Azure OpenAI

Azure OpenAI lets RealTimeX use an OpenAI deployment hosted in your Azure environment instead of a direct OpenAI API account.

Open it from Settings > AI Providers > LLM, then choose Azure OpenAI.

What you need

Before you configure it in RealTimeX, you need:

• an Azure OpenAI resource
• the Azure OpenAI endpoint URL
• an Azure OpenAI API key
• the deployment name you want RealTimeX to call
• the context-window size that matches that deployment

Current setup flow

1. Open Settings > AI Providers > LLM.
2. Choose Azure OpenAI.
3. Paste the Azure endpoint URL.
4. Paste the Azure API key.
5. Enter the chat deployment name.
6. Choose the token limit that matches that deployment.
7. Choose the model type.
8. Save the settings.
9. Test a normal chat.

Fields the current UI exposes

Service endpoint

This is the Azure OpenAI endpoint URL for your resource.

API key

This is the Azure OpenAI key tied to that resource.

Chat deployment name

This is the deployment identifier RealTimeX should call.

Use the Azure deployment name, not just the underlying model family name.

Chat model token limit

The current UI asks you to choose the context-window size explicitly.

Choose the value that matches the deployment you created in Azure. If this is wrong, your effective context assumptions inside RealTimeX will also be wrong.

Model type

The current selector supports:

• Default
• Reasoning

Use Reasoning only when the Azure deployment is a reasoning-style model that should be treated differently by the current provider integration.

When to choose Azure OpenAI instead of native OpenAI

• Choose Azure OpenAI when your organization standardizes on Azure-hosted OpenAI resources.
• Choose it when compliance, networking, or enterprise procurement requirements are tied to Azure.
• Use native OpenAI when you want the simpler API-key-plus-model-list workflow.

Troubleshooting

The provider saves but responses fail

Check the four most common mismatches:

• wrong endpoint
• wrong API key
• wrong deployment name
• wrong token-limit assumption

The deployment exists in Azure but RealTimeX still cannot use it

Make sure the deployment is a chat-capable Azure OpenAI deployment and that the key belongs to the same resource as the endpoint you entered.

You are not sure whether to pick Default or Reasoning

Start with Default unless you intentionally created a reasoning deployment and know that is the behavior you want RealTimeX to use.