Azure OpenAI Realtime: Can GPT Realtime processing remain in Australia East?

Question

Azure OpenAI Realtime: Can GPT Realtime processing remain in Australia East?

Teah Rota 0

We need GPT Realtime for a speech-to-speech healthcare use case with Australian data-sovereignty requirements.

The Azure Speech regional documentation lists gpt-realtime as available for australiaeast, but only with a Global Standard deployment. Microsoft’s deployment documentation states that Global Standard may process prompts and responses in any Azure region, while single-region processing requires a Standard or Regional deployment.

Can GPT Realtime currently be deployed with processing restricted to Australia, either through a regional deployment, a private/approved configuration, or another supported Azure option?

SRILAKSHMI C 19,825 Reputation points Microsoft External Staff Moderator

2026-08-18T04:32:13.83+00:00

Hello @Teah Rota

Thank you for Reaching out to Microsoft Q&A, and I apologize that our previous response did not fully address your specific requirement. I understand that your key concern is strict Australia-only processing for GPT Realtime, particularly for a healthcare workload with Australian data-sovereignty requirements.

GPT Realtime and Australia-only processing

Microsoft's current documentation states that deployment type determines where inference data is processed:

Global Standard - prompts and responses may be processed in any Azure region where the model is deployed.

Data Zone - processing is restricted to the applicable Microsoft-defined data zone. For APAC, this includes Australia, Japan, Korea, Singapore, and India.

Standard / Regional Provisioned - processing occurs in the specific Azure region associated with the deployment.

Therefore, if your requirement is specifically "processing must remain in Australia/Australia East", a Global Standard gpt-realtime deployment should not be treated as providing an Australia-only processing guarantee.

The fact that gpt-realtime appears in an Australia East availability view does not, by itself, mean that every deployment type is available there. Microsoft explicitly notes that not all models support all deployment types, and the model availability page should be checked by both region and deployment type.

Is there a supported way to force Global Standard to Australia East?

Based on Microsoft's documented deployment model, there is no customer-side setting that converts a Global Standard deployment into a single-region Australia East deployment.

If the model is available in Australia East only through Global Standard for your subscription/deployment scenario, Global Standard cannot be configured to guarantee that inference processing will remain in Australia East. Global deployments dynamically route traffic to available Azure infrastructure.

Similarly, a Private Endpoint, VNet, firewall, or other network configuration does not change the geographic processing boundary of the model deployment. Those controls govern network access; the deployment type determines the inference processing location.

What about an APAC Data Zone deployment?

This is a possible alternative only if your requirement is APAC residency rather than Australia-only residency.

Microsoft documents that the APAC Data Zone includes Australia, Japan, Korea, Singapore, and India. Consequently, a Data Zone deployment provides an APAC processing boundary, but it does not guarantee that processing occurs specifically in Australia or Australia East.

Recommended approach for your healthcare workload

For a requirement of "GPT Realtime inference data must be processed only within Australia."

the deployment needs to support a Standard/regional or Regional Provisioned deployment in an Australian region. Microsoft's documentation specifically recommends Standard or Regional Provisioned when a workload requires single-region processing.

If gpt-realtime is not currently offered in Australia East under one of those single-region deployment types, then there is currently no documented Azure configuration that can guarantee Australia-only processing for that model.

I recommend checking the current Foundry model catalog --> GPT Realtime --> Region/Deployment type availability for your subscription, because Microsoft notes that model availability can differ by deployment type and region.

Please refer this

Understanding deployment types in Microsoft Foundry Models

Region availability for Foundry Models sold by Azure

Data, privacy, and security for Foundry Models sold by Azure

I Hope this helps. Please do let me know if you have any further queries.

Thank you!
Manish Deshpande 8,220 Reputation points Microsoft External Staff Moderator

2026-08-18T16:37:56.0466667+00:00

Hello @Teah Rota

We sincerely apologize that your issue has not yet been resolved. Your satisfaction is important to us, and we are committed to helping you reach a resolution. Could you please share additional details here so we can assist further? If preferred, you may also share your email address and a convenient time to connect via Private Message.
SRILAKSHMI C 19,825 Reputation points Microsoft External Staff Moderator

2026-08-19T06:30:36.95+00:00

Hi @Teah Rota

I’m following up on our previous response regarding GPT Realtime and the Australia-only processing requirement for your healthcare workload.

Could you please confirm whether the information and documentation provided addressed your data-sovereignty requirements?

In particular, please let us know if you were able to confirm the available GPT Realtime deployment types in Australia East from the Foundry model catalog.

If you still have questions or require further clarification regarding Global Standard, Data Zone, or single-region processing, please share the specific requirement and we’ll be happy to assist further.

