Question about auto-update of gpt-4o model version in Azure OpenAI

Question

I received an email from Microsoft that the model of gpt-4o in Azure OpenAI will be auto-update .

Prepare for a gpt-4o model version update in Azure OpenAI Service You’re receiving this notice because you use Azure OpenAI Service. On 11 December 2024, for Azure OpenAI deployments that are set to Auto-update by default and that use gpt-4o version 2024-05-13, we’ll begin updating them to use version 2024-08-06. The update will be completed within two weeks of that date. Because version 2024-08-06 offers improved capabilities, you may notice some changes in the model behavior and compatibility after the upgrade.

Currently the configuration is as follows.

• Model name: gpt-4o
• Model Version: 2024-05-13
• Deployment Type: Global Standart
• Region: Japan East

There are the following problems with automatic updates

• Currently, gpt-4o 2024-08-06 cannot be selected for Japan East.
• We have not purchased a quota for PTU quota.

Will it still auto-update?

Answer 1

@dott-nagayama At this moment Japan East does not support version 2024-08-06 as listed in this table. What the above auto update means is that your current version gets updated to 2024-08-06 under the same Global Standard offering. The current offerings for PTU also does not have support for Japan East for this model version. So, this is an automatic upgrade that lets you use the latest features, you need not purchase additional PTU quota as the endpoint will still be global standard.

If you have additional questions specific to the deployment, you can always raise a support case from Azure Portal under Health + Support page. I hope this helps!!

If this answers your query, do click Accept Answer and Yes for was this answer helpful. And, if you have any further query do let us know.