Dear VAPI Support Team, We’ve been experiencing repeated call interruptions due to the following error: call.in-progress.error-providerfault-eleven-labs-500-server-error This occurs mid-conversation and appears to originate from a server-side issue with ElevenLabs. Below are some recent callIds where this error occurred: Call ID Date & Time (UTC) 8338f286-b1f6-499d-96ec-6120c7cca8ee August 7, 2025, 15:59 8801e645-a711-46e5-bfec-5e87aacce791 August 6, 2025, 18:29 f91ea64c-43b0-45e1-aa36-78c210a6d509 August 6, 2025, 16:44 76eb4a46-69a8-48cd-9222-76074f1fe8f8 August 6, 2025, 16:28 9ac21030-ea41-4363-853e-a5dd61c374e0 July 6, 2025, 22:51 Could you please investigate these calls and confirm whether the issue lies with ElevenLabs? Additionally, we would appreciate any suggestions or workarounds such as implementing a fallback provider, retry mechanism, or any limitations we should be aware of when using ElevenLabs. Please let me know if you need any additional details. Best regards,

To mitigate interruptions from ElevenLabs errors, configure a voice fallback plan so your assistant automatically switches to another provider if the primary fails. Here’s a sample configuration snippet for your assistant’s voice settings:

json
"voice": {
  "provider": "elevenlabs",
  "voiceId": "your-elevenlabs-voice-id",
  "fallbackPlan": {
    "voices": [
      { "provider": "openai", "voiceId": "shimmer" },
      { "provider": "playht", "voiceId": "jennifer" }
    ]
  }
}

Set this via the API; Vapi will try each fallback in order if a failure occurs. [See full details in the documentation.](https://docs.vapi.ai/voice-fallback-plan) Sources: - [Voice Fallback Plan](https://docs.vapi.ai/voice-fallback-plan) - [Call ended reasons](https://docs.vapi.ai/calls/call-ended-reason)

Hi, is it fixed now?

We are getting the same, is it related to Optimize Streaming Latency parameter?

We are sssing this still, around 2-3% of calls get this error.

We are not getting this anymore. so solved from our perspecitve