익명 03:44

What are the practical advantages of provisioning an Azure OpenAI resource inste...

What are the practical advantages of provisioning an Azure OpenAI resource instead of an Azure AI Foundry resource?

Azure AI Foundry resources seem like a superset of Azure OpenAI resources as an Azure OpenAI resource may only contain OpenAI models while Azure AI Foundry may contain OpenAI models or some non-OpenAI models such as Grok, Kimi or Qwen. I feel I may have missed something. What are the practical advantages of provisioning an Azure OpenAI resource instead of an Azure AI Foundry resource?



Top Answer/Comment:

There is no advantage. Initially Azure offered only OpenAI so they had separate kind for it. When more models were available they have added another kind that is able to deploy any model. Keep in mind that underneath the resources are both cognitive services account just different kind. Also many of the other kinds of cognitive services are also now part of Azure AI foundry resource. Overall the platform had a lot of renames overtime and a lot of transformations thus the different kinds and why they still exist. There is also small information about this in "What's the difference between Azure OpenAI and Microsoft Foundry Models?". In general if you use only OpenAI and nothing more what Foundry offers like different models, routing, etc. you could just use OpenAI. A little more information on differences Upgrade from Azure OpenAI to Microsoft Foundry.

상단 광고의 [X] 버튼을 누르면 내용이 보입니다