Best Practice for Azure AI service cost configuration

2025-02-12T08:28:25.8366667+00:00

Hi All, Good Day to you

I have question regarding Azure AI services

How can I temporarily stop or pause Azure AI services (like Azure Machine Learning and Search) to reduce costs, without deleting them, and reactivate them later when needed? Are there any best practices or cost-saving configurations for this scenario? So basically we just dont want unneccessary charges since we are not using this currently and we probably will reactivate this again when needed.

Hope can get some insights and recommendation.

Thanks!

(Below is my resource visualizer)User's image

Azure Machine Learning
Azure Machine Learning
An Azure machine learning service for building and deploying models.
3,120 questions
Azure AI services
Azure AI services
A group of Azure services, SDKs, and APIs designed to make apps more intelligent, engaging, and discoverable.
3,120 questions
{count} votes

1 answer

Sort by: Most helpful
  1. santoshkc 12,265 Reputation points Microsoft Vendor
    2025-02-12T10:22:37.05+00:00

    Hi @Syed muhammad harith al gadri syed abdul rahman,

    Thank you for reaching out to Microsoft Q&A forum.

    To temporarily stop or pause Azure AI services like Azure Machine Learning and Azure AI Search to reduce costs without deleting them, consider the following best practices:

    1. Pause Compute Resources:
      • Azure Machine Learning: Enable idle shutdown for compute instances or set up an auto-stop schedule. Alternatively, delete compute clusters when not in use, as stopping alone may still incur disk and load balancer costs.
      • Azure AI Search: Scale down the search service to a lower tier when not in use instead of keeping a high-performance SKU active.
    2. Monitor and Optimize Costs:
      • Use Azure Cost Management to track spending trends.
      • Set budgets and alerts to avoid unnecessary expenses.
      • Regularly review the Azure Pricing Calculator before deploying new resources.
    3. Delete instances, clusters and/or online deployments if you don't plan on using them soon.
    4. Use Auto-scaling and Low-Priority VMs:
      • Enable auto-scaling for training clusters and managed endpoints.
      • Use low-priority VMs to reduce compute costs for non-critical workloads.
    5. Leverage Azure Prepayment or Reserved Instances:
      • If you anticipate high future usage, consider Azure Reserved VM Instances for discounts.

    For more details, please refer to:

    Plan to manage costs for Azure ML & Manage and optimize Azure ML costs

    Plan and manage costs of an Azure AI Search service

    I hope this information helps. Do let us know if you any further queries.


    If this answers your query, do click Accept Answer and Yes for was this answer helpful.

    0 comments No comments

Your answer

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