https://vapi.ai/ logo
Join Discord
Powered by
assistant-request-returned-invalid-assistant End R...
# support
m
I need help here :((( I'm getting an assistant-request-returned-invalid-assistant End Reason. I'm not sure why. This is the payload that my server is responding with when the server url is hit by vapi: {"assistant": {"voice": {"model": "eleven_turbo_v2", "voiceId": "KaJWDrV3K2SAk5HGQWB2", "provider": "11labs", "stability": 0.5, "similarityBoost": 0.75}, "model": {"model": "gpt-3.5-turbo", "messages": [{"role": "You are an experienced financial department support agent. Your role is this this and this. their payments for the trips they completed.", "content": "Testing content"}], "provider": "openai", "maxTokens": 250, "temperature": 0.7, "recordingEnabled": "false", "firstMessage": "Hello John Doe! Thank you for calling the financial department. How can i help you today?", "voicemailMessage": "You have reached the our voicemail please try again later", "transcriber": {"model": "nova-2", "language": "en", "provider": "deepgram"}}, "id": "2u598320dhjskhr0312", "orgId": "fhdkjrh1983h89f1h88133", "createdAt": "2024-06-03T13:20:51.480215", "updatedAt": "2024-06-03T13:20:51.480248"}} https://cdn.discordapp.com/attachments/1247240628866388099/1247240629323563069/Screenshot_2024-06-03_at_1.27.43_PM.png?ex=665f4eda&is=665dfd5a&hm=8cb87c9baf95f1aba5fa1e60b8420acb4e3989290ed38ac8db149789b806261b& https://cdn.discordapp.com/attachments/1247240628866388099/1247240629977878568/Screenshot_2024-06-03_at_1.28.54_PM.png?ex=665f4eda&is=665dfd5a&hm=ac78a3500a725d75bef44ff9ca0aec33b034b21f4d33cef2cdf54aca339ddf20&
v
Hi @MikeKraz, I'm an AI assistant for Vapi. I'll analyze available sources and get back to you shortly.
----------------------------- **While I generate an answer, here are some sources that might be helpful**: * [Create Assistant]() * [Create Phone Call]() * [Update Assistant]() * [Call Ended Reason]() -----------------------------
### Answer It appears that the 
assistant-request-returned-invalid-assistant
error may be occurring because the response from your server does not conform to the expected schema that Vapi requires for its assistant configuration. This could happen for various reasons such as missing required fields, incorrect data types, or fields that Vapi does not recognize or expects differently. To resolve this issue, ensure that the payload correctly follows the structure and specifications expected by Vapi. A typical configuration should include: 1. Valid 
model
and 
voice
settings that match the expected formats. 2. All required fields like 
transcriber
, 
firstMessage
, 
voicemailMessage
, and others must adhere to the expected formats and data types. 3. Ensure that there are no additional fields that are not recognized by Vapi, or that the values provided do not violate the schema's expectations. This error often relates to how the assistant data is structured or how it is processed on Vapi's end when it tries to utilize the assistant configuration provided by your server. For a detailed understanding of the correct schema and required fields, you may want to refer to the official documentation on how to structure the assistant's configuration. If this continues to be an issue, reviewing the server-side logs for any discrepancies in the request handling might be necessary. Consider validating your JSON response against the expected schema using tools or libraries designed for JSON schema validation to catch any mismatches or errors before sending the response to Vapi. This preventive measure can help ensure the response is fully compatible with what Vapi expects to receive, reducing the likelihood of errors such as 
assistant-request-returned-invalid-assistant
.
Got it. Tagging @User for additional help.
m
FYI this payload was also invalid: {"assistant": {"firstMessage": "Hello Michael Krasnik! Thank you for calling this department. How can i help you today?", "model": {"provider": "openai", "model": "gpt-3.5-turbo", "messages": [{"role": "You are an experienced financial department support agent for company Mile.", "content": "You must make jokes."}]}}}
s
@MikeKraz I think this issue was solved during OH
m
It was thank you so much!
@Sahil do you mind hopping into vapi-support i have a super quick question?
s
Can you just ask the question directly in support thread and if it is something private send me a message on dm
8 Views
PreviousNext
Powered by