I have been using vertex ai services to work on my Gen AI , but the billing is too high and has hit $100 in just 2 days. I need to stop services VertexAi services but i dont see any option. This simply does not say anything else, from here how do i get the breakdown and how do i stop any compute clusters, the google support team sucks badly.
Last month i had a similar issue and they took 1 month and did not resolve the issue.
Hi @Shivi82,
Welcome to Google Cloud Community!
It is entirely understandable to experience concern about unexpected charges, particularly when they are significant and their origin is unclear. Here’s a breakdown of how you can get more clarity and potentially stop the charges:
To get a detailed breakdown of your charges and costs:
Go to the Google Cloud Console and navigate to Vertex AI Dashboard. Search “Detailed Billing Dashboard". See below image for reference:
Here’s another workaround that you may try to get an insight on how Vertex AI usage is being charge:
Go to Google Cloud Pricing Calculator. Click the “Add to estimate”. Then enter “Vertex AI” in the search box. It will populate all Vertex AI related services. Select the service that you will utilize. Then from there you’ll be able to gain insight as to which aspects may be billed in your account if you use a specific service in Vertex AI. See below image for reference:
You may also try referring to Vertex AI pricing documentation for more information.
Then, if you need help with your Cloud Billing account, Google Cloud offers free Cloud Billing Support to all customers. If you’d like to talk to a support representative about a Cloud Billing question or a billing-related issue, visit our Cloud Billing Support page for contact options.
You may consider these when contacting Cloud Billing Support:
Now, to stop the unnecessary resources:
First, for compute clusters**,** even if you don’t see an option to stop all Vertex AI services directly, you should be able to stop or delete individual compute clusters (which are often the biggest source of costs). Look for a “Stop” or “Delete” button within the Vertex AI console.
Next, for model training, if you have it, consider whether they are necessary at this moment. You can pause or stop model training to save costs.
Lastly for deployments, if you have models deployed, and they are not currently in use, you can remove them to stop them from consuming resources.
Here are a few helpful tips to keep your Vertex AI spending in check:
Monitor Usage: Set up alerts or notifications for your Google Cloud account so you get timely warnings if your costs are exceeding a certain limit.
Use Cost Management Tools: Google Cloud has tools that can help you analyze your spending and identify potential areas for cost savings.
Consider Preemptible Instances: For less critical tasks, consider using preemptible instances. These are much cheaper but can be terminated with little notice.
I hope the above information is helpful.
Hi @ruthseki thank you for such a detailed and comprehensive reply, this is very helpful. I really appreciate your time to pin through all possible options.
However some of these i had already tried, my cost was coming from Vertex AI API and there was no way and i feel is no way to do a detailed breakdown, I eventually had to delete the project atleast for now.
After researching and asking GEMINI i had to disable the API, I guess Google should make it easy to analyze, stop api once the budget has been reached as i did set the budget and notification based on which i got to know the budget has been exceeded.
Appreciate again all your suggestions thank you.
