Need help understanding the options and road map for countries that are loosing regional standard Open AI model options with gpt4.1 mini depreciation

Hi @Camden Knight

Thank you for reaching out to Microsoft Q&A.

I understand your concern regarding the deprecation of GPT-4.1-mini and GPT-4o, especially since your solution relies on Regional Standard deployments in Canada East or Canada Central to meet your data residency requirements.

At this time, Microsoft has not published a roadmap or timeline indicating when replacement models (such as GPT-5.x series) will become available as Regional Standard deployments in specific Azure regions. Model availability is determined by several factors, including regional capacity, infrastructure readiness, and rollout plans, and these details are announced as they become available.

In the meantime, here are a few recommendations that may help you evaluate your available options.

1. Verify Current Model Availability

The most up-to-date source for model availability is the Azure AI Foundry portal, as it reflects the current deployment inventory and may be updated before the documentation.

Navigate to:

Azure AI Foundry → Models + Endpoints → Deploy Model

Then filter by:

• Region: Canada East or Canada Central

Deployment Type: Regional Standard

This will show the models that are currently available for deployment in your selected region.

2. Understand Model Successors

When a model is deprecated, its replacement may not become available simultaneously across all Azure regions or deployment types.

If GPT-4.1-mini or GPT-4o is no longer available for new Regional Standard deployments in your region, the next step is to identify which successor models are currently supported in Canada East or Canada Central under the Regional Standard deployment type.

3. Verify Quota and Access

If you expect a model to be available but do not see it in Azure AI Foundry, it's also worth verifying whether availability is being affected by:

Regional quota or capacity constraints.

Limited-access or gated model requirements.

Subscription-specific availability.

You can review your available quota in Azure AI Foundry Management Center under Quota.

4. Evaluate Alternative Deployment Options

If Regional Standard is not currently available for the desired replacement model, you may also consider whether one of the following deployment types meets your business and compliance requirements:

Global Standard

Provisioned Throughput (PTU)

Another Azure region that supports Regional Standard for the required model

However, if Canadian data residency is a strict requirement, Regional Standard deployment in a supported Canadian region remains the recommended approach.

Regarding Your Question on the Roadmap

Based on the currently published Microsoft documentation, there is no announced timeline for when GPT-5.x Regional Standard deployments will become available in Canada East or Canada Central.

The recommended approach is to:

Continue monitoring the Azure AI Foundry model availability documentation for updates.

Periodically check the Azure AI Foundry portal, as it reflects the latest supported models and deployment types.

Follow the published model lifecycle and retirement guidance as replacement models become available.

You can review the latest model availability here:

https://learn.microsoft.com/azure/ai-foundry/how-to/deploy-models#region-availability

Azure OpenAI models and regions for Foundry Agent Service (classic) — Standard availability tables (example region/model availability reference): https://learn.microsoft.com/azure/foundry-classic/agents/concepts/model-region-support?wt.mc_id=knowledgesearch_inproduct_azure-cxp-community-insider#available-models

Verify model support steps (how to check region/model/capabilities in Foundry classic): https://learn.microsoft.com/azure/foundry-classic/agents/concepts/model-region-support?wt.mc_id=knowledgesearch_inproduct_azure-cxp-community-insider#verify-model-support

I Hope this helps. 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.

Thank you!

Hello Camden Knight,

Greetings! Thanks for raising this question in the Q&A forum

I looked into this closely and confirmed that what you're seeing is a real and currently unresolved gap, not something you are missing on your end. This is a known pain point that other customers have run into as well, and Microsoft has not published a committed timeline for closing it, so I want to give you an accurate picture rather than a false reassurance.

Here is the situation as it stands today. Regional Standard deployment in Canada Central and Canada East has historically only supported a narrow set of models, mainly the gpt-4o family and older GPT-4 generations. Newer model families are rolled out to Standard deployment region by region on a rolling schedule that Microsoft does not publish in advance, and Canada has consistently lagged. As one example from the official docs, when gpt-5.1 Standard was added following the gpt-4o upgrade, it landed in only eight regions, centralus, eastus, eastus2, northcentralus, southcentralus, swedencentral, westus, and westus3, and Canada Central or Canada East were not among them.

On the retirement side, here is the current published schedule that affects you.

gpt-4o (2024-05-13)   Deprecated   Retires 2026-10-01   Replacement: gpt-5.1
gpt-4o (2024-08-06)   Deprecated   Retires 2026-10-01   Replacement: gpt-5.1
gpt-4o (2024-11-20)   GA           Retires 2026-10-01   Replacement: gpt-5.1
gpt-4.1               GA           Retires 2026-10-14   Replacement: not yet listed
gpt-4.1-mini          GA           Retires 2026-10-14   Replacement: not yet listed
gpt-4.1-nano          GA           Retires 2026-10-14   Replacement: not yet listed

The listed replacement for gpt-4o, gpt-5.1, is not yet available as a Standard deployment in either Canada region, and gpt-4.1 mini does not have a published replacement model at all yet. That combination is exactly why you cannot find a like-for-like regional standard option today.

Step 1: Verify the current state directly rather than relying on a documentation snapshot

Region availability tables in the docs are periodically refreshed and can lag the live catalog. Check the authoritative live list for your subscription.

Foundry portal path: Operate > Quota > toggle "Show all" to see every model and region, including ones not yet deployed

You can also query it programmatically per region:

GET https://management.azure.com/subscriptions/{sub}/providers/Microsoft.CognitiveServices/locations/canadacentral/models?api-version=2024-10-01
GET https://management.azure.com/subscriptions/{sub}/providers/Microsoft.CognitiveServices/locations/canadaeast/models?api-version=2024-10-01

Step 2: Rule out Data Zone as a substitute, since it does not cover Canada

Azure OpenAI Data Zone deployments currently only support the United States and European Union zones. There is no Canada Data Zone, so this deployment type cannot be used as a workaround for your Canadian residency requirement, you would need to stay on true regional Standard or Provisioned deployment in Canada Central or Canada East.

Step 3: Check whether Provisioned throughput has different regional coverage than Standard

Provisioned deployment types sometimes follow a different rollout schedule than pay-as-you-go Standard. If your workload can tolerate the Provisioned Throughput Units commitment model, check the Foundry Models sold by Azure page for the current model by deployment type by region matrix, since gpt-5.1 or a newer family may already be available as Provisioned in Canada even though Standard is not.

https://learn.microsoft.com/en-us/azure/foundry/foundry-models/concepts/models-sold-directly-by-azure

Step 4: File a formal regional availability request

Since Microsoft does not publish forward looking rollout dates for adding models to specific Standard regions, the only way to influence or get visibility into the roadmap for Canada Central and Canada East is through the official feedback channel, which product teams do monitor for prioritizing regional expansion.

https://feedback.azure.com/d365community/forum/79b1327d-d925-ec11-b6e6-000d3a4f06a4

Step 5: Open a support case citing your compliance deadline if this is business critical

Given that gpt-4o retires on 2026-10-01 and you have no in-region Standard replacement today, I would recommend opening an Azure support request now rather than waiting, referencing the retirement date and your data residency requirement explicitly. If you have a Microsoft account team or Unified Support contract, this is also worth raising with them directly, since they sometimes have visibility into unpublished regional rollout plans that are not available through public documentation or standard support channels.

If this answer helps you kindly accept the answer which will help others who have similar questions.

Best Regards,

Jerald Felix.