Action Step: Go to your [Vapi dashboard](https://dashboard.vapi.ai/), navigate to Observe > Call Logs, and review the error messages and call flow for failed calls. Also, check the [ElevenLabs Status Page](https://status.elevenlabs.io/) to ensure there are no ongoing service issues. This will help you determine if the problem is with your API key, rate limits, or a provider-side outage before making further changes to your setup. Tip: Focus on any error codes or disconnect reasons in the logs for clues about the WebSocket drops. For more details, see the [Vapi debugging guide](https://docs.vapi.ai/debugging). Source: - [Debugging voice agents](https://docs.vapi.ai/debugging)

I’ve already checked Call Logs and reviewed all the failed calls. In the Logs tab, I can see multiple ElevenLabs WebSocket closed entries but no clear close code or reason in the JSON. Everything else in the flow appears normal. From what I can tell, the Vapi side is functioning correctly, so the issue might be related to ElevenLabs or how the WebSocket stream is handled now that I’m using my own keys. Also, how can I get direct support from Vapi? I pay for the service, and while Discord is helpful, it doesn’t feel like a proper support channel when we’re dealing with technical service-level issues.

Hi socialbrainstorm, If you encounter issues with your voice assistant, you can start by using the debugging tools available in the [Vapi dashboard](https://dashboard.vapi.ai/). Here are some steps you can follow for common issues: 1. **Test the Voice Agent**: - Use the "Talk to Assistant" feature to directly test your agent. - Use the "Call" option to test workflows. 2. **Check Call and API Logs**: - Navigate to

Observe > Call Logs

to review call details, identify failures, and analyze workflow flows. - Check

Observe > API Logs

to monitor API requests and authenticate errors. 3. **Analyze Webhook Deliveries**: - Go to

Observe > Webhook Logs

to ensure webhook deliveries and check server responses. 4. **Verify Integration and Service Status**: - Check the [Vapi Status Page](https://status.vapi.ai/) and corresponding provider status pages, like [ElevenLabs Status](https://status.elevenlabs.io/). If issues persist, consider these practical steps: - Switch voice providers if the agent sounds robotic. - Adjust model configurations, such as prompts and temperature, to improve response quality. - Test individual tools under the

Tools

section to debug parameters or API calls. For more detailed troubleshooting, refer to the [debugging guide](https://docs.vapi.ai/debugging). If this doesn’t resolve the issue, please let us know and we’ll be happy to step in and assist further.