Microsoft Foundry SDKs and Endpoints

Note

This document refers to the Microsoft Foundry (classic) portal.

🔄 Switch to the Microsoft Foundry (new) documentation if you're using the new portal.

This article describes the SDKs and endpoints you can use with your Foundry resource. It shows you how to connect to your project, access models from different providers, and use Foundry Tools. The SDK offers a unified way to work with AI resources through client libraries in multiple programming languages.

The Microsoft Foundry SDK simplifies AI application development on Azure. It lets developers:

• Access models from various providers through one interface
• Combine models, data, and Foundry Tools to build AI-powered applications
• Evaluate, debug, and improve application quality and safety in development, testing, and production

The Microsoft Foundry SDK integrates with other client libraries and services that work together.

Foundry SDK

Developers working with Microsoft Foundry need flexibility to integrate multiple AI capabilities into unified workflows. These SDKs provide the building blocks for provisioning resources, orchestrating agents, and connecting to specialized Foundry Tools. By choosing the right library, you can streamline development, reduce complexity, and ensure your solutions scale across Foundry projects and external endpoints.

Note

This article applies to a Foundry project. The code shown here doesn't work for a hub-based project. For more information, see Types of projects.

Prerequisites

• An Azure account with an active subscription. If you don't have one, create a free Azure account, which includes a free trial subscription.

• Create a Foundry project if you don't have one already.
• Microsoft Foundry Models allows customers to consume the most powerful models from flagship model providers using a single endpoint and credentials. This means that you can switch between models and consume them from your application without changing a single line of code.
  Copy the Foundry project endpoint in the Overview section of your project. You'll use it in a moment.

  Tip

  If you don't see the Foundry project endpoint, you're using a hub-based project. (See Types of projects). Switch to a Foundry project, or use the preceding steps to create one.

• Sign in with the Azure CLI using the same account that you use to access your project:

  az login
  

The following examples show how to authenticate and create a client for your project endpoint.

Tip

These code samples are starting points. Use these clients to interact with models, run evaluations, and more, as explained in the client libraries section.

The Azure AI Projects client library for Python is a unified library that enables you to use multiple client libraries together by connecting to a single project endpoint.

• Install the project client library

  pip install azure-ai-projects azure-identity openai
  

• Create a project client in code. Copy the Foundry project endpoint from the Overview page of the project and update the endpoint string value.

  from azure.identity import DefaultAzureCredential
  from azure.ai.projects import AIProjectClient
  
  project = AIProjectClient(
    endpoint="your_project_endpoint",  # Replace with your endpoint
    credential=DefaultAzureCredential())
  # The AIProjectClient lets you access models, data, and services in your project.
  

The Azure AI Projects client library for Java (preview) is a unified library that enables you to use multiple client libraries together by connecting to a single project endpoint.

Important

Items marked (preview) in this article are currently in public preview. This preview is provided without a service-level agreement, and we don't recommend it for production workloads. Certain features might not be supported or might have constrained capabilities. For more information, see Supplemental Terms of Use for Microsoft Azure Previews.

• Add these packages to your installation:

  • com.azure.ai.projects
  • com.azure.core

• Create a project client in code. Copy the Foundry project endpoint from the Overview page of the project and update the connections string value.

  import com.azure.ai.projects.ProjectsClient;
  import com.azure.ai.projects.ProjectsClientBuilder;
  import com.azure.core.credential.AzureKeyCredential;
  
  String endpoint ="your_project_endpoint"; // Replace with your endpoint
  
  ProjectsClient projectClient = new ProjectsClientBuilder()
      .credential(new DefaultAzureCredential())
      .endpoint(endpoint)
      .buildClient();
  // The ProjectsClient enables unified access to your project's resources.
  

The Azure AI Projects client library for JavaScript is a unified library that enables you to use multiple client libraries together by connecting to a single project endpoint.

• Install dependencies (preview):

  npm install @azure/ai-projects @azure/identity
  

• Create a project client in code. Copy the Foundry project endpoint from the Overview page of the project and update the endpoint string value.

  import { AIProjectClient } from '@azure/ai-projects';
  import { DefaultAzureCredential } from '@azure/identity';
  
  const endpoint = "your_project_endpoint"; // Replace with your actual endpoint
  
  const project = new AIProjectClient(endpoint, new DefaultAzureCredential());
  // The AIProjectClient lets you access models, data, and services in your project.
  

The Azure AI Projects client library for .NET is a unified library that enables you to use multiple client libraries together by connecting to a single project endpoint.

