Hi everyone, We’re still running into this error: Error Type: pipeline-error-openrouter-llm-failed It happens in less than 1% of calls, but when it does, the model never seems to connect to OpenRouter at all. We’re confident it’s not an OpenAI or OpenRouter issue, since all our other applications using the same setup work fine. What makes it tricky: - It sometimes happens right at the start of a call, and sometimes mid-call - We haven’t been able to reliably reproduce it - No tokens are used, so it looks like the request never actually reaches the model Relevant log:

{
  "level": 50,
  "body": "Model sequence failed - all attempts exhausted",
  "provider": "openrouter",
  "model": "openai/gpt-5.2-chat",
  "error": {
    "message": "providerfault-model-no-response"
  }
}

This makes us suspect it might be something on the Vapi side, possibly in how retries or routing are handled. Has anyone seen similar behavior or knows what could cause this? Not looking for paid consults, just trying to understand what’s going on. Thanks!