Microsoft announced that GPT-Live-1 would be available through Microsoft Foundry, but I still can't find it in the model catalog. I've been checking regularly and haven't seen it appear, even though our organization has an enterprise subscription. Has anyone actually been able to access or deploy the model in Foundry? I'm hoping to use it for a few proofs of concept. If it isn't available there yet, are there any practical alternatives for testing it, especially for workloads that must stay on EU-based servers?
3 Answers
I can’t see it in our enterprise environment either. You may be able to experiment with the model through the OpenAI API, but that won’t necessarily meet regional hosting requirements.
It doesn’t appear to be published in the Foundry model catalog yet. The announcement seems to have come before the model was broadly released, so there may not be anything wrong with your subscription or tenant.
It sounds like the model is still in the process of being rolled out. Several people are seeing the same thing, and access may appear gradually rather than being enabled for every Foundry account at once.

That’s the issue for us—we need EU-only processing, and the OpenAI API seems to require a more expensive business arrangement for that kind of regional setup.