### Answer I don't have specific information about an "Opt in Stable Server" setting in the available sources. However, regarding ElevenLabs TTS errors during live calls, the available sources show that when TTS providers like ElevenLabs experience issues, Retell AI automatically tries fallback options. [(1) ]() In one documented case, error logs showed "11labs TTS error in streaming" followed by "Tts provider elevenlabs is experiencing issues, trying fallback," and the issue was resolved by changing the voice setting. [(2) ]() For addressing poor agent naturalness or TTS-related issues, you can: [(1) ]() - **Change voice**: Custom-cloned voices tend to have more naturalness issues; switching to a platform voice often improves stability - **Switch voice provider**: Different providers have different strengths - **Adjust voice temperature**: Change the temperature setting to affect vocal expressiveness If you continue experiencing ElevenLabs TTS errors, I'd recommend checking your call logs for specific error messages and considering switching to a different voice or voice provider as a troubleshooting step.

​ To learn more, see the sources I considered: [[1] Address Metric Issues](https://docs.retellai.com/ai-qa/address-metric-issues) [[2] Inbound Calls Not Working. It shows on the transcription voice but when i listen. No Voice is heard](https://discord.com/channels/1186424376300679209/1347250794826563736) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

@AI Evy The description for the Opt in Stable Server is this: " When enabled, calls and API requests from this workspace will be routed to the stable server cluster. The stable server runs a version that lags behind the main cluster for enhanced stability. An additional $0.02/min charge applies."

### Answer Based on your description of the "Opt in Stable Server" feature, this setting routes your workspace to a server cluster running a slightly older version of Retell's platform for enhanced stability, at an additional cost of $0.02/min. Regarding whether this would help prevent ElevenLabs TTS errors specifically: I don't have documentation that directly addresses whether the stable server cluster would prevent third-party TTS provider errors like the ElevenLabs issue you experienced. However, Retell AI does have built-in TTS fallback and retry mechanisms that automatically activate when TTS providers experience issues. [(1) ]() These resilience features are designed to improve reliability across the platform. For your specific ElevenLabs TTS error, the stable server option might help if the error was related to a recent platform change or instability in the main cluster. However, if the issue originated from ElevenLabs' service itself (rather than Retell's integration) , the stable server wouldn't prevent that. **My recommendation**: Before opting into the stable server (which adds cost) , I'd suggest: 1. Checking if the ElevenLabs error was a one-time occurrence or recurring 2. Reviewing your call logs for specific error details 3. Considering TTS fallback configuration or switching voice providers if the issue persists If you're experiencing ongoing stability issues beyond just TTS errors, the stable server option could provide additional reliability, but it's designed for broader platform stability rather than specifically addressing third-party provider issues.

​ To learn more, see the sources I considered: [[1] Reliability Overview](https://docs.retellai.com/reliability/reliability-overview) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.