### Answer
---
Here are some steps to troubleshoot and reduce latency in your Conversation Flow agent:
## Check Your Latency Metrics
You can monitor latency of individual calls in the "Call History" section of your dashboard. [(1) ](
) Key metrics to review:
- **P90 (90th Percentile) **: 90% of calls have latency below this value
- **Median (50th Percentile) **: Half of calls have latency less than this value
- **End-to-End Latency**: Total time from when the user stops speaking until the AI begins responding
Production voice AI agents typically aim for 800ms or lower latency to maintain conversational flow. [(2) ](
)
## Optimization Recommendations
**Change to a Faster LLM**: Use a more efficient language model to reduce response time. [(3) ](
)
**Change to a Faster TTS**: Implement a faster text-to-speech system for quicker audio output.
## Conversation Flow-Specific Considerations
- Large prompts and many transitions in a node can add latency [(4) ](
)
- Complex tool coordination or multiple decision branches may impact performance [(5) ](
)
- Review your node transitions and ensure proper configuration
## Possible Platform Issues
Community reports indicate there have been temporary infrastructure issues causing high E2E latency that were resolved by Retell's team (such as GPU provider issues) . [(4) ](
) [(6) ](
) If your agent was working fine previousl