If the information provided has addressed your concern, please let us know so we can proceed accordingly.

Thank you!
SRILAKSHMI C 19,825 Reputation points Microsoft External Staff Moderator

2026-08-20T09:41:11.6966667+00:00

Hi @Teah Rota

I’m following up on our previous response regarding GPT Realtime and the Australia-only processing requirement for your healthcare workload.

Could you please confirm whether the information and documentation provided addressed your data-sovereignty requirements?

In particular, please let us know if you were able to confirm the available GPT Realtime deployment types in Australia East from the Foundry model catalog.

If you still have questions or require further clarification regarding Global Standard, Data Zone, or single-region processing, please share the specific requirement and we’ll be happy to assist further.

If the information provided has addressed your concern, please let us know so we can proceed accordingly.

Thank you!

1 answer

Your answer

Manish Deshpande 8,220 Reputation points Microsoft External Staff Moderator

2026-08-18T16:37:56.0466667+00:00

Hello @Teah Rota

We sincerely apologize that your issue has not yet been resolved. Your satisfaction is important to us, and we are committed to helping you reach a resolution. Could you please share additional details here so we can assist further? If preferred, you may also share your email address and a convenient time to connect via Private Message.
SRILAKSHMI C 19,825 Reputation points Microsoft External Staff Moderator

2026-08-19T06:30:36.95+00:00

Hi @Teah Rota

I’m following up on our previous response regarding GPT Realtime and the Australia-only processing requirement for your healthcare workload.

Could you please confirm whether the information and documentation provided addressed your data-sovereignty requirements?

In particular, please let us know if you were able to confirm the available GPT Realtime deployment types in Australia East from the Foundry model catalog.

If you still have questions or require further clarification regarding Global Standard, Data Zone, or single-region processing, please share the specific requirement and we’ll be happy to assist further.

If the information provided has addressed your concern, please let us know so we can proceed accordingly.

Thank you!
SRILAKSHMI C 19,825 Reputation points Microsoft External Staff Moderator

2026-08-20T09:41:11.6966667+00:00

Hi @Teah Rota

I’m following up on our previous response regarding GPT Realtime and the Australia-only processing requirement for your healthcare workload.

Could you please confirm whether the information and documentation provided addressed your data-sovereignty requirements?

In particular, please let us know if you were able to confirm the available GPT Realtime deployment types in Australia East from the Foundry model catalog.

If you still have questions or require further clarification regarding Global Standard, Data Zone, or single-region processing, please share the specific requirement and we’ll be happy to assist further.

If the information provided has addressed your concern, please let us know so we can proceed accordingly.

Thank you!

Answer 1

Hi @Teah Rota

Thanks for reaching out to Microsoft Q&A I understand you are looking to validate data sovereignty for a health-care workload.

I would not use the Australia East Global standard option for a workload that requires inference processing to remain in Australia.

The Azure Speech supported-regions page lists Voice Live gpt-realtime in australiaeast, but it lists the deployment type as Global standard. The Foundry deployment-types documentation claims that Global deployments can process inference data in any Azure region. Therefore, deployment in Australia East does not by itself provide an Australia-only processing commitment.

For a single-region processing requirement, the same documentation identifies Standard or Regional Provisioned as the relevant deployment types, where supported. I could not find a documented gpt-realtime Standard or Regional Provisioned option in Australia East. A private endpoint or other network configuration would not alter the documented processing location for a Global deployment. Data Zone processing is also not an Australia-only alternative: it is limited to a Microsoft-defined APAC zone.

There is a documentation inconsistency worth calling out: the Speech region table shows gpt-realtime Global standard in Australia East, while the Foundry regional availability table does not show gpt-realtime in Australia East. I would open an Azure support request before proceeding, including the exact service path (Speech Voice Live or Azure OpenAI Realtime), model version, resource region, subscription, and the requirement for Australia-only processing. Ask support to confirm whether an Australia-only deployment type is available for that exact configuration and to investigate the conflicting availability tables.

For a healthcare use case with a strict Australia-only requirement, I would treat this as a current product-availability dependency and wait for written confirmation of a supported single-region option rather than relying on Global standard.

Please 'Upvote' (Thumbs-up) and 'Accept' as answer if the response was helpful. This will be helpful to other community members who face the same issue.

Best regards, Andrew S Taylor

Teah Rota 0 Reputation points

2026-08-18T02:58:09.5766667+00:00

Thanks for the corroboration, Andrew - this confirms my reading of the current documentation. The remaining question is whether Microsoft offers any supported Australia-only deployment option for GPT Realtime that isn't reflected in the public docs, such as a request-based/private deployment or other approved configuration.