I want to connect fibremax.au voip to vapi using sip trunking. How do I go about this for success

Hi Vapi Support, I’m running into a persistent issue where webhooks sent from Vapi to our GoHighLevel (GHL) workflow are not always triggering as expected. ✅ What's Working: When calls end normally (e.g., "voicemail" or successful engagements), the end-of-call-report webhook is received and contains: transcript analysis recordingUrl assistantOverrides.variableValues customer.number These trigger GHL workflows correctly. ❌ What's Failing: When the call ends with: json Copy Edit "endedReason": "customer-did-not-answer" the webhook is: Still received at our GHL endpoint BUT missing key fields (no transcript, analysis, etc.) So it does not match the GHL webhook sample structure and is ignored silently As a result, we have no visibility in GHL logs and the lead is completely missed in the automation. 🔎 Questions for Support: Can Vapi force inclusion of transcript, analysis, and recordingUrl fields in all end-of-call-report payloads, even if empty or null? Can we add a fallback message or enforce structure from our side (e.g., in assistant config or via assistantOverrides) to normalize webhook payloads, especially for calls that are missed or unanswered? Is there a recommended way to ensure that customer.number or assistantOverrides.variableValues.phone is always included in the top-level payload for mapping downstream? Can you confirm if there's any webhook delivery logic that filters or suppresses minimal call logs based on endedReason? We would prefer to resolve this within Vapi if possible before setting up external proxying/filtering middleware. Let me know what options you recommend to standardize webhook output and ensure all end-of-call-report events can be reliably processed. Thanks so much for your help!

The Problem When I trigger outbound calls via GoHighLevel (GHL) using a batch of 10 unique contacts (each a separate object in the customer array, not a shared array like customers), I notice that not all calls consistently appear in Vapi’s "call logs." As a result, my inbound webhook (which listens for Vapi's call log events) does not always receive a callback for every individual contact that was triggered through GHL. This leads me to question: Is this an issue with how Vapi logs or emits events? Could GHL be firing webhooks too quickly, overwhelming Vapi or causing missed logs? Is it normal behavior for Vapi to occasionally drop logs, especially under high-frequency bursts? Questions I Need Clarity On Reliability: Is there a known limitation or occasional "bug" where Vapi might miss certain call logs even though the API was successfully triggered? Rate Handling: Should I be throttling or dripping my GHL webhook requests (e.g., spacing them out with a delay per contact)? Best Practices: Is there a threshold or recommended request frequency when sending bulk but individual requests to Vapi (10-50 contacts, one-by-one)? Goal Ensure 100% delivery and detection of call logs for each triggered contact. If I’m sending 10 contacts to call, I expect 10 Vapi logs and 10 webhook callbacks — no exceptions.

Trying to signup via email, I get a verification email, open it and it leads to a blank grey screen. Tried: - Deleting cache - Incognito mode - Different emails and login methods (google and github) - Different browser

Analysis summary of my calls stopped generating randomly. It was wokring for many weeks as expected. to debug i made very simplified prompt and it worked in some case, but for complex prompt it is still NOT working. Plese help urgently. Call id - fee64acd-cb6c-4de4-9928-2e50bb59a64a Time - 11 Jul 2025, 16:52 I use Vapi from Dasboard directly. In case any param needs to be updated, please provide detailed step by step instructions.

I need assistant to do some calculations like addition, mult, etc. But it reads them outloud. Is there any way to force this to do it on the background and just say the result rather than reading the whole calculation? #1211483291191083018

Vapi assistants are not able to recognize injected variableValues and sometimes will hallucinate. How can I resolve this?

When i try to integrate a vapi agent inside of telnyx it says an unexpected error happened what to do

I was part of the voice AI Maven course by kwindla, and I filled out the form to get my credits, but I don't see them on my account. please help

When my agent is booking a call and it’s confirming contact info. It’s saying the random contact id to the person on the phone how can this be removed?

call ID : 20abd46b-0eff-41d1-8e3e-e52ad80291aa timestamp 2025-07-10T06:20:00.574Z expectation : phone call initiated response recived : phone call cannot be established , though everything was taken care of like concurrent calls were with in the limit, this particular call got error , but another subsequent call after 20 seconds was initiated successfully

@Vapi Support Bot would it be possible for you guys to update your service with out new SSL certificate? or what does this error mean?

When I try to save the transfer it says there's an error https://cdn.discordapp.com/attachments/1393329133676658718/1393329134205403196/image.png?ex=6872c660&is=687174e0&hm=bd6a849fb6f4df927e918ad8fdf8d5b73652cf4183d6385b4ed61943119eed68&

Hi, I've been trialing Vapi but having some issues getting started, would appreciate some help 🙂 I noticed that when using Rime voices, the call isn't starting at all (testing in the vapi platform) Also with default transcription settings, it seems like transcription isn't really working. Are there any reasonable defaults your recommend here?

Hello! I have this weird issue where the assistant repeats itself around 3m15sec, do you know what could have happened? https://dashboard.vapi.ai/calls/c2439d03-ee18-407d-b964-32f246c071bb I checked and it's not requesting a tool or anything, it just says: 00:05:54:455 [LOG] Model output: Thank you, Jamie. 00:05:54:456 [CHECKPOINT] Voice audio requested 00:05:54:456 [LOG] Voice input: Thank you, Jamie. You'll be bringing three bags when you travel to Washington D C to see your mother.Thank you, 00:05:54:499 [LOG] Model output: You'll be bringing three 00:05:54:549 [LOG] Model output: bags when you travel to 00:05:54:597 [LOG] Model output: Washington DC to see your 00:05:54:637 [LOG] Model output: mother. Do 00:05:54:638 [LOG] Voice input: Jamie. You'll be bringing three bags when you travel to Washington D C to see your mother. 00:05:54:682 [LOG] Model output: you have any specific 00:05:54:693 [CHECKPOINT] Voice audio received 00:05:54:728 [INFO] Turn latency: 2118ms (transcriber: 0ms, endpointing: 501ms, kb: n/a, model: 983ms, voice: 600ms) 00:05:54:728 [CHECKPOINT] Assistant speech started

I'm not able to publish my changes whenever I try to write a prompt for the success evaluation. The publish button remains disabled. This does not happen, however, when I try to change the summary prompt.