• Install packages:

  dotnet add package Azure.Identity
  dotnet add package Azure.Core
  dotnet add package OpenAI
  

• Create a project client in code. Copy the Foundry project endpoint from the Overview page of the project and update the endpointUrl string value.

  using Azure.Identity;
  using Azure.Core;
  using Azure.Core.Pipeline;
  using Azure.AI.Projects;
  using System;
  
  string endpointUrl = "your_project_endpoint"; // Replace with your endpoint
  
  DefaultAzureCredential credential = new();
  BearerTokenPolicy tokenPolicy = new(credential, "https://cognitiveservices.azure.com/.default");
  
  AIProjectClientOptions clientOptions = new AIProjectClientOptions();
  // The PerRetry position ensures the authentication policy is applied to every retry attempt.
  // This is important for robust authentication in distributed/cloud environments.
  clientOptions.AddPolicy(tokenPolicy, HttpPipelinePosition.PerRetry);
  
  AIProjectClient projectClient = new(new Uri(endpointUrl), new DefaultAzureCredential(), clientOptions);
  // The AIProjectClient lets you access models, data, and services in your project.
  

OpenAI SDK

The OpenAI SDK lets you interact with the Azure OpenAI service. It offers a simple interface for making API calls and managing authentication. The OpenAI SDK directly calls the Azure OpenAI endpoint. The following code snippet shows how to create the OpenAI client from the Project client for proper scoping and context management.

Which endpoint should you use?

• Managing a Project or calling Agents v2? Use the Foundry Project endpoint with the Foundry SDK. Get your OpenAI client from the Project using Microsoft Entra ID for authentication.
• Calling a model directly? Use the Azure OpenAI endpoint with the OpenAI SDK with Microsoft Entra ID as the preferred authentication method. If using API keys, choose the v1 endpoint: https://<YOUR-RESOURCE-NAME>.openai.azure.com/openai/v1/.

Create an OpenAI client from your project

# Use the AIProjectClient to create an OpenAI client for your project
openai_client = project.get_openai_client(api_version="api_version")
response = openai_client.responses.create(
    model="gpt-4.1-mini",
    input="What is the size of France in square miles?",
)
print(f"Response output: {response.output_text}")

The following code snippet demonstrates how to use the Azure OpenAI v1 endpoint with the OpenAI client for responses.

from openai import OpenAI
from azure.identity import DefaultAzureCredential, get_bearer_token_provider

token_provider = get_bearer_token_provider(
    DefaultAzureCredential(), "https://cognitiveservices.azure.com/.default"
)

client = OpenAI(  
  base_url = "https://<YOUR-RESOURCE-NAME>.openai.azure.com/openai/v1/",  
  api_key=token_provider,
)

response = client.responses.create(
    model="model_deployment_name",
    input= "What is the size of France in square miles?" 
)

print(response.model_dump_json(indent=2)) 

For more information on using the OpenAI SDK, see Azure OpenAI supported programming languages.

Important

Items marked (preview) in this article are currently in public preview. This preview is provided without a service-level agreement, and we don't recommend it for production workloads. Certain features might not be supported or might have constrained capabilities. For more information, see Supplemental Terms of Use for Microsoft Azure Previews.

// 
OpenAIClient openAIClient = projectClient.getOpenAIClient();

For more information on using the OpenAI SDK, see Azure OpenAI supported programming languages.

// Use the AIProjectClient to create an OpenAI client for your project
const openAIClient = await project.getOpenAIClient();

For more information on using the OpenAI SDK, see Azure OpenAI supported programming languages.

1. Install the OpenAI package:

   dotnet add package OpenAI
   

2. The following code snippet demonstrates how to create the OpenAI client directly using the Azure OpenAI v1 endpoint.

   using Azure.Identity;
   using Azure.Core;
   using Azure.Core.Pipeline;   
   using OpenAI;
   using System;
   using System.ClientModel.Primitives;
   
   endpointUrl = "https://<YOUR-RESOURCE-NAME>.openai.azure.com/openai/v1/"
   
   DefaultAzureCredential credential = new();
   BearerTokenPolicy tokenPolicy = new(credential, "https://cognitiveservices.azure.com/.default");
   
   OpenAIClientOptions clientOptions = new()
   {
       Endpoint = new Uri(endpointUrl)
   };
   
   // The PerRetry position ensures the authentication policy is applied to every retry attempt.
   // This is important for robust authentication in distributed/cloud environments.
   clientOptions.AddPolicy(tokenPolicy, HttpPipelinePosition.PerRetry);
   
   var projectClient = new ResponseClient(
       endpointUrl, 
       credential,
       clientOptions
   );
   // The ResponseClient lets you interact with models and services in your project.
   

For more information on using the OpenAI SDK, see Azure OpenAI supported programming languages.

After you create a client, use it to access models, run evaluations, and connect to other Foundry Tools.

• Using the project endpoint, you can:
  • Use Foundry Models, including Azure OpenAI
  • Use Foundry Agent Service
  • Run evaluations in the cloud
  • Enable tracing for your app
  • Fine tune a model
  • Retrieve endpoints and keys for external resource connections, such as Foundry Tools, local orchestration, and more.

