What is Microsoft Foundry?

Microsoft Foundry is a trusted platform that empowers developers to drive innovation and shape the future with AI in a safe, secure, and responsible way.

Microsoft Foundry is a unified Azure platform-as-a-service offering for enterprise AI operations, model builders, and application development. This foundation combines production-grade infrastructure with friendly interfaces, enabling developers to focus on building applications rather than managing infrastructure.

Microsoft Foundry unifies agents, models, and tools under a single management grouping with built-in enterprise-readiness capabilities including tracing, monitoring, evaluations, and customizable enterprise setup configurations. The platform provides streamlined management through unified role-based access control (RBAC), networking, and policies under one Azure resource provider namespace.

• Coming from Azure OpenAI? Upgrade your Azure OpenAI resource to a Foundry resource while preserving your endpoint, API keys, and existing state.
• Using hub-based projects? Hub-based projects are accessible in the Foundry (classic) portal. New investments are focused on Foundry projects in the new portal.

Evolution of Foundry

Foundry consolidates several previous Azure AI services and tools into a unified platform. The following table maps previous concepts to their current equivalents. For detailed guidance on transitioning, see Navigate from classic to the new experience.

Dimension	Previous	Current
Brand	Azure AI Studio / Azure AI Foundry	Microsoft Foundry
Brand	Azure AI Services	Foundry Tools
Portal	Foundry (classic)	Foundry
Agent API	Assistants API (Agents v0.5/v1)	Responses API (Agents v2)
API versioning	Monthly api-version params	v1 stable routes (/openai/v1/)
Resource model	Hub + Azure OpenAI + Azure AI Services	Foundry resource (single, with projects)
SDKs & endpoints	Multiple packages (azure-ai-inference, azure-ai-generative, azure-ai-ml, AzureOpenAI()) against 5+ endpoints	Unified project client (azure-ai-projects 2.x) + OpenAI() against one project endpoint.
Terminology	Threads, Messages, Runs, Assistants	Conversations, Items, Responses, Agent Versions

Your first API call

Get started now — Quickstart: Build with models and agents | Open Foundry portal | Get an Azure account

Send a prompt and get a response from a model in a few lines of code:

Python

import os
from dotenv import load_dotenv
from azure.identity import DefaultAzureCredential
from azure.ai.projects import AIProjectClient

load_dotenv()

print(f"Using PROJECT_ENDPOINT: {os.environ['PROJECT_ENDPOINT']}")
print(f"Using MODEL_DEPLOYMENT_NAME: {os.environ['MODEL_DEPLOYMENT_NAME']}")

project_client = AIProjectClient(
    endpoint=os.environ["PROJECT_ENDPOINT"],
    credential=DefaultAzureCredential(),
)

openai_client = project_client.get_openai_client()

response = openai_client.responses.create(
    model=os.environ["MODEL_DEPLOYMENT_NAME"],
    input="What is the size of France in square miles?",
)
print(f"Response output: {response.output_text}")

C#

using Azure.AI.Projects;
using Azure.AI.Projects.OpenAI;
using Azure.Identity;
using OpenAI.Responses;

#pragma warning disable OPENAI001

string projectEndpoint = Environment.GetEnvironmentVariable("PROJECT_ENDPOINT")
?? throw new InvalidOperationException("Missing environment variable 'PROJECT_ENDPOINT'");
string modelDeploymentName = Environment.GetEnvironmentVariable("MODEL_DEPLOYMENT_NAME")
?? throw new InvalidOperationException("Missing environment variable 'MODEL_DEPLOYMENT_NAME'");

AIProjectClient projectClient = new(new Uri(projectEndpoint), new AzureCliCredential());

ProjectResponsesClient responseClient = projectClient.OpenAI.GetProjectResponsesClientForModel(modelDeploymentName);
ResponseResult response = await responseClient.CreateResponseAsync("What is the size of France in square miles?");

Console.WriteLine(response.GetOutputText());

TypeScript

import { DefaultAzureCredential } from "@azure/identity";
import { AIProjectClient } from "@azure/ai-projects";
import "dotenv/config";

const projectEndpoint = process.env["PROJECT_ENDPOINT"] || "<project endpoint>";
const deploymentName = process.env["MODEL_DEPLOYMENT_NAME"] || "<model deployment name>";

async function main(): Promise<void> {
    const project = new AIProjectClient(projectEndpoint, new DefaultAzureCredential());
    const openAIClient = project.getOpenAIClient();
    const response = await openAIClient.responses.create({
        model: deploymentName,
        input: "What is the size of France in square miles?",
    });
    console.log(`Response output: ${response.output_text}`);
}

main().catch(console.error);

REST API

Replace YOUR-FOUNDRY-RESOURCE-NAME with your values:

curl -X POST https://YOUR-FOUNDRY-RESOURCE-NAME.services.ai.azure.com/api/projects/YOUR-PROJECT-NAME/openai/responses?api-version=2025-11-15-preview \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $AZURE_AI_AUTH_TOKEN" \
-d '{
        "model": "gpt-4.1-mini",
        "input": "What is the size of France in square miles?"
}'

For the full walkthrough, see the Microsoft Foundry quickstart.

