hey folks, does anyone have a beginner guide how to send a post call webhook to a make automation? Thank you! I have the sdk but i have no idea what to update.

To send a post-call webhook to your Make automation, set your assistant's

server.url

to your Make webhook endpoint. For example, in your assistant configuration:

json
"server": {
  "url": "https://hook.make.com/your-make-webhook-id"
}

Vapi will POST an

end-of-call-report

event to this URL after each call, which Make can use to trigger your automation. For more details on the payload, see the [Vapi server events documentation](https://docs.vapi.ai/server-url/events). Source: - [Server Events Vapi Documentation](https://docs.vapi.ai/server-url/events)

@Vapi Support Bot where do i make that edit, how do i see and edit my bots code

To edit your bot's server URL for post-call webhooks, go to the [Vapi dashboard](https://dashboard.vapi.ai/assistants), select your assistant, and in the Advanced tab, set the Server URL field to your Make webhook endpoint:

text
https://hook.make.com/your-make-webhook-id

Save your changes, and Vapi will send post-call events to this URL automatically. For more details, see the [Vapi server URL documentation](https://docs.vapi.ai/server-url/setting-server-urls). Source: - [Setting Server URLs](https://docs.vapi.ai/server-url/setting-server-urls)

how do i do it from the dashboard? if i add a server url that is a make webhook will it impact anything else?

Yes, the webhook data sent to your Make endpoint includes the call transcript. You'll find it in the

transcript

field of the

end-of-call-report

payload. For example:

json
{
  "message": {
    "type": "end-of-call-report",
    "transcript": "AI: How can I help? User: What's the weather? ..."
    // ...other fields
  }
}

You can use this field in your Make automation to access the full call transcript. [See more in the Vapi server events documentation.](https://docs.vapi.ai/server-url/events) Source: - [Server Events Vapi Documentation](https://docs.vapi.ai/server-url/events)

@Vapi Support Bot is it possible to only run that webhook on the end of call reports rather than 5 times a call?

You can't configure Vapi to only send webhooks for

end-of-call-report

events, but you can filter them in your Make automation. In your Make scenario, add a filter after the webhook module:

json
{{ if(equals(triggerBody.message.type; "end-of-call-report"), true, false) }}

This ensures only

end-of-call-report

events trigger further actions in your automation, while other event types are ignored. [See Vapi server events documentation for event types.](https://docs.vapi.ai/server-url/events) Source: - [Server Events Vapi Documentation](https://docs.vapi.ai/server-url/events)

@User any idea how I can control costs? make charges me everytime the scenario runs, but 1/5 scenarios will be the right one

Hi, to get better understanding of the cost please refer to our pricing page

Thanks Aditya, I meant from the perspective of, during the call the make webhook runs like 10 times because it is firing for every "status update", so it cost a make credit every single time. i would love to avoid firing the make webhook every time, only on the "end of call report"

Hi, do this set assistant.serverMessages to endOfCall report only.

Guys, both N8N and Make charge by workflow executions, this renders VAPI unusable because you cannot send specific post-call webhooks. I had to create a zapier account specially for that purpose using premium subscription (pro) in order to be able to filter all the numerous events that are not the "end-of-call-report" since Zapier filters (when enabled) do not count as an execution which is a big plus, and actually the only offering i found to have this feature, unless you want to host your own API endpoint and provision it... A Vapi conversation can create hundreds of events, imagine each execution would cost you money... The problem with Zapier is that they do not support wbhook responses, thus you cannot execute a webhook and get back a custom response as an api call would expect, this is besides that issue but just to keep in mind, while N8N and make do support this essential. So i eventually stopped using VAPI and started using Eleven Labs agents (Conversational AI) since you get a nice post-call webhook that is always fired, and it comes with an HMAC key to verify the payload as well. (also because of the low audio quality compared to an 11Labs agent my clients complained about that) Guys at Vapi, please listen to users and try to think as a user, it's not enough to have a running agent, the post-processing is more important or at least as important as the runtime itself, and you should make it easier for developers, add another specific endpoint or filter to send only end-of-call, this would bring many developers to you, because many of us use automations and your current paradigm is an anti-pattern. Also I've been complaining about the fact that agent does not speak first with many of the TTS providers, which again renders it useless using these, again besides the point... I'm ranting here because i had high hopes for Vapi, but those "small" things...