The next section lists the Foundry Tools client libraries and shows how to use them.

Foundry Tools SDKs

To use Foundry Tools, you can use the following SDKs with the endpoints listed.

Which endpoint should you use?

Choose an endpoint based on your needs:

Use the Azure AI Services endpoint to access Computer Vision, Content Safety, Document Intelligence, Language, Translation, and Token Foundry Tools.

Azure AI Services endpoint: https://<YOUR-RESOURCE-NAME>.services.ai.azure.com/

For Speech and Translation Foundry Tools, use the endpoints in the following tables. Replace placeholders with your resource information.

Speech Endpoints

Foundry Tool	Endpoint
Speech to Text (Standard)	https://<YOUR-RESOURCE-REGION>.stt.speech.microsoft.com
Text to Speech (Neural)	https://<YOUR-RESOURCE-REGION>.tts.speech.microsoft.com
Custom Voice	https://<YOUR-RESOURCE-NAME>.cognitiveservices.azure.com/

Translation Endpoints

Foundry Tool	Endpoint
Text Translation	https://api.cognitive.microsofttranslator.com/
Document Translation	https://<YOUR-RESOURCE-NAME>.cognitiveservices.azure.com/

The following sections include quickstart links for the Foundry Tools SDKs and reference information.

C# supported Foundry Tools

Foundry Tool	Description	Quickstarts and reference documentation
Speech	Add speech to text, text to speech, translation, and speaker recognition capabilities to applications.	• Speech to text quickstart • Text to speech quickstart • Speech translation quickstart • Speech SDK for .NET • Speech NuGet package (Speech CLI)
Content Safety	Detect harmful content in applications and services.	• Analyze text content quickstart • Use a text blocklist quickstart • Analyze image content quickstart • Content Safety SDK for .NET • Content Safety NuGet package
Vision	Analyze content in digital images and rich media assets.	• Azure Vision in Foundry Tools v3.2 GA Read quickstart • Image Analysis quickstart • Use the Face service quickstart • Vision SDK for .NET • Vision NuGet package
Language	Build applications with natural language understanding capabilities.	• Custom question answering (CQA) quickstart • Entity linking quickstart • Language detection quickstart • Key Phrase extraction quickstart • Detecting named entities (NER) quickstart • Detect Personally Identifiable Information (PII) quickstart • Sentiment analysis and opinion mining quickstart • Using text, document and conversation summarization quickstart • Using Text Analytics for health quickstart • Language SDK for .NET (text analysis) • Language NuGet package (text analysis) • Language SDK for .NET (Question Answering) • Language NuGet package (question answering)
Translator	Use AI-powered translation technology to translate more than 100 in-use, at-risk, and endangered languages and dialects.	• Translator SDK for .NET (text) • Translator NuGet package (text) • Translator SDK for .NET (batch) • Translator NuGet package (batch)
Document Intelligence	Turn documents into intelligent data-driven solutions.	• Document Intelligence quickstart • Document Intelligence SDK for .NET • Document Intelligence NuGet package
Azure AI Search	Bring AI-powered cloud search to your mobile and web apps.	• Use agentic retrieval quickstart • Vector search quickstart • Classic generative search (RAG) using grounding data quickstart • Full-text search quickstart • Semantic ranking quickstart • Chat with Azure OpenAI models using your own data quickstart • Azure AI Search SDK for .NET • Azure AI Search NuGet package

Java supported Foundry Tools

Foundry Tool	Description	Quickstarts and reference documentation
Speech	Add speech to text, text to speech, translation, and speaker recognition capabilities to applications.	• Speech to text quickstart • Text to speech quickstart • Speech translation quickstart • Speech SDK for Java • Speech Maven package
Content Safety	Detect harmful content in applications and services.	• Analyze text content quickstart • Use a text blocklist quickstart • Analyze image content quickstart • Content Safety SDK for Java • Content Safety Maven package
Vision	Analyze content in digital images and rich media assets.	• Image Analysis quickstart • Use the Face service quickstart • Vision SDK for Java • Vision Maven package
Language	Build applications with natural language understanding capabilities.	• Entity linking quickstart • Language detection quickstart • Key Phrase extraction quickstart • Detecting named entities (NER) quickstart • Detect Personally Identifiable Information (PII) quickstart • Sentiment analysis and opinion mining quickstart • Using text, document and conversation summarization quickstart • Using Text Analytics for health quickstart • Language SDK for Java (text analysis) • Language Maven package
Translator	Use AI-powered translation technology to translate more than 100 in-use, at-risk, and endangered languages and dialects.	• Translator SDK for Java (text) • Translator Maven package (text)
Document Intelligence	Turn documents into intelligent data-driven solutions.	• Document Intelligence quickstart • Document Intelligence SDK for Java • Document Intelligence Maven package
Azure AI Search	Bring AI-powered cloud search to your mobile and web apps.	• Use agentic retrieval quickstart • Vector search quickstart • Classic generative search (RAG) using grounding data quickstart • Full-text search quickstart • Semantic ranking quickstart • Chat with Azure OpenAI models using your own data quickstart • Azure AI Search SDK for Java • Azure AI Search Maven package

