Azure OpenAI: Sudden Cost Increase — Need Explanation and Billing Adjustment Consideration

Question

Azure OpenAI: Sudden Cost Increase — Need Explanation and Billing Adjustment Consideration

David Hughes 0

Problem description

I am experiencing a significant increase in costs related to my Azure OpenAI workload, specifically with the Microsoft Foundry service and a supporting Azure-hosted application. The charges have risen from an expected baseline of approximately $600 per month to over $30,000 in the current month. I am seeking to understand what triggered this spike in charges, as no specific resource or usage details are provided in my case record.

Environment

The environment involves the Azure OpenAI service, with a workload comprising a Foundry service and an Azure-hosted application component, in the context of a billing issue. Specific region, model, deployment details, or traffic patterns are not provided.

What I've already tried

Based on the available case information, I have not performed any troubleshooting steps myself. The case record indicates that self-help guidance focused on reviewing cost analysis, but no resource-level breakdown or explanation for the charge increase is documented.

Current status

I am currently seeking an explanation for the cause of the cost spike and would like to explore the possibility of a one-time billing adjustment or service credit. I am requesting assistance to identify the specific resource or meter driving the increase and any relevant usage or deployment changes during the affected period.

Anshika Varshney 15,625 Reputation points Microsoft External Staff Moderator

2026-10-23T16:54:25.0733333+00:00

Hello @David Hughes

Thank you for bringing this to our attention. We are currently checking the details and working on resolving the issue. We’ll update you as soon as we have more information.

1 answer

Your answer

Anshika Varshney 15,625 Reputation points Microsoft External Staff Moderator

2026-10-23T16:54:25.0733333+00:00

Hello @David Hughes

Thank you for bringing this to our attention. We are currently checking the details and working on resolving the issue. We’ll update you as soon as we have more information.

Answer 1

Likely cause: the spike is usually driven by one or more specific meters/resources (for example token-heavy Azure OpenAI usage, traffic pattern changes, or deployment/config changes), but your current case lacks the breakdown needed to prove which one. Start by isolating the exact cost driver, then request billing review with evidence.

Confirm scope and access first.
- In Cost Management, open the same billing/subscription scope where charges were incurred.
- Ensure you have read access to cost data at that scope.
Identify the exact charge source in Cost Analysis.
- Use Invoice details view and the affected date range.
- Group/filter by:
  - Service name
  - Meter
  - Resource
  - Daily granularity
- Why: this pinpoints the meter/resource causing the jump instead of relying on total spend.
Isolate Azure OpenAI contribution.
- Filter to Azure OpenAI/Foundry-related services.
- Compare the current period vs previous baseline period with the same grouping.
- Record which resource and meter changed most (absolute and % increase).
Correlate spend jump with workload changes.
- Check deployment/config changes and traffic changes in the same timeframe.
- For AI workloads, confirm whether prompt/output token volume increased, concurrency increased, or retry loops appeared.
Put guardrails in place immediately.
- Create budget thresholds (actual + forecast alerts).
- Add alert recipients and monitor daily until stabilized.
- Note: budgets alert only; they do not automatically stop resources.
Billing adjustment path (if you suspect abnormal/platform-side behavior).
- Open/continue support with: impacted subscription, exact UTC window, top cost-driving meter/resource, before-vs-after comparison screenshots, and any incident/correlation evidence.
- Ask for formal billing review and goodwill/credit consideration based on findings.

Validation checks:

You can name the top 1-3 meters/resources responsible for the delta.
You can show when the increase started (day/hour trend).
A budget alert is active to prevent recurrence.

Official references:

Cost Analysis quickstart: https://learn.microsoft.com/en-us/azure/cost-management-billing/costs/quick-acm-cost-analysis
Common Cost Analysis investigations: https://learn.microsoft.com/en-us/azure/cost-management-billing/costs/cost-analysis-common-uses
Create budgets/alerts: https://learn.microsoft.com/en-us/azure/cost-management-billing/costs/tutorial-acm-create-budgets
Cost Management access/scopes: https://learn.microsoft.com/en-us/azure/cost-management-billing/costs/assign-access-acm-data

If you find this answer helpful, please upvote it. If this resolves your issue, please mark it as the accepted answer so that it can help others in the community.