Getting a pipeline-first-error-first-message-faile...
# support
t
Calls keep timing out ID: 4c5e764c-caf4-46be-8815-e6520a9edd0a Also, the phone agent here says jibberish: d1e6b03e-4e5d-4cd8-bae0-bdc2c3ac33a9
v
Hi @toni.jay, I'm an AI assistant for Vapi. While a team member reviews this question, I'll look through relevant sources and see if I can provide suggestions. Please tag @Vapi Support with any follow-up questions.
----------------------------- **While I generate an answer, here are some sources that might be helpful**: * [Get Call]() * [Set up your Server URL]() * [Glossary]() * [Dashboard]() -----------------------------
### Answer The "pipeline-first-error-first-message-failed" message and the issue with the phone agent saying gibberish indicate there might be problems with the initial setup or execution of the call. This could be due to a variety of issues, such as incorrect assistant configuration, issues with the voice model or transcriber settings, or problems with the integration of the assistant with the phone system. For the calls timing out, ensure that your setup meets the expected configuration and that all required parameters are correctly set. It could also help to review the server logs for any errors or warnings that could give more insight into why the calls are not completing as expected. For the assistant saying gibberish, this could be related to the voice model configuration or an issue with the text-to-speech (TTS) service. Ensure that the language and model settings are correctly configured for your use case and that the TTS service is functioning as expected. Unfortunately, without specific details on the configuration and setup, it's challenging to provide a more detailed solution. Reviewing the assistant and call setup configurations against the documentation might help identify any discrepancies or misconfigurations. For further assistance or if the problem persists, it's recommended to reach out to Vapi support for a more in-depth investigation and assistance.
s
This is happening due to some unusual activity detected on your ElevenLabs free plan account. So, kindly change it or upgrade to a new plan.
Consider using the GPT-4 model for the gibberish part. It will provide you with a much more accurate response.
t
Thank you. GPT-4 model's increases the latency. Any ideas how I can improve this?
s
You can use rime-AI as your voice to cover up the latency increased by the GPT-4 Model.
t
Thank you Sahil.