Hello. A Vapi agent I created recently received a call from a customer. While helping the customer, the agent got stuck repeating a set of lines over and over again until the caller interrupted it. The call then continued as normal. What may have caused this repeating error? The call ID: 582c99ad-d6eb-4816-884b-c61e9491bb52 Call Start Time: 2025-07-11T14:36:08.723Z (July 11, 2025 10:36AM EST) https://cdn.discordapp.com/attachments/1393312064616792206/1393312064843419788/image.png?ex=6872b67a&is=687164fa&hm=ff09c041b649b75e5773e9e2b95f6dc46915d199d74e81d46e191d4202422422&

Dear VAPI Support, I’m reaching out to report an issue with the identity verification process on your platform. I’ve attempted to complete the verification multiple times over the past two days, but the system gets stuck in an endless loading state and never finishes. There is no result, confirmation, or error message — it just keeps spinning. Since your platform uses Veriff for identity verification, the issue might be related to that integration. I’m attaching a screenshot of the screen I see during the process, in case it helps with troubleshooting. At the moment, I’m unable to access my VAPI account because of this, and my AI agent is locked inside that account — so this issue is preventing me from continuing my precious work. I’d appreciate it if you could look into this as soon as possible or let me know how to proceed. Thank you for your help. https://cdn.discordapp.com/attachments/1393289086319722546/1393289086927765574/image.png?ex=6872a114&is=68714f94&hm=0f744d77e59b32f64c543305418c324833a40e71b5d5d23e17b647689cffca5d&

Trying to send out invitations to a team member so they can get access to a vapi org and they are telling me no email is being received.

When I login with my Google account to VAPI dashboard, I am correctly redirected to the dashboard... However, the dashboard is just a grey screen and it does not load anything. I do not know if it has something to do with my OS (Windows 8.1 - I know, it's not the best), but I need help. Thanks. https://cdn.discordapp.com/attachments/1393273481923137608/1393273488684351669/image.png?ex=6872928d&is=6871410d&hm=b4733843081dc84dfbadedb028d5a7fe7b956f26db084c691130ff27daee3ee4&

When I am using Groq Llama4 Maverick on test call it works great but when I use it on real call the latency goes to 5000ms whereas it shows 640 ms on Web and 1140 ms on Twilio ( I am using VAPI Number)

I’m am using the control url, but it keeps timing out. I have been trying to use update session but I think that is for SMS conversation as checking the response to a call there is not “session_id”. There is “id” but I have tried that whilst on a live call and it says the session does not exist. Is there a workaround? I am trying to have my agent read the address received via SMS because sometimes to prevent confusion. Thanks

Please provide me an api to change the protocol

Good afternoon, all my assistants are getting some issues, after some seconds the call get hanged up and there's this errors: call.in-progress.error-vapifault-azure-speech-transcriber-failed

Hi Guys, I'm having an issue with call transfers when I need to have the AI dial an extension. We currently have this setup however our customer's phone system doesn't recognize this in the same way as a human dialing the extension and thus the call recipient's Voice Mail never kicks in. Have you guys come across this issue before? Do you have any insight as to how we can trouble shoot this? These people with the GoTo system for some reason when VAPI AI makes the transfer to the person at their extension, their phone will ring but for some reason their voicemail never kicks in. I’ve tested this many times today. VAPI making the transfer vs. My calling in their direct line and pressing their extension manually, in that case their voicemail will kick in. organizationid: 07c645bc-b098-43c4-b631-7fc5fae7d0c5 assistantid: cfe7d417-58ef-4487-84e7-a69696c2be96 callid: a8d46633-1bf4-432d-9510-a1ace7cad621

I’ve created a test “Transfer Call” tool between two assistants I set up, intending for one assistant to transfer the call to the other. However, when I trigger the transfer, I receive a couldn't get assistant error. When I remove the tool from usage, everything works fine again — the issue disappears entirely. Looks like the transfer tool is causing the assistant resolution to fail. Let me know if logs or assistant IDs are needed.

The user voice isn't being heard by the AI agent.

I have configuerd my server URL at an org level in Vapi. When HIPAA is disabled, on my server, I am getting full payload including transcript and recordingURL (.wav file). I was able to copy this .wave file to my S3 bucket. When HIPAA compliance is enabled, I am getting payload with transcript but recordingURL of .wave file is missing. Our environment is HIPAA compliant and we have a client who needs transcript + recording audio. What is the best way to go about this?

It worked before. But we notice recently the placeholder is just being kept as it is. For example: - Customer name is '{{customerName}}' the provided parameter is: {customerName: ''} So it may end up make the AI understand customer name is {{customerName}} in real text. Is there any suggestion on improvement this. We want to verify if the customer name is empty, prompt the AI to collect for the name.

We’re currently exploring high-volume usage of your platform and would like to know more about your agency plans. We’re expecting to exceed 50,000 minutes of call usage, and we're keen to understand if there are custom packages, volume discounts, or tailored offers available for agencies operating at this scale. Looking forward to your response and a potential collaboration. Best regards, Kishan Sankar Automation Engineer @Sahil @Vapi Support Bot