Creating a new API Tool or updating an existing Tool errors out, with the attached error https://cdn.discordapp.com/attachments/1438787707085258837/1438787707324203078/Screenshot_2025-11-14_at_6.08.12_pm.png?ex=691826f8&is=6916d578&hm=1f923fa4b648693a73c9f252b0372b0564f1c4b3e671a179b8281eae58fde712&

VAPI Assistant - I'm trying to find a reliable way to use a unique callId to correlate all tool calls within the same Vapi call session - in API tools and custom function tools. Better if dynamic variable available. In the logs => Tool category => I can see the callId in the attributes property, but not sure how can I use it. Didn't find anything useful Community thread - https://vapi.ai/community/m/1381689099240472586?utm_source=chatgpt.com Dynamic variable - https://docs.vapi.ai/assistants/dynamic-variables#default-variables

VAPI TTS is down the agent will not speak back, if I change it to openai or any other TTS the agent start speaking, the only thing that works is the first message, and the voice ID is spenser.

019a7bbc-94ce-7885-898a-53c2c60e79c8

Hi! I'm in need of assistant. I'm helping a customer to integrate an AI agent to help their clients. This customer has a third party company that does the first calling for them and then transfers the clients (while on the call) to their SDRs. I understood that i need to create a "Free Vapi SIP" in order to be able to receive transfer as a fallback in case all SDRs are busy. The problem is that after i created the Vapi SIP and assign an AI assistant to it, the call never get's through when I try to test with Linphone. It just keeps calling for ever.

Hi vapi Support team, In the account, how do we change name? Does this affect the outbound caller id? Appreciate the help to fix, thanks

Hello, I'm having trouble with double booking using the google calendar integration. I had a promp that made the AI check for double booking and avoid it. It used to work well back in september/october but now it stopped working, same set up, same promp. I made sure the right calendar is selected, the right time frame too. It seems like something changed on Vapi or Google's end and now the tool doesnt work properly. Any current fix for that? Thank you https://cdn.discordapp.com/attachments/1438617982044737659/1438617983068012735/rn_image_picker_lib_temp_0c3e5809-c2dc-4430-8108-9022161e855a.jpg?ex=691788e7&is=69163767&hm=8b20090322612069163c28803d35a0a60d565726ff61999d1f5c9d2bc75d1245&

is open ai down ? all assistants are doing the same will not respond to customer after first message is spoken #1211483291191083018

Reference call id: 019a7e3e-b300-7113-8a1e-f355efff124e Previously, if we send the type mcp to the assistant, the tool calls were calling function based on the mcp get_tools. However, since few days ago, the assisstant seems to use mcp_tools like a regular function instead of mcp server. Nothing changed on the backend side, not sure what's going on in here. Working example call id: 019a17b7-2814-7999-b0ce-bbd052d41801 Thank you

Good morning, Following up about this. Attached is a screenshot showing a charge of $71.99. Could you please assist us in reviewing this transaction? The accounts we currently have registered are under the following email addresses: support@marketingdpt.co admin@marketingdpt.co devteam@marketingdpt.co Thank you, and we look forward to your assistance. Regards, Daniella Kuan DPT Group https://cdn.discordapp.com/attachments/1438524886619258910/1438524886912991313/unnamed.png?ex=69173233&is=6915e0b3&hm=ac172a4b9a3d8140e18de9a5b24697eb64f3a6e2f64268fe4f82a4f84a7004bc&

Hi vapi Support team I tried to create vapi api request tool and I can create this But I can't update this, I hope you fix this issue

Hi everyone, I'm trying to find the complete JSON structure for the SIP credentials API call (https://api.vapi.ai/credential). The documentation at https://docs.vapi.ai/advanced/sip/sip-trunk shows a Zadarma example with limited fields, but the GUI (screenshot attached) displays many more options like domain, realm, usePublicIP, etc. Question: Does anyone have the complete API documentation or can share the full JSON body structure that matches all the keys from the GUI screenshot? I want to create the SIP trunk via API instead of using the potentially buggy GUI. Thanks! https://cdn.discordapp.com/attachments/1438510588194979942/1438510589092696114/image.png?ex=691724e2&is=6915d362&hm=68591e3e2806de0815ea70434c8861b510484fea53eaf00fb990b57431c342c6&

Hi, I've noticed that my campaigns are not triggering, long after their scheduled time. An example is campaign_id=

f6b64f95-b48a-41ce-bf2e-7a71b5c8a615

, which was due to be scheduled at 11:16am GMT today. I ended up cancelling it at ~12:00 noon, as I was attempting to demo this to a customer, but it never worked. This isn't the first time that I've seen this happen, but it was the first time that was in real-time with a customer, so it looked very bad. Can you see what went wrong, and are there additional parameters that I need to make my customers aware of (e.g. a campaign will trigger within X hours of the scheduled start datetime?) My Org ID is

bbef089b-7bd1-4385-981a-8fec3b32ea36

The Agent uses the Content in the Request Response Delayed field the same way in every language. This message needs to be changed in every language. How can I do this?

