I am facing the issue as Vapi is not getting the response back from make. JSON returned:{ "results": [ { "toolCallId": "call_EPeJcpGomST5L4XEwR85tCy4", "result": "Unfortunately, that time isn’t available. How about one of these instead? 9:00 AM 12:30 PM 2:30 PM"}] }

Heywargod Thank you for creating the ticket! A member of our team will reach out to you shortly.

@wargod It seems like there might be an issue with the format of the JSON response you're receiving from the server. The

result

field contains a string with line breaks, which is not standard JSON formatting. JSON strings should not contain unescaped line breaks. To resolve this, you should ensure that the server encodes the response correctly. Line breaks within strings should be represented by the escape sequence

\n

. Here's how the JSON should be formatted:

json
{
    "results": [
        {
            "toolCallId": "call_EPeJcpGomST5L4XEwR85tCy4",
            "result": "Unfortunately, that time isn’t available. How about one of these instead?\n9:00 AM\n12:30 PM\n2:30 PM"
        }
    ]
}

Please check the server-side code responsible for generating the JSON response and make sure it correctly escapes special characters like line breaks. If you need further assistance with this, please let me know!

Yes, but it's still not working!

@wargod I'm sorry to hear that you're still experiencing issues. To assist you further, I'll need some more information about the problem. Could you please provide me with the following details? 1. What specific action or command are you trying to execute with VAPI that is not working? 2. Are there any error messages or codes being returned by VAPI? 3. Have you made any recent changes to your system or code that might have affected the VAPI integration? Once I have more context, I'll do my best to help you resolve the issue. If the problem is urgent or complex, it might be best to reach out to our support team for more in-depth assistance.

In the tool call id, I don't think you should add the "call" to the id

the "call_" is in vapi id by default.

Can you send me the screenshot?

SS is here. And yes I am using this to return to vapi from make https://cdn.discordapp.com/attachments/1334658296291921920/1334666831176273981/Call_id.PNG?ex=679d5cd9&is=679c0b59&hm=5d0f26ac9e4bfa42749a1dbf641ef78af632e1384882b044e72fbaaa734116f3&

@wargod It looks like there's an issue with the JSON response format you're sending back to VAPI from Make. The

result

field appears to be a string containing JSON data, which is not correct. Instead, the

result

field should directly contain the response data as a string without nested JSON. Here's how the corrected JSON response should look:

json
{
    "results": [
        {
            "toolCallId": "wwXqTC5kZBqbBmInmrbS2I1T",
            "result": "I have some time slots available for Friday, January 31: \n\n  - 9:00 AM \n  - 12:30 PM \n  - 3:30 PM\n\nWhen would you like to book?"
        }
    ]
}

Please make the following changes: 1. Remove the nested JSON from the

result

field. 2. Ensure the

result

field is a properly formatted string. 3. Remove the prefix "call_" from the

toolCallId

if it's not part of the actual ID you received from VAPI. After making these changes, try sending the response again. If the VAPI agent is still not speaking the response, please let me know, and we can investigate further.

@VapiIts still not working. and "call_" is part of toolCallID. CODE: "results": [ { "toolCallId": "call_EPeJcpGomST5L4XEwR85tCy4", "result": "Unfortunately, that time isn’t available. How about one of these instead? 9:00 AM, 12:00 PM, 3:15 PM" } ] }

@wargod Thank you for providing the updated code. The JSON format looks correct, and if "call_" is part of the

toolCallId

, then it should be included as you have done. If the response is still not being picked up by the VAPI agent despite the status 200 in the logs, it's possible that there might be an issue with how the VAPI agent is processing the response or an issue with the integration setup. Here are a few steps you can take to troubleshoot the issue: 1. Double-check that the

toolCallId

matches the ID expected by VAPI. 2. Ensure that there are no extra spaces or formatting issues in the JSON response. 3. Verify that the content-type of the response is set to

application/json

. 4. Review the VAPI documentation to confirm that the response structure aligns with the expected format. 5. Check if there are any additional logs or error messages that can provide more insight into why the VAPI agent is not speaking the response. If after these steps the issue persists, it may be best to reach out to VAPI support for further assistance. They will have the necessary tools and access to help diagnose and resolve the issue more effectively.

Can you screenshot the webhook response then?

