Azure OpenAI GPT-5.4 Availability

Question

Azure OpenAI GPT-5.4 Availability

Cory Tomlinson 55

Hi,

Can someone from Microsoft provide some answers on when Azure Foundry will have full support for OpenAI gpt-5.4 models?

https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/introducing-gpt-5-4-in-microsoft-foundry/4499785

And perhaps why comments on the above post have been turned off?

My observations as a long time Azure OpenAI customer is that something is going very poorly with the entire gpt-5.x class model deployment. Given that gpt-5.1 was released in Nov '25, followed by gpt-5.2 in Dec '25, and now gpt-5.4 on March 5, and that Azure still doesn't have meaningful support for any of these models tells me there are larger underlying problems.

To announce these new models as "generally available in Microsoft Foundry" is dishonest at best because even 2 weeks later you can only get a very limited set of deployment options, in a very small number of data center regions.

What is going on? When will gpt-5.4 be broadly available across all regions and for all deployment types as it should be, and with support for all of the new features and capabilities in the API?

0 comments

1 answer

Your answer

Answer 1

Anshika Varshney 12,115 Microsoft External Staff Moderator

Hi Cory Tomlinson

I understand the confusion around GPT‑5.4 availability, so let me clarify how it currently works in Azure OpenAI and Azure AI Foundry.

GPT‑5.4 has been announced as generally available in Microsoft Foundry, but that does not mean it is instantly available in all regions, deployment types, or subscription setups. New OpenAI models are usually rolled out in phases. They first appear in a limited set of regions and with limited deployment options and then expand over time as capacity and backend validation complete. This staged rollout is normal for large models and helps ensure stability and performance.

Right now, GPT‑5.4 availability depends on three main things. The region you are using, the deployment type you choose, and whether capacity is available for that model in your subscription. Even if the model is visible in documentation or blogs, it may not yet show up as deployable in every region or support every feature.

To check availability correctly, please look at the official model and region matrix. This page is updated as new regions and deployment options are added https://learn.microsoft.com/azure/ai-services/openai/concepts/models

If GPT‑5.4 does not appear in your region, try switching to another supported region and refresh the deployment list. Many users are able to see newer models earlier in core regions before they expand globally.

Also note that feature support can lag behind the initial release. Some advanced capabilities or deployment modes may be enabled gradually after the base model becomes available. This is why you might see GPT‑5.4 mentioned as available, but not yet usable for every scenario.

For general context on how Azure OpenAI models are released and managed through Foundry, this overview explains the rollout and deployment behavior https://azure.microsoft.com/products/ai-foundry/models/openai

So in short, GPT‑5.4 is being rolled out, but availability is still limited by region and capacity. If you do not see it yet, it usually means your region has not been enabled or capacity is still being expanded. Checking the model availability page and trying another supported region is the best way to confirm what is currently accessible.

Hope this helps clear things up. Do let me know if you have any further queries.

If this answers your query, please do click Accept Answer and Yes for was this answer helpful.

Thankyou!

Cory Tomlinson 55 Reputation points

2026-03-17T21:10:37.1733333+00:00

That doesn't answer my question. I didn't ask for clarification about "how it works" because I am already fully aware of how it works. This model has been out for over 2 weeks and it's only available in East US 2 as PAYG Global or Data Zone. There is not a single NA region that supports Regional or Provisioned deployments. It was the exact same with gpt-5.2 which still has extremely limited availability. If it was being "rolled out" as you say, I would expect it to be slowly showing up in more regions with more deployment types, but that is not what is happening. Nothing is happening.

My point is that Microsoft should not be lagging behind by weeks and months on new model releases. OpenAI can serve the models on day 1 of release as can downstream providers like Cursor. AI is far too competitive and the alternative which is to cancel Azure PTUs and buy API service directly from OpenAI is looking like the only viable path forward.
SRILAKSHMI C 18,990 Reputation points Microsoft External Staff Moderator

2026-03-18T09:42:31.95+00:00
Hi Cory Tomlinson,

I completely understand your concern here especially given your experience and the expectation for faster parity with direct OpenAI access. Let me address your question more directly.

At this time, Azure OpenAI Service / Foundry does not have a published ETA for full regional and deployment-wide availability of GPT-5.4. While it has been announced as generally available in Microsoft Azure AI Foundry, this “GA” status applies to the model being production-ready, not universally deployed across all regions or deployment types yet.

The current limitation you’re seeing (for example, availability primarily in East US 2 and lack of Regional/Provisioned support) is due to:

Capacity constraints for high-demand models

Phased infrastructure rollout across regions

Additional validation required for PTU/Provisioned deployments

These rollouts can take time, and expansion does not always happen linearly across regions, which is why you may not yet see gradual visibility changes.

Please Continue using Global Standard / Data Zone deployments where available

Monitor the official model availability matrix : https://learn.microsoft.com/azure/ai-services/openai/concepts/models

Your feedback about the gap between announcement and broad availability is valid. Azure and OpenAI direct APIs operate on different infrastructure models, and Azure introduces additional layers (regional compliance, enterprise SLAs, provisioned capacity), which can lead to delays in full rollout.

Thank you!
Anshika Varshney 12,115 Reputation points Microsoft External Staff Moderator

2026-03-19T17:04:12.0333333+00:00

Hi Cory Tomlinson,

Did you get any chance to review the response.

Thankyou!
Anshika Varshney 12,115 Reputation points Microsoft External Staff Moderator

2026-03-20T18:44:59.39+00:00

Hi Cory Tomlinson,

Thank you for sharing your feedback. I truly apologies if the initial support did not fully meet your expectations.

Shrilakshmi updated the comment with additional details for clarity. Please review it and let me know if there are any remaining questions or if further guidance is needed. I’ll be happy to assist.

Thankyou!

Share via

Azure OpenAI GPT-5.4 Availability

1 answer

Your answer