> ## Documentation Index
> Fetch the complete documentation index at: https://agno-v2-update-deprecated-models.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Azure OpenAI

> Use OpenAI models through Azure with Agno agents.

Use OpenAI models through Azure's infrastructure. Learn more [here](https://learn.microsoft.com/azure/ai-services/openai/overview).

Azure OpenAI provides access to OpenAI's models like `GPT-4o`, `gpt-5-mini`, and more.

## Authentication

Navigate to Azure OpenAI on the [Azure Portal](https://portal.azure.com/) and create a service. Then, using the Azure AI Studio portal, create a deployment and set your environment variables:

<CodeGroup>
  ```bash Mac theme={null}
  export AZURE_OPENAI_API_KEY=***
  export AZURE_OPENAI_ENDPOINT=***  # Of the form https://<your-resource-name>.openai.azure.com/openai/deployments/<your-deployment-name>
  # Optional:
  # export AZURE_OPENAI_DEPLOYMENT=***
  ```

  ```bash Windows theme={null}
  setx AZURE_OPENAI_API_KEY ***  # Of the form https://<your-resource-name>.openai.azure.com/openai/deployments/<your-deployment-name>
  setx AZURE_OPENAI_ENDPOINT ***
  # Optional:
  # setx AZURE_OPENAI_DEPLOYMENT ***
  ```
</CodeGroup>

## Example

Use `AzureOpenAI` with your `Agent`:

<CodeGroup>
  ```python agent.py theme={null}
  from agno.agent import Agent
  from agno.models.azure import AzureOpenAI
  from os import getenv

  agent = Agent(
      model=AzureOpenAI(id="gpt-5-mini"),
      markdown=True
  )

  # Print the response on the terminal
  agent.print_response("Share a 2 sentence horror story.")
  ```
</CodeGroup>

## Prompt caching

Prompt caching will happen automatically using our `AzureOpenAI` model. You can read more about how OpenAI handle caching in [their docs](https://platform.openai.com/docs/guides/prompt-caching).

## Advanced Examples

View more examples [here](/models/providers/cloud/azure-openai/usage/basic-stream).

## Parameters

| Parameter                 | Type                       | Default         | Description                                                                |
| ------------------------- | -------------------------- | --------------- | -------------------------------------------------------------------------- |
| `id`                      | `str`                      | (required)      | The deployment name or model name to use                                   |
| `name`                    | `str`                      | `"AzureOpenAI"` | The name of the model                                                      |
| `provider`                | `str`                      | `"Azure"`       | The provider of the model                                                  |
| `api_key`                 | `Optional[str]`            | `None`          | The API key for Azure OpenAI (defaults to AZURE\_OPENAI\_API\_KEY env var) |
| `api_version`             | `Optional[str]`            | `"2024-10-21"`  | The API version to use                                                     |
| `azure_endpoint`          | `Optional[str]`            | `None`          | The Azure endpoint URL (defaults to AZURE\_OPENAI\_ENDPOINT env var)       |
| `azure_deployment`        | `Optional[str]`            | `None`          | The deployment name (defaults to AZURE\_OPENAI\_DEPLOYMENT env var)        |
| `base_url`                | `Optional[str]`            | `None`          | Alternative base URL for the service                                       |
| `azure_ad_token`          | `Optional[str]`            | `None`          | Azure AD token for authentication                                          |
| `azure_ad_token_provider` | `Optional[Any]`            | `None`          | Azure AD token provider for authentication                                 |
| `default_headers`         | `Optional[Dict[str, str]]` | `None`          | Default headers to include in all requests                                 |
| `default_query`           | `Optional[Dict[str, Any]]` | `None`          | Default query parameters to include in all requests                        |

`AzureOpenAI` also supports the parameters of [OpenAI](/reference/models/openai).