Hello team, I want to build a call center from scratch using VAPI. I don’t have much software knowledge. I’ve watched many articles and videos, but I haven’t been successful. Where can I get support on this topic?

I'm building my own custom DTMF tool which is going to be using typescript and it takes in multiple arguments. I'd like to know how the original is built, so I can build upon it.

Hi Vapi team, I am facing an issue connecting Google Calendar to my Vapi account. Steps to reproduce: -Go to More → Integrations → Tools Provider → Google Calendar -Click Connect -Select my Google account and allow all requested permissions -Google shows the flow as finished and redirects back to Vapi, but the button still shows “Connect” and the integration is not saved Details: -Assistant ID: d78737b7-1c9d-4ef3-a31a-eda8b32ad1fd -Google account email used: yapdiyan@gmail.com -Tried multiple times, different browsers, and incognito; issue persists -Screenshot attached (shows the Google Calendar integration page with the Connect button) I saw that other users are reporting similar Google Calendar integration issues (OAuth finishes but connection does not persist). Could someone from the team please help check my account on the backend or share a current workaround so I can use calendar-based scheduling for my client? #1211483291191083018 https://cdn.discordapp.com/attachments/1449248676806660146/1449248677184016384/WhatsApp_Image_2025-12-13_at_09.18.06_5abf9755.jpg?ex=693e3582&is=693ce402&hm=6387057b24187e241c9bbf48b65f19e0acbd0607fa7f881b8252985859297daf&

Hi, I have connected Make at the org level but I do not see Make as an option inside Tool nodes in agent flows. Can you enable Tool-based integrations (Make) for my org?

https://dashboard.vapi.ai/calls/019b143f-24d4-700e-8fc2-133914c5aed1 I’m trying to make the voice agent more conversational so that the user can interrupt the agent at any time. However, this doesn’t seem to work, the agent only recognizes user speech after it finishes speaking. I want the user to be able to speak and interrupt the agent at any moment. I’ve tried enabling the Stop Speaking Plan in the advanced settings, but it doesn’t seem to have any effect. Regardless of the configuration, the agent still ignores the user while it’s talking. Has anyone else encountered this issue or found a workaround? https://cdn.discordapp.com/attachments/1449150460291448852/1449150460731592715/image.png?ex=693dda09&is=693c8889&hm=e6f95972903ae3e34c96eaa6c8191cbe891b0b0a457fe7f3c1b57e2e7d0a26c0&

We are seeing crazy high charges for chat messsages like 5 cents just for a single input and output. I see the token is high is this because of something we have configured incorrectly? @Vapi Support Bot

the transcriber is really struggling on emails. i want to inject emails into keywords when an api call returns, to better recognize the user's email. is that possible today? can it be?

Hello! I currently have 27 concurrency lines and no calls being made. When I try to make a new call, it says my concurrency limit has been hit when I know this is not possible. Would you please be able to reset my concurrency lines?

Hi Vapi Team, We are currently using your platform to design and operate two types of assistants: Inbound and Outbound. Our goal is to clarify the best architectural approach to manage concurrency limits, ensuring that inbound calls are always available, even while large outbound campaigns are running. Context Inbound calls must be available at all times (we are using Redis to manage a queue on our side). Outbound campaigns may involve a high volume of calls (e.g. 1,000–2,000 calls). The main concern is that outbound campaigns can fully consume the available concurrency, making inbound lines temporarily unavailable. Increasing the total number of concurrency lines alone does not fully solve the problem. For example, even with 1,000 concurrency lines, a sufficiently large outbound campaign could still saturate all available channels and block inbound calls. Proposed Approach Our current idea is to logically separate inbound and outbound traffic by using two different Vapi organizations: Inbound Org Dedicated to inbound assistants Allocated with X concurrency lines Ensures inbound availability at all times Outbound Org Dedicated exclusively to outbound campaigns Allocated with Y concurrency lines Used only for outbound execution This way, outbound campaigns would never compromise inbound availability, as each flow would have its own isolated concurrency pool. Questions If we create a new organization using “Switch organization → New organization” within our existing account: Will this new org automatically receive its own set of concurrency lines (e.g. +10)? Or does it share the same concurrency pool as the parent organization? If concurrency is shared, should the outbound org be created entirely outside the current organization (with a separate account / email) to guarantee isolation? Is there any native or technical mechanism in Vapi to: Reserve or cap concurrency for outbound campaigns (e.g. “use only X lines for outbound”)

Hi, I'm struggling with this integration. I've created the Azure speech resource, and have my keys. In the VAPI integrations page I try selecting region UK and add my key, but it keeps getting rejected with error because VAPI is trying https://uk.tts.speech.microsoft.com whereas apparently it should be https://uksouth.tts.speech.microsoft.com Can anyone advise?