Transcript: Assistant Hello, this is Eva from ABC Lawyers. Can I help you today? User I want to speak to a lawyer, please, about a workplace injury. My name is Grace Stevenson, and I am uh, uh, it happened in my workplace in, uh, in Queensland. And I was I got some back injury. Assistant I'm sorry to hear about your accident, Grace. Do you mind if I ask you a few questions to better assist you? User That's fine. Assistant Thank you, Grace. Can I please have your best contact number? User Yes. It's o 4, double 1, 0, double 1, 0 3 5. Assistant Let me confirm that number. That's 0 4 1 1 0 1 0 3 5. That's only 9 digits. Could you please repeat the full● PROMPT: ## Phone Number Protocol ### Simple Rules 1. "double [digit]" means write that digit TWICE - "double 1" = 1 1 - "double 6" = 6 6 2. "triple [digit]" means write that digit THREE TIMES - "triple 9" = 9 9 9 3. Single digits = write once - "oh" or "o" or "zero" = 0 - "one" = 1, "two" = 2, etc. 4. Process each word separately - don't combine them - "double 1 double 1" = 1 1 + 1 1 = four separate ones (1 1 1 1) 5. Australian numbers are always 10 digits including the zero at the beginning - count before confirming ### Quick Examples - "o 4 double 1 0 double 1 6 3 8" = 0 4 1 1 0 1 1 6 3 8 (10 digits) - "oh 4 double 1 double 7 6 7 8 5" = 0 4 1 1 7 7 6 7 8 5 (10 digits) - "oh 4 double 1 oh double 6 triple 9" = 0 4 1 1 0 6 6 9 9 9 (10 digits) CALL ID: 019a7d02-ac99-7222-963c-190ec6b2ce01

Please can you send it directly to me for privacy -- thanks.

Hello, I have a question regarding call latency in my recordings (see attached screenshot). In the Latency Summary section, it shows me Average Turn Latency 1292ms. However, when listening to a real recording, the actual response latency (from the end of the user's call to the start of the AI's response) is subjectively close to 3 seconds. My questions are: Why is there such a difference between the measured value of 1292ms and the actual real-time latency that I hear? How can I effectively reduce the overall response latency? Which of the components (Voice, LLM, Transcriber, Endpointing) have the greatest potential for optimization? Context and Goal: For comparison: when testing voice models (e.g. via Google AI Studio with Gemini Flash Live), latency is often less than one second and transcription is virtually flawless. I would like to achieve similar speed and fluency within VAPI. Do you have any specific recommendations for setup, model selection, or configuration that would help me minimize this latency and achieve near-instantaneous and smooth conversation? https://cdn.discordapp.com/attachments/1438460648982843472/1438460649473839104/Snimka_obrazovky_2025-11-13_o_10.28.10.png?ex=6916f660&is=6915a4e0&hm=ae7efeb5dfcd135dcc0457a3da660c7e44856a13e059d91ee08fdbc4c673f43d&

How to fix pls https://cdn.discordapp.com/attachments/1438349577878245417/1438349578150739968/Screenshot_20251113-070449_Gmail.jpg?ex=69168eee&is=69153d6e&hm=df6475992a5852fedd7ffafe2dc7ca98b8152911c7d0e058b32a086242459673&

I have this call: https://dashboard.vapi.ai/calls/019a7a23-607b-7bb9-88c7-58d12e14f287 I don't understand why it hung up while the person was talking, without warning. How do we prevent this from happening again?

Hello, When sending a call from the dashboard, how can I control how long it dials the customer before hanging up OR reaching voicemail? Is this accessible in the dashboard? Thanks

I am writing to seek urgent clarification regarding the cost breakdown in my recent call usage, as I am facing confusion with your pricing system, specifically concerning the Speech-to-Text (STT) component. have noted the standard call cost, which I understand to be $0.05 per minute. However, I have also observed a separate charge for STT (Speech-to-Text) within the VAPI call cost breakdown. This is confusing for me because I am under the impression that the STT service is handled by my own configured external provider. Specifically: My current configuration utilizes an Azure Transcriber connected via my own API key. Therefore, I anticipate that all Speech-to-Text processing costs should be billed directly to my Azure account, not through VAPI. I urgently request clarification on the following points: Why is the STT charge being deducted on the VAPI side when I am using a self-managed external transcriber (Azure)? Is it possible that VAPI is providing and charging for a separate, default STT service, even when an external API key is configured? Could you please confirm the exact service or process that is incurring this specific STT charge in my usage report? Thank you for your prompt attention to this matter and for clarifying the billing details, especially concerning the unexpected STT charge.

