Project ID: (PII Removed by Staff)
Service: Vertex AI
SKU: Online/Batch Prediction Nvidia H100 80GB GPU running in Iowa (SKU ID: AE9C-DB60-DF46)
Region: us-central1 (and verified across other regions)
Date of issue start: 2025/sep/03
Actions already taken:
Undeployed the model from all Vertex AI endpoints.
Verified that gcloud ai endpoints list shows no deployed endpoints.
Checked for custom jobs, batch prediction jobs, and notebooks — none are running.
Verified across multiple regions (us-central1, us-east5, us-west1, etc.).
Issue:
Despite taking these steps, billing reports continue to increase every hour for the H100 GPU SKU.
Request:
Please investigate why H100 GPU charges are still being applied after resources have been undeployed, and confirm whether this is a billing delay or an active resource that is not visible in the console.
Thank you for your assistance.