We have a workflow for appointment bookings that is close to getting to production. The issue I am facing is that the API request nodes seem to timeout after 5 seconds, even if I set the timeoutSeconds to a bigger value. Please fix this as soon as possible, or let me know what I need to change, since we are on a tight timeline.

@Vapi @Vapi Support Bot I’m working with a tool call named

checkAvailability

that successfully returns available slots as an array of objects. What I’d like to do is store this tool response in a variable within the dashboard, so that I can later reference it in my flow using

{{availableSlots}}

. Could you please guide me on: - How to capture the tool’s response and assign it to a variable in the dashboard?

Hello, Please add an API endpoint that allows purchasing additional concurrency lines for voice agents on demand. This is required to scale traffic automatically without manual actions. Thank you.

Hi team, I’m building a Voice Agent on the Vapi.ai platform and ran into an issue with the Tool (Query). The agent successfully calls the tool, but the response says “No document found for field.” I’ve already checked both Files and the Knowledge Base, and the IDs are correct. Could you help me figure out how to fix this? (By the way The content in My Files and the Knowledge Base is in Chinese. ) https://cdn.discordapp.com/attachments/1448967954497536124/1448967955072290951/123.png?ex=693d3011&is=693bde91&hm=b5e25056378cf3bbe50f5ee1013d666cfd97ace74dc7b05e098b487001a973ac& https://cdn.discordapp.com/attachments/1448967954497536124/1448967955890176000/Query_Chatbot_ploblem.png?ex=693d3011&is=693bde91&hm=e23b4887a0a03c1fc73dad235a9e86c77f75d73aba24c096a4240fdbc80de1b2&

Hi team, seems like 'model-output' webhook message is not working for GPT realtime model, it doesn't output the model response (except tool calls for some reason), please help to check on this

Hi team, so I wanted to use model-output webhook server message to do some background processing, so I implemented a buffer and everything, but when I try it out it doesn't give me the correct sentence, it will be scrambled like this:

glad screen now looking really that hear.I'm I'm to forward you about how has over changed the. image an have I I years to here with — like'd should you share you chatting What on your it. notice you to this image do in? dancing that with see

can I get the sequence index somehow?

Hi Support, This is regarding the cost of calls. I realize the total cost is more than breakdown added up. In Vapi console under call logs, the total cost of my call shows $0.13. However the "total cost" shown under the end of call report which i configured a webhook to extract to googlesheet < 1. message: costBreakdown: Total> shows total cost as $0.1818. There is an additional $0.0518 charged which i can't find the cause, all the call charges are not tallying. I've also checked the $0.13 breakdown and it already consist of the VAPI platform fee, Please advice what's causing the additional cost. you can use this call ID as reference : 019b1076-f5cc-7221-84b3-ff669a5ba6ee https://cdn.discordapp.com/attachments/1448899446547943595/1448899446824898623/Image_12-12-25_at_12.45_PM.png?ex=693cf043&is=693b9ec3&hm=d7f0dea37271d15b8d89312d9c60cbdef43903f87835732b3babedc600f222f9& https://cdn.discordapp.com/attachments/1448899446547943595/1448899447424815104/Image_12-12-25_at_12.47_PM.png?ex=693cf043&is=693b9ec3&hm=29b42f20232a398129786c40321ea0da63ecf7ea1db3e76b3829b231cd38e9e7&

I'm using this with my custom LLM. So far, the transfer and all happens well. However, if the number to which the conversation is being transferred declines the call, or doesn't pick up, the fallback does not seem to work. This is what my config looks like when sent to the conversation's control URL { "type": "transfer", "destination": { "type": "number", "number": phone_number, "callerId": '{{phoneNumber.number}}', "transferPlan": { "mode": "warm-transfer-experimental", "holdAudioUrl": "https://file.mp3", "summaryPlan": { "enabled": True, "messages": [ { "role": "system", "content": "Please provide a summary of the call." }, { "role": "user", "content": "Here is the transcript:\n\n{{transcript}}\n\n" } ] }, "fallbackPlan": { "message": "It seems the person is currently busy or on another call. Can I take a message?", "endCallEnabled": False } } } } CC: @Prashant Abbi

"This is not something you can fix from your side — it's a backend issue that Vapi needs to resolve. " Please help. https://cdn.discordapp.com/attachments/1448791538854395954/1448791539584073791/Screenshot_1.png?ex=693c8bc4&is=693b3a44&hm=051127de0aee62bc5b0df1cf0ec52f35054b0630076da58cc7286734fea8a2b0&

