Will I get a refund after cancelling?

Azure ML is pay-as-you-go, so you are only billed for actual usage up to the deletion date — there are no cancellation fees. If you have prepaid Reserved Instances, early termination may involve fees; contact Azure Support. Azure Support plan subscriptions do not receive prorated refunds — you pay through the end of the billing month.

Can I recover my workspace after deletion?

Yes, if you used the default soft delete option: go to Azure portal → Machine Learning → click 'Recently deleted' at the top of the page → select your workspace → click Recover. You have 14 days to do this. If you chose 'Delete the workspace permanently,' recovery is not possible. Note: compute resources, role assignments, and linked Databricks workspaces are not guaranteed to be restored and may need to be manually recreated.

How long until my data is permanently deleted?

With soft delete (default): 14 days after workspace deletion. With permanent delete (optional checkbox at deletion): immediately and irreversibly. If you cancel your entire Azure subscription, Microsoft retains subscription data for 30–90 days before permanent deletion, after which the subscription cannot be reactivated.

Does deleting the workspace cancel my Azure subscription?

No. Deleting an Azure ML workspace only removes that specific workspace — your Azure subscription and all other services remain active. To cancel your entire Azure subscription, go to Cost Management + Billing → Subscriptions in the Azure portal, select your subscription, and click Cancel. Only a Subscription Owner can do this.

Can I export my models and training data before deleting?

Yes. In Azure ML Studio, go to Jobs → All Jobs, select a job, and click 'Download all' to export artifacts and logs. For models, go to Models, select the model, and click 'Download all.' Your data in the associated Azure Storage account is not deleted with the workspace by default, so blobs remain accessible until you delete the storage account manually.

Azure Machine Learning 해지 방법 (2026)

Azure Machine Learning 플랜은 US$0.00/월부터 시작합니다. 7단계 가이드를 따라 약 5분 만에 구독을 해지하세요 — 유지 제안을 피하는 팁 포함.

업데이트: 2026년 4월

보통

웹사이트 방문

예상 소요 시간:5 분

해지 방법:Website

Sign in to Azure Portal

Go to https://portal.azure.com and sign in with your Azure credentials. You must have the Owner role on the subscription or workspace to delete it.

Navigate to Machine Learning

In the top search bar, type 'Machine Learning' and select it from the dropdown results to see your list of Azure ML workspaces.

Select your workspace

From the list of workspaces, click the name of the workspace you want to delete.

Click Delete

At the top of the workspace overview page, click the 'Delete' button. A confirmation panel will slide in from the right side.

Choose permanent or soft delete

By default, Azure performs a soft delete (14-day recoverable window). To delete immediately and permanently, check the option 'Delete the workspace permanently' in the confirmation panel. Note: compute resources (clusters, instances, endpoints) are hard-deleted immediately regardless of this choice — billing for compute stops right away.

Confirm deletion

Type the workspace name in the confirmation field if prompted, then click 'Delete' to confirm. The workspace will be removed from your active list within a few minutes.

Delete associated resources (optional)

The workspace deletion does NOT automatically remove the linked Storage account, Key Vault, Container Registry, or Application Insights. To delete these, return to the portal, search for each service by name, and delete them individually — or use the Python SDK with delete_dependent_resources=True.

더 저렴한 대안

Microsoft 365

부터 AU$16.00/월

Google Workspace

Wasabi

부터 US$6.99/월(≈ AU$9.81)

さくらのクラウド

Salesforce Essentials

부터 US$25.00/월(≈ AU$35.10)

자주 묻는 질문

구독 관리에 도움이 필요하신가요?

Subger로 모든 구독을 추적하고, 해지 리마인더를 받고, 최고의 혜택을 찾아보세요.

Subger 무료 체험

제휴 공시: 이 페이지의 일부 링크는 제휴 링크일 수 있습니다. 해당 링크를 통해 구매하시면 Subger가 귀하에게 추가 비용 없이 소정의 수수료를 받을 수 있습니다. 이를 통해 사이트를 무료로 유지하고 정확한 구독 정보를 계속 제공할 수 있습니다. 당사는 가치가 있다고 판단하는 서비스만 추천합니다.