Available models

Foundry gives you access to over 1,900 models from Microsoft, OpenAI, Anthropic, Mistral, xAI, Meta, DeepSeek, Hugging Face, and more. The following table highlights popular model families to help you choose a starting point.

Model family	Best for
GPT-5	Most capable — complex reasoning, multi-step tasks, and multimodal scenarios
GPT-4.1	Best balance of capability and cost for production workloads
GPT-4.1 mini	Fastest — low-latency, high-throughput scenarios
Claude	Advanced reasoning, code generation, and multimodal tasks
Grok	Reasoning, coding, and data extraction
Mistral	Code generation, multilingual, and general-purpose chat
DeepSeek-R1	Open-weight reasoning at scale
Phi-4	Small language model — on-device or resource-constrained environments
Meta Llama	Open models — customization and fine-tuning

For help choosing between models, see the GPT-5 vs GPT-4.1 model choice guide. Browse the full catalog in the Foundry Models overview.

What’s new

Foundry is evolving fast. Here are some of the latest additions:

• Model routing with the Responses API — Use a single API to route to any model automatically or by name.
• A2A agent endpoints (preview) — Expose agents as Agent2Agent endpoints so other agents can discover and call them.
• Instant models (preview) — Call any supported model by name without creating a deployment first.
• Voice agents (preview) — Build voice-enabled agents with hosted agents.
• Routines (preview) — Automate multi-step agent workflows with reusable routines.
• Fabric IQ (preview) — Connect agents to Microsoft Fabric data.
• Work IQ (preview) — Connect agents to Microsoft 365 content.

See What’s new in Microsoft Foundry for the full list.

Choose your path

Foundry supports multiple developer surfaces. Use the following table to find the right starting point for your scenario.

I want to…	Start here
Call a model from code	Quickstart: Your first API call
Build an agent with tools and memory	Agent Service overview
Explore models in the browser	Foundry portal playgrounds
Deploy and manage models at scale	Foundry Models overview
Develop in VS Code	Foundry for VS Code
Set up governance and security	Foundry Control Plane

Who is Foundry for?

Microsoft Foundry serves three primary audiences:

• Application developers building AI-powered products with agents, models, and tools. Start with the quickstart.
• ML engineers and data scientists who fine-tune models, run evaluations, and manage model deployments.
• IT administrators and platform engineers who govern AI resources, enforce policies, and manage access across teams. See security and governance and Foundry Control Plane.

Key capabilities

Build agents

Multi-agent orchestration — Build collaborative agent behavior and complex workflow execution using SDKs for C# and Python.

Tool catalog — Connect over 1,400 tools through public and private catalogs.

Memory — Retain and recall contextual information across interactions without requiring repeated input.

Foundry IQ knowledge integration — Ground agent responses in enterprise or web content with citation-backed answers.

Publishing — Publish agents to Microsoft 365, Teams, BizChat, or containerized deployments.

Operate and govern

Real-time observability — Monitor performance and governance with built-in metrics and model tracking.

Centralized AI asset management — Manage all agents, models, and tools from the Operate section, including agents registered from other clouds.

Enterprise controls — Full authentication support for MCP and A2A, AI gateway integration, and Azure Policy integration.

Microsoft Foundry API and SDKs

The Microsoft Foundry API provides a consistent contract for building agentic applications across different model providers. SDK client libraries are available for:

• Python
• C#
• JavaScript/TypeScript (preview)
• Java (preview)

The Microsoft Foundry for VS Code Extension helps you explore models and develop agents directly in your development environment.

Foundry portal

The Microsoft Foundry portal is where you manage projects, deploy models, build agents, and monitor your AI assets. To use the current version, make sure the New Foundry toggle in the banner is set to on.

See Find features in the Foundry portal if you’re used to the Foundry (classic) portal and not sure where to find things now.

For details on switching between projects or finding resources created in Foundry (classic), see Find features in the Foundry portal.

Pricing and billing

The platform is free to use and explore. Pricing occurs at the deployment level. Each product within Foundry (models, agents, tools) has its own billing model and price.

Using Foundry also incurs costs associated with the underlying services. To learn more, read Plan and manage costs for Foundry Tools.

Use the Total Economic Impact calculator for Foundry to estimate your return on investment.

Region availability

Foundry is available in most regions where Foundry Tools are available. For more information, see region support for Microsoft Foundry.

How to get access

You need an Azure account. Then sign in to Microsoft Foundry.

Related content

• Quickstart: Get started with Microsoft Foundry
• Quickstart: Set up Foundry resources
• Instant models in Microsoft Foundry (preview)
• Create a project
• Get started with an AI template
• Use the Microsoft Foundry Skill in coding agents
• What’s new in Microsoft Foundry documentation?