I’m experiencing an issue with the Query Tool in Vapi AI. Previously, I was able to upload a .txt file as a knowledgebase, assign it to an AI assistant, and the assistant could read and use it without problems. However, recently this no longer works—the AI assistant cannot access or read the uploaded knowledgebase. Could you please advise on what might be causing this and how to resolve it? For reference: File type: .txt Steps I followed: i uploaded the text file to vapi ai, created a query tool, described it, assigned the knowledgebase (text file), i assign it to my assistant, when i request to use it, the tool fails and the ai cannot read and he gives me this notification when he tried to access the tool "Cannot read properties of undefined (reading 'message" The json code for example looks like this: { "id": "26e3fd95-ae66-4f66-ba3d-8d563f6a0b18", "createdAt": "2025-12-11T15:49:55.275Z", "updatedAt": "2025-12-11T15:51:24.875Z", "type": "query", "function": { "name": "query_tool", "description": "this is a tool used to answer FAQs", "parameters": { "type": "object", "properties": {}, "required": [] } }, "messages": [ { "type": "request-start", "blocking": false } ], "orgId": "c4bad15c-08a9-4947-97b2-267c898b3629", "knowledgeBases": [ { "name": "new_knowledge_base", "provider": "google", "model": "gemini-2.0-flash", "description": "this is a tool used to answer FAQs", "fileIds": [ "ba9e64e9-ce8b-4d57-9b3d-8028d86fd02f" ] } ] } The issue that i think is the blocking.

Call: 019b0edc-fea5-7ff2-a144-23bde81ed2db Model: GPT Realtime Tool call finishes, but assistant only says: "I'm checking on that for you right now. Let's wait just a moment for the response." Very frustrating.

{
  "id": "log-28",
  "level": 30,
  "time": 1765480937089,
  "timestamp": 1765480937089000000,
  "severityText": "INFO",
  "severityNumber": 9,
  "body": "Tool calls finished",
  "attributes": {
    "category": "tool",
    "callId": "019b0edc-fea5-7ff2-a144-23bde81ed2db",
    "orgId": "9fc10dfb-0f59-4ced-8157-345511dd806a",
    "runDurationMs": 5272,
    "totalResults": 1,
    "errorsCount": 0,
    "results": [
      {
        "name": "get_response_2",
        "toolCallId": "call_Dujyv9wGmxRLsp1U",
        "type": "function",
        "serverUrl": "https://trustful-uncherishing-braiden.ngrok-free.dev",
        "result": "Our dentures are made of high quality materials.",
        "metadata": {}
      }
    ]
  }
}

After muliple uploads of files, our files are disappearing from the dashboards like it has been deleted https://cdn.discordapp.com/attachments/1448763243655987338/1448763243869769728/Screenshot_2025-12-11_at_11.46.37_PM.png?ex=693c716a&is=693b1fea&hm=83770a377c29b82e1d3bc734bbbd7065232455efaff542a2db7176edf7744029&

Tried with multiple webhooks to different phone numbers. I'm using vonage for telephony but I dont see any call logs there. I've checked over all my logs im pretty sure the problem is with vapi. https://cdn.discordapp.com/attachments/1448748791745351690/1448748793091588127/image.png?ex=693c63f4&is=693b1274&hm=7574a5fe76000121b1300163b734532a5658c7476db1b46d2cbdcb82358c0fa9&

I try to transfer calling to a free vapi number to my number in Colombia, is it possible?

I’m building an AI automation agency and will be creating multiple voice agents for different clients, as well as internal agents for my own company. What are you looking to build with Vapi? I want to create a scalable SaaS-style system where I can build and manage multiple agents for my agency and for each client—ideally all under one main account, with separate usage tracking per client. Why would you like to speak with the sales team? I need guidance on how to set up a SaaS model using Vapi: • How to manage multiple client agents under one parent account • How usage can be attributed and billed per client (client pays for their own usage) • How to structure this for monthly recurring revenue + usage billing • Best practices and recommended setup for agencies offering AI agents as a service

To configure your Query Tool for retrieving TXT file content, ensure your assistant's system prompt explicitly instructs when to use the tool. For example:

json
{
  "role": "system",
  "content": "You are a helpful assistant. When users ask about products or need information, use the 'knowledge-search' tool to search our knowledge base for accurate answers. Always call the knowledge-search tool before responding."
}

This helps the assistant properly access your TXT-based knowledge base and prevents retrieval errors. For more details, see the [Vapi Query Tool documentation](https://docs.vapi.ai/knowledge-base/using-query-tool). Source: - [Using the Query Tool for Knowledge Bases](https://docs.vapi.ai/knowledge-base/using-query-tool)

When I go through the process of the google calendar integration, everything goes through fine and I get the "success" message at the end, but then in Vapi the connect button does not change to "Connected" and when the agent tries to access the calendar, I get the Missing Nango Config error.

I'm having an issue with query tools. They can't retreive data from txt files

Hi, I just wanted to ask if I was on a Vapi Pay as You Go plan with a credit card on file and credits on the account, am I still limited to only 10 calls/day on the free Vapi phone number or am I free to call more?