https://vapi.ai/ logo
Join Discord
Powered by
access entire message from client app
# support
r
i am trying to show the messages too in client as the conversation is going on, how i tried transcribe and conversation update both keep sending same stuff multiple times and it's no the entire message
v
@research_boy It sounds like you're trying to display real-time transcriptions of the conversation in the client. If you're experiencing issues with the 
transcript
and 
conversation-update
messages being repetitive or not showing the entire message, there might be a configuration issue or a misunderstanding of how these messages should be handled. The 
transcript
message typically provides real-time transcription of what's being said, while the 
conversation-update
message can include various types of updates about the conversation state. To assist you better, could you please provide more details about how you're currently handling these messages in your client? This will help me understand what might be going wrong and how to correct it. If there's a specific function you'd like to call or a particular issue you need help with, please let me know!
You can continue this conversation by mentioning me in the message.
Got it. Tagging @User for additional help.
s
which server events your are looking into.
r
all messages, i am following the react sample app
s
hey can you share the config/code-snippet for receiving client messages.
r
vapi.on("conversation-update", onConversation); vapi.on("message", onMessage); tried using this two
const onConversation= (conversation) => { console.log("Event: conversation", conversation); setConversation(true) }
const onMessage = (message) => { // console.log("Event: message", message); if ( message.type === "TRANSCRIPT" && message.transcriptType === "PARTIAL" ) { // console.log("Active Transcript: ", message.content); setActiveTranscript(message); } else { // console.log("New Message: ", message.content); setMessages((prev) => [...prev, message]); setActiveTranscript(null); } };
here
s
use ClientMessageTranscript event and check if transcriptType is final so you will get the final transcript of both user and assistant.
You can also filter based on the role.
r
transcript is basically deepgram output right? what i need is deepgram output in case of user dialogue and model generated text for the assistant part
s
you get that but we are using streaming and end of speech detection so you also get partial script.
r
so i will have to handle that in client app by devising some approach to differentiate this right?
s
no the differentation of transcript by role and is partial/final si already aavilable using api/events.
r
but i have seen specific spelling mistakes in names when it comes to transcript output , this names are model generated and verified at model endpoint
s
please share a example with call id and timestamp.
no response closing the ticket.
3 Views
PreviousNext
Powered by