OpenAI API / OpenAI API
Azure OpenAI Realtime API token usage mismatch between response.done server event and actual consumption
Understand why Azure OpenAI Realtime API token counts from the response.done event don't match expected/consumed tokens; fix billing measurement accuracy Includes evidence for OpenAI API troubleshooting demand.
- Category
- OpenAI API
- Error signature
response.done server event usage object does not match actual token consumption — measuring token usage via OpenAI.Realtime .NET SDK shows discrepancy between reported and expected values- Quick fix
- Compare the failing environment with a known working setup, then change one configuration value at a time.
- Updated
What this error means
response.done server event usage object does not match actual token consumption — measuring token usage via OpenAI.Realtime .NET SDK shows discrepancy between reported and expected values is a OpenAI API failure pattern reported for developers trying to understand why azure openai realtime api token counts from the response.done event don’t match expected/consumed tokens; fix billing measurement accuracy. Based on the imported evidence, treat this as a tool-specific troubleshooting page rather than a generic API error.
Why this happens
Stack Overflow question (id 79918080) asked Apr 1 2025: developer using Azure OpenAI Realtime C# SDK logs token usage from response_done events and finds discrepancies. Real-time audio sessions are high-value production use cases. Category: OpenAI API per mapping rules.
Common causes
- Stack Overflow question (id 79918080) asked Apr 1 2025: developer using Azure OpenAI Realtime C# SDK logs token usage from response_done events and finds discrepancies. Real-time audio sessions are high-value production use cases. Category: OpenAI API per mapping rules.
Quick fixes
- Confirm the exact error signature matches
response.done server event usage object does not match actual token consumption — measuring token usage via OpenAI.Realtime .NET SDK shows discrepancy between reported and expected values. - Check the OpenAI API account, local tool state, and provider configuration involved in the failing workflow.
- Compare the failing environment with a known working setup, then change one configuration value at a time.
Platform/tool-specific checks
- Verify the command, editor, extension, or API client that produced the error.
- Compare local settings with CI, deployment, or editor-level settings when the error appears in only one environment.
- Avoid deleting credentials, local model data, or project settings until the failing scope is clear.
Step-by-step troubleshooting
- Capture the exact error message and the command, editor action, or request that triggered it.
- Check whether the failure is account/auth, quota/rate, model/provider, local runtime, or deployment configuration.
- Review the source evidence below and compare it with your environment.
- Apply one change at a time and rerun the smallest failing action.
- Keep the working fix documented for the team or deployment environment.
How to prevent it
- Keep provider/tool configuration documented.
- Record non-secret diagnostics such as tool version, provider name, model name, and command path.
- Add a lightweight check before CI or production workflows depend on the tool.
Sources checked
Evidence note: Stack Overflow question (id 79918080) asked Apr 1 2025: developer using Azure OpenAI Realtime C# SDK logs token usage from response_done events and finds discrepancies. Real-time audio sessions are high-value production use cases. Category: OpenAI API per mapping rules.
Related errors
- OpenAI API
FAQ
What should I check first?
Start with the exact response.done server event usage object does not match actual token consumption — measuring token usage via OpenAI.Realtime .NET SDK shows discrepancy between reported and expected values text and the smallest action that reproduces it.
Can I ignore this error?
No. Treat it as a failed OpenAI API workflow until the root cause is understood.
Is this guaranteed to have one fix?
No. The imported evidence supports the troubleshooting path above, but tool behavior can vary by account, plan, version, provider, and local configuration.
How do I know the fix worked?
Rerun the same command, editor action, or request. The fix is working when that action completes without response.done server event usage object does not match actual token consumption — measuring token usage via OpenAI.Realtime .NET SDK shows discrepancy between reported and expected values.