> ## Documentation Index
> Fetch the complete documentation index at: https://symphony-docs.fcamara.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Azure AI Foundry - Add models

> Step-by-step guide to deploy new models in Azure AI Foundry for use in Symphony.

## What it is

**Azure AI Foundry** is Microsoft's centralized environment for managing AI models, agents, and integrations. This page describes how to add new models so they become available to applications and services consuming AI in Foundry. The entire process is done directly in the [Azure portal](https://portal.azure.com).

## Step by step

<Steps>
  <Step title="Access Foundry in Azure">
    On the Azure portal home page, click **Foundry**. If the option doesn't appear in the home screen shortcuts, use the top search bar and search for **Foundry**.
  </Step>

  <Step title="Select the Foundry resource">
    On the **Microsoft Foundry | All resources** screen, locate the Foundry resource used by the project (e.g., `ai-fcopilothub973968710647`). Click the resource to open its details.
  </Step>

  <Step title="Open the Foundry portal">
    Within the selected resource, click **Go to Foundry portal**. This button redirects to the Microsoft Foundry portal, where you can manage models, agents, playgrounds, and deployments.
  </Step>

  <Step title="Check resource information">
    When accessing the Foundry portal, the resource home screen is displayed. There you can view important information for integration:

    * **Endpoint/URL**
    * **Access keys**
    * **Resource name**
    * **Region**
    * **Project data**

    This information can be used to configure applications, pipelines, or services that need to communicate with Azure AI Foundry.
  </Step>

  <Step title="Access the model catalog">
    In the portal's side menu, click **Model catalog** to view the models available to add to the bus.
  </Step>

  <Step title="Select or search for a model">
    Use the search bar to locate the desired model, or browse models evaluating criteria such as **quality**, **safety**, **cost**, and **throughput**. Select the model to start deployment.
  </Step>

  <Step title="Verify deployment">
    After creating the deployment, the model appears in the **Model deployments** list. Check that the **Status** field shows **Succeeded** — then the model will be available for use in Azure AI Foundry.
  </Step>
</Steps>

<Note>
  The deployed model will be used by applications, agents, or integrations that need to consume AI within the Foundry environment. To make it available in Symphony, connect the provider/endpoint in **Admin Panel → Connection** and enable it in **Models** (see [SaaS Manual](/en/admin/manual-saas)).
</Note>