Not the result, the configuration

@Pen 💻 I am also having the same problem, this is my response which worked previously, something seems to have changed on the VAPI side { "results": [ { "toolCallId": "call_lHZzsNi9vjMGtIdVZ8KUIIa2", "result": "The following timeslots are available: 2025-01-14T15:30:00-05:00,2025-01-14T17:00:00-05:00,2025-01-14T17:30:00-05:00" } ] }

You don't need to add the toolcall id anymore since it is the same webhook and it is running presently...

I updated it now the response is { "results": [ { "result": "The following timeslots are available: 2025-01-31T14:00:00-05:00,2025-01-31T14:30:00-05:00,2025-01-31T15:30:00-05:00,2025-01-31T16:30:00-05:00,2025-01-31T17:00:00-05:00,2025-01-31T17:30:00-05:00" } ] } However it is still not working, in VAPI log I see { "role": "tool", "tool_call_id": "call_vc3aXU9FTM2yJl4c9hFuWZR8", "content": "No result returned." },

@Pen 💻 No change, still nothing https://cdn.discordapp.com/attachments/1334658296291921920/1334673024292753418/s3.PNG?ex=679d629e&is=679c111e&hm=ff0b699428e01a8723b37a3e9e9fdc23e416932bbe920ead40684f0ece0156e4&

Copy the code, send it to chatgpt to format to the right JSON format

got it, I just removed the extra results, now I have { "result": "The following timeslots are available: 2025-01-31T14:00:00-05:00,2025-01-31T14:30:00-05:00,2025-01-31T15:30:00-05:00,2025-01-31T16:30:00-05:00,2025-01-31T17:00:00-05:00,2025-01-31T17:30:00-05:00" }

You tested from vapi and the agent didn't give results?

correct

Did you add in the prompt for the assistant to relay the available date if the time isn't available?

in make or vapi?

In the vapi prompt.

no, let me do that

It might be needed, I always add it.

which model are you using?

Gpt4o, sometimes mini

have you tested this recently? mine was working but stopped without me changing anything

I have it, but still their is no response from vapi

And the make webhook was triggered?

For me as well somthings were working but now nothing

Vapi must have changed something

Yeah, that should be the only explanation

could you check if it still works for you?

I can't test RN, not at my desktop... Will give you feedback once I do.

@wargod can you share the call ID so I can take a look?

Here https://cdn.discordapp.com/attachments/1334658296291921920/1335331070911975495/S2.PNG?ex=679fc778&is=679e75f8&hm=2a862c42c2906bd36a36eb5a3a7a6119875bca37befb7cf8b4c6566d942894d3& https://cdn.discordapp.com/attachments/1334658296291921920/1335331071155372094/s3.PNG?ex=679fc778&is=679e75f8&hm=8c187202275025d1bc7fdfc392272ebeffd0b00bd77d234b992442cfa5368c11&

@wargod

your_server_response

🔵 23:52:11:256 Server Message Successful. (https://hook.us2.make.com/s78km). Response: { "result": "I have some time slots available for Friday, January 31: - 9:00 AM - 12:00 PM - 1:30 PM When would you like to book?" } 🔵 23:51:49:144 Server Message Successful. (https://hook.us2.make.com/yohh). Response: Accepted

expecte_response_format

json
{
    "results": [
        {
            "toolCallId": "X",
            "result": "Y"
        }
    ]
}

expected_response_from_your_server

json
{
    "results": [
        {
            "toolCallId": "X",
            "result": "I have some time slots available for Friday, January 31: - 9:00 AM - 12:00 PM- 1:30 PM When would you like to book?"
        }
    ]
}

NOTE: Make sure to return the result as single paragraph without any line breaks or new lines.

I was asked to remove tool id and try. The origional code was { "results": [ { "toolCallId": "wwXqTC5kZBqbBmInmrbS2I1T", "result": "I have some time slots available for Friday, January 31: \n\n - 9:00 AM \n - 12:30 PM \n - 3:30 PM\n\nWhen would you like to book?" } ] }

@wargod I do not recall suggesting this change, nor do I believe I would have. However, I acknowledge that the current format is indeed correct.

so is their any solution?

Is it still not working yet?

unfortunately no.

same here. I return something like this and it doesn't work

{
  "results": [
    {
      "toolCallId": "X",
      "result": "some results with no line breaks"
    }
  ]
}

Thanks. The issue has been solved. It was a JSON error.