Hello everyone I need your help As you can see photo, I call webhook which send message to phone. But only in webchat, I always receive empty message. Please help about this

I want to transfer the call but the call ends. Context: I have created a Vapi number (US number), and when I tested it calling from my phone and asked to be transfered using the Forwarding Phone Number function to a mexican sumber the call ends. I added to the system prompt the instruction: if the intention of the user is to... transfer the call using the transfer call function to the number .... I also tried using the transfer call tool. I added the E164 format and all. In the log it appears: [ { "name": "transferCall", "toolCallId": "call_axDPotgfnHm9M9942QBxWmRA", "type": "transferCall", "result": "Transfer initiated." } ]. I also run: import { VapiClient } from "@Vapi-ai/server-sdk"; const client = new VapiClient({ token: "xxx" }); const call = await client.calls.get("xxx"); console.log("End reason:", call.endedReason); and it appears: End reason: assistant-forwarded-call how can i solve it?

Built a voice agent that’s connected to a form. As soon as the form is submitted, the VAPI agent makes an outbound call to the number and starts qualifying the lead. I sent a demo to a prospect to test it. He filled out the form, got the call, and picked up—but something strange happened. When he answered, the phone still kept “ringing” for a few seconds during the call (you can hear it in the recording). Then it suddenly showed that the call was connected. Has anyone else run into this before? Any idea what might cause it or how to fix it? Recording link: https://storage.vapi.ai/019a78ab-da44-7cce-b903-eb72a7961484-1762961166034-d6c62340-0ff7-4950-b55d-dc1b92410bc7-mono.wav

Trying to setup outbound calls. However i keep getting 'customer-busy' but I even don't see the call on the BYO SIP trunk in my SBC logging. callid= 019a78ca-6736-7ffd-91fe-e9bfeca435ff Timestamp = 2025-11-12 15:58:48.986+00 JSON message = { "assistantId": "588f8e84-432d-4cfc-8824-xxxxx", "phoneNumberId": "dce51408-5a5a-47b7-b51a-xxx", "type": "outboundPhoneCall", "customer": { "number": "+316XXXXXXXX", "numberE164CheckEnabled": false } } Any ideas how to fix this? ps; when i forward an inbound call to VAPI assistant to a PSTN number i see that the calls is sent via the BYO SIP Trunk. ^Marty

Hello Vapi Support, We’re trying to understand why Azure STT (Vapi-managed) billing is higher than our measured user speech on a specific call. For call : 019a2619-9be9-7992-ac29-e76ee6536d27, the dashboard shows Azure STT cost = $0.25310806. In the STT selector the rate shown for Azure is $0.017/min (in pricing page). At that rate, the charge implies about 14.89 minutes of billable audio ($0.25310806 ÷ $0.017). Our own aggregation of role:"user" utterances for this call totals ≈1.4526 mins, so we expected roughly cost=0.0242103 h×$1.00/h=$0.0242103 If the underlying Azure plan is $1.00/hour. If the underlying Azure plan is closer to $2.00/hour ≈ $0.03333/min, then $0.25310806 would correspond to ~7.59 minutes, still above our measured user speech. That’s why this looks unexpected on our side. 1- Could you please confirm the exact billing model used for Azure (Vapi-managed) and help reconcile this call? 2- Do you bill both directions when “transcribe agent/both sides” is enabled? And if it’s disabled, can agent/TTS audio still be captured or billed in any path? 3- If Call recording is ON, does auto-transcribe recordings add minutes in addition to live streaming STT? 4- What rounding/minimums apply (per-utterance rounding, trailing-silence padding, minimum block per stream, etc.)? With many short turns, how much overhead should we expect? 5- Do reconnections or STT restarts start new minimum blocks? 6- Which exact Azure Speech SKU/region backs the “Azure — $0.017/min” option (pure pass-through of a negotiated rate, or any additional factor)? 7- Could you share a minute breakdown that sums to $0.25310806 (user live STT, agent live STT if any, recording-transcription minutes, rounding/blocks, restarts)? so we need to understand the math behind : "costs": [ { "cost": 0.25310806, "type": "transcriber", "minutes": 7.5968, "transcriber": { "provider": "azure" } } ]

I can’t create or update tools in my workspace — even new ones fail to save (tried incognito). Can you check if my workspace is locked on the backend?

I’m trying to transfer the call. In my logs, it shows that the tool is being triggered, but I’m not receiving the call transfer. Call ID: 019a781e-7471-7ff1-97fc-3837ee976ae5 Org ID: 61d09b13-129f-4353-9eb3-0e7ee89eb6b8 https://cdn.discordapp.com/attachments/1438160017034448906/1438160018884268146/image.png?ex=6915de64&is=69148ce4&hm=c8e6ad2808fd178839cc7ac8bdeb49deb7372c2479bcab005afdd4477c0704c8& https://cdn.discordapp.com/attachments/1438160017034448906/1438160020008075425/image.png?ex=6915de64&is=69148ce4&hm=1f3e6ed0c7727c8968c8ce3acdd2103bf1fff5d78e180b6d2fa7ef34efbb4633&

Could someone tell me if the assistant can recognize when the user types on the phone’s virtual keyboard during a call?