JavaScript supported Foundry Tools

Foundry Tool	Description	Reference documentation
Speech	Add speech to text, text to speech, translation, and speaker recognition capabilities to applications.	• Speech to text quickstart • Text to speech quickstart • Speech translation quickstart • Speech SDK for JavaScript • Speech npm package
Content Safety	Detect harmful content in applications and services.	• Analyze text content quickstart • Use a text blocklist quickstart • Analyze image content quickstart • Content Safety npm package
Vision	Analyze content in digital images and rich media assets.	• Azure Vision in Foundry Tools v3.2 GA Read quickstart • Image Analysis quickstart • Use the Face service quickstart • Vision SDK for JavaScript • Vision npm package
Language	Build applications with natural language understanding capabilities.	• Entity linking quickstart • Language detection quickstart • Key Phrase extraction quickstart • Detecting named entities (NER) quickstart • Detect Personally Identifiable Information (PII) quickstart • Sentiment analysis and opinion mining quickstart • Using text, document and conversation summarization quickstart • Using Text Analytics for health quickstart • Language SDK for JavaScript (text analysis) • Language npm package
Translator	Use AI-powered translation technology to translate more than 100 in-use, at-risk, and endangered languages and dialects.	• Translator SDK for JavaScript (text) • Translator npm package (text)
Document Intelligence	Turn documents into intelligent data-driven solutions.	• Document Intelligence quickstart • Document Intelligence SDK for JavaScript • Document Intelligence npm package
Azure AI Search	Bring AI-powered cloud search to your mobile and web apps.	• Use agentic retrieval quickstart • Vector search quickstart • Classic generative search (RAG) using grounding data quickstart • Full-text search quickstart • Semantic ranking quickstart • Chat with Azure OpenAI models using your own data quickstart • Azure AI Search SDK for JavaScript • Azure AI Search npm package

Python supported Foundry Tools

Foundry Tool	Description	Quickstarts and reference documentation
Speech	Add speech to text, text to speech, translation, and speaker recognition capabilities to applications.	• Speech to text quickstart • Text to speech quickstart • Speech translation quickstart • Speech SDK for Python • Speech PyPi package
Content Safety	Detect harmful content in applications and services.	• Analyze text content quickstart • Use a text blocklist quickstart • Analyze image content quickstart • Content Safety SDK for Python • Content Safety PyPi package
Vision	Analyze content in digital images and rich media assets.	• Azure Vision in Foundry Tools v3.2 GA Read quickstart • Image Analysis quickstart • Use the Face service quickstart • Vision SDK for Python • Vision PyPi package
Language	Build applications with natural language understanding capabilities.	• Custom question answering (CQA) quickstart • Entity linking quickstart • Language detection quickstart • Key Phrase extraction quickstart • Detect named entities (NER) quickstart • Detect Personally Identifiable Information (PII) quickstart • Sentiment analysis and opinion mining quickstart • Using text, document and conversation summarization quickstart • Using Text Analytics for health quickstart • Language SDK for Python (text analysis) • Language PyPi package (text analysis) • Language SDK for Python (question answering) • Language PyPi package (question answering) • Language SDK for Python (language conversations) • Language PyPi package (language conversations)
Translator	Use AI-powered translation technology to translate more than 100 in-use, at-risk, and endangered languages and dialects.	• Translator SDK for Python (text) • Translator PyPi package (text) • Translator SDK for Python (batch) • Translator PyPi package (batch)
Document Intelligence	Turn documents into intelligent data-driven solutions.	• Document Intelligence quickstart • Document Intelligence SDK for Python • Document Intelligence PyPi package
Azure AI Search	Bring AI-powered cloud search to your mobile and web apps.	• Connect to a search service quickstart • Use agentic retrieval quickstart • Vector search quickstart • Classic generative search (RAG) using grounding data quickstart • Full-text search quickstart • Semantic ranking quickstart • Chat with Azure OpenAI models using your own data quickstart • Azure AI Search SDK for Python • Azure AI Search PyPi package

Using the Agent Framework for local orchestration

Microsoft Agent Framework is an open-source development kit for building AI agents and multi-agent workflows for .NET and Python. It provides a way to build and manage AI agents that can interact with users and other services. It can orchestrate agents in Foundry, or have local agents that use Foundry models.

For more information, see the Microsoft Agent Framework overview