1. In Azure Foundry, under “Use with Foundry” → Foundry, select the aiportal foundry project

   1. From there, click “Go to Foundry Portal”

   

2. Once in the Foundry, select “Models and Endpoints” in the left side menu

1. Click Deploy Model → Deploy Base Model

1. Select the model from the list of providers, click confirm and then click deploy

1. if you are not able to deploy a model make sure you have:
   1. Requested GPT-5 access
   2. Have enough quota; If you need quota, you can request more model quota
2. Once deployed, you should be taken to the model’s screen (if not, select the model from the list of model deployments)

4a. Take note of:

Endpoint The “azure endpoint” in the right code screen (in the above it is: https://demo-aiportal.cognitiveservices.azure.com/)

Deployment The “Name” in the left table (in the above it is: gpt-5.4)

Model Deployment Quota The “Rate Limit” in the lower left table (in the above it is: 150,000)

1. Once the model has been added to Azure and the values have been noted, add the provider to your AI Portal using Adding / Updating a ChatGPT Provider in AI Portal