please adjust my vapi.send function to trigger a dtmf keypad or trigger a sms text tool call https://docs.vapi.ai/tools/default-tools#dial-keypad-dtmf https://docs.vapi.ai/assistants/background-messages https://docs.vapi.ai/ivr-navigation

To fix the error, