Clarification needed: GPT Standard deployments in Switzerland North

Michael Bock 5 Reputation points
2026-06-30T08:31:20.06+00:00

We are currently working on a production workload on Azure OpenAI in the Switzerland North region using any GPT Version that can be deployed with the Standard deployment type.

About one to two weeks ago, GPT-4.1 was available in the Switzerland North region with the Standard deployment type. However, this is no longer the case. Currently, no GPT large language models (LLMs) are available for Standard deployments in Switzerland North—only embedding and speech recognition models are listed.

Could you clarify when GPT LLMs will be available again in the Switzerland North region with the Standard deployment type?

Azure OpenAI in Foundry Models

3 answers

Sort by: Most helpful
  1. SRILAKSHMI C 19,735 Reputation points Microsoft External Staff Moderator
    2026-07-07T11:50:02.77+00:00

    Hello @Michael Bock

    Thank you for providing the screenshot and the additional details.

    I was able to review what you've shared, and it confirms your observation. When filtering for:

    • Region: Switzerland North
    • Deployment SKU: Standard
    • Lifecycle: Generally Available

    only the following models are currently available:

    • Whisper
    • text-embedding-ada-002
    • text-embedding-3-small
    • text-embedding-3-large

    At this time, no GPT large language models are listed for Standard deployments in Switzerland North. As you mentioned, GPT-5 series models are available under Data Zone Standard, but not under the Standard deployment SKU.

    Based on the current publicly available documentation and model availability information, there is no published ETA for when GPT LLMs will become available again for Standard deployments in Switzerland North.

    Model availability is determined by factors such as regional rollout, deployment type, model lifecycle, and available capacity, and may change over time. While GPT-4.1 was previously available in this region, we don't have any public information indicating when or if it will return for Standard deployments.

    At this point, the available options are:

    • Continue using Data Zone Standard, if it satisfies your workload and compliance requirements.
    • Deploy the required GPT model in another supported region, if your architecture permits.
    • Monitor the Azure OpenAI model availability documentation and Azure AI Foundry portal, as these are updated when additional models become available.

    If having GPT models available specifically under the Standard deployment SKU in Switzerland North is a production requirement, we recommend opening an Azure support request. This allows the product team to verify whether this is an expected regional availability change or whether there are any subscription- or capacity-specific factors affecting your tenant.

    Thank you!

    Was this answer helpful?


  2. Michael Bock 5 Reputation points
    2026-06-30T15:23:05.8666667+00:00

    Those are the only models available when filtered Region Switzerland North, Deployment SKU Standard, Lifecycle Generally available. For DataZone Standard there are available gpt 5 series that would be fine, but for Deyployment SKU Standard there are only the listed model available .User's image

    Was this answer helpful?

    0 comments No comments

  3. Wang Wang 5 Reputation points
    2026-06-30T10:25:49.46+00:00

    Same issue here. According to the official docs, GPT-4o and GPT-4.1 shouldn't be deprecated until October this year — so either Microsoft moved the date up without telling anyone, or this is a bug.

    If the timeline was changed, the documentation needs to be updated immediately so people stop relying on wrong info. If it wasn't changed, then this is clearly a bug and it needs to be fixed ASAP.

    This is already seriously affecting our work. We shouldn't have to find out about critical changes through broken deployments instead of proper communication. Please fix this or update the docs — this isn't acceptable as it stands.

    Was this answer helpful?

    0 comments No comments

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.