($"[ASSISTANT]: {response.GetOutputText()}");
   

JavaScript

JavaScript v1 examples

1. Install the Azure Identity client library before you can use DefaultAzureCredential:

   npm install @azure/identity
   

2. Use the following code to configure the OpenAI client object in the project route, specify your deployment, and generate responses.

   import { AIProjectClient } from "@azure/ai-projects";
   import { DefaultAzureCredential } from "@azure/identity";
   
   const endpoint = "https://YOUR-RESOURCE-NAME.services.ai.azure.com/api/projects/YOUR_PROJECT_NAME";
   const deploymentName = "MAI-DS-R1"; // Replace with your deployment name, not the model ID 
   
   async function main() {
       const projectClient = new AIProjectClient(endpoint, new DefaultAzureCredential());
       const openAIClient = await projectClient.getOpenAIClient();
   
       const response = await openAIClient.responses.create({
           model: deploymentName,
           input: "What is the capital/major city of France?"
       });
       console.log(response.output_text);
   }
   
   main();
   

Java

Authentication with Microsoft Entra ID requires some initial setup. First, install the Azure Identity client library. For more options on how to install this library, see Azure Identity client library for Java.

1. Add the Azure Identity client library:

   <dependency>
       <groupId>com.azure</groupId>
       <artifactId>azure-identity</artifactId>
       <version>1.18.0</version>
   </dependency>
   

   After setup, choose which type of credential from azure.identity to use. For example, use DefaultAzureCredential to authenticate the client. Authentication is easiest with DefaultAzureCredential, as it finds the best credential to use in its running environment.

2. Use the following code to configure the OpenAI client object in the project route, specify your deployment, and generate responses.

   import com.azure.ai.agents;
   import com.azure.core.util.Configuration;
   import com.azure.identity.DefaultAzureCredentialBuilder;
   import com.openai.models.responses.Response;
   import com.openai.models.responses.ResponseCreateParams;
   
   public class Sample {
       public static void main(String[] args) {
           String endpoint = "https://YOUR-RESOURCE-NAME.services.ai.azure.com/api/projects/YOUR_PROJECT_NAME";
           String deploymentName = "MAI-DS-R1"; // Replace with your deployment name, not the model ID
   
           ResponsesClient responsesClient = new AgentsClientBuilder()
                   .credential(new DefaultAzureCredentialBuilder().build())
                   .endpoint(endpoint)
                   .serviceVersion(AgentsServiceVersion.V2025_11_15_PREVIEW)
                   .buildResponsesClient();
   
           ResponseCreateParams responseRequest = new ResponseCreateParams.Builder()
                   .input("What is the capital/major city of France?")
                   .model(deploymentName)
                   .build();
   
           Response response = responsesClient.getResponseService().create(responseRequest);
       }
   }
   

REST

curl -X POST https://YOUR-RESOURCE-NAME.services.ai.azure.com/api/projects/YOUR_PROJECT_NAME/openai/responses?api-version={{API_VERSION}} \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $AZURE_OPENAI_AUTH_TOKEN" \
-d '{
    "model": "MAI-DS-R1",
    "input": "What is the capital/major city of France?"
}'

The response includes the generated text along with model and usage metadata.

Supported Foundry Models

A selection of Foundry Models are supported for use with the Responses API.

View supported models in the Foundry portal

To see a full list of the supported models in the Foundry portal:

Sign in to Microsoft Foundry. Make sure the New Foundry toggle is on. These steps refer to Foundry (new).

1. Go to the Model catalog.
2. Filter the models by Capabilities and select Agent supported.

::: moniker-end

::: moniker range=“foundry” Sign in to Microsoft Foundry. Make sure the New Foundry toggle is on. These steps refer to Foundry (new).

1. Select Discover in the upper-right navigation, then Models in the left pane.
2. Open the Capabilities dropdown and select the Agent supported filter.

::: moniker-end

List of supported models

This section lists some of the Foundry Models that are supported for use with the Responses API. For the Azure OpenAI models that are supported, see Available Azure OpenAI models.

Models sold directly by Azure:

• MAI-DS-R1: Deterministic, precision-focused reasoning.
• grok-4: Frontier-scale reasoning for complex, multiple-step problem solving.
• grok-4-fast-reasoning: Accelerated agentic reasoning optimized for workflow automation.
• grok-4-fast-non-reasoning: High-throughput, low-latency generation and system routing.
• grok-3: Strong reasoning for complex, system-level workflows.
• grok-3-mini: Lightweight model optimized for interactive, high-volume use cases.
• Llama-3.3-70B-Instruct: Versatile model for enterprise Q&A, decision support, and system orchestration.
• Llama-4-Maverick-17B-128E-Instruct-FP8: FP8-optimized model that delivers fast, cost-efficient inference.
• DeepSeek-V3-0324: Multimodal understanding across text and images.
• DeepSeek-V3.1: Enhanced multimodal reasoning and grounded retrieval.
• DeepSeek-R1-0528: Advanced long-form and multiple-step reasoning.
• gpt-oss-120b: Open-ecosystem model that supports transparency and reproducibility.

Troubleshoot common errors

Error	Cause	Resolution
401 Unauthorized	Invalid or expired credential	Verify your DefaultAzureCredential has the Cognitive Services OpenAI User role assigned on the resource.
404 Not Found	Wrong endpoint or deployment name	Confirm your endpoint URL includes /api/projects/YOUR_PROJECT_NAME and the deployment name matches your Foundry portal.
400 Model not supported	Model doesn’t support Responses API	Check the supported models list and verify your deployment uses a compatible model.

Related content

• Migrate from Azure AI Inference SDK to OpenAI SDK
• Azure OpenAI supported programming languages
• Switch between OpenAI and Azure OpenAI endpoints
• Model support for v1 Azure OpenAI API