Make sure your openai api key is working or that your billing limit on vapi isn't being hit, if you're using a custom model make sure you made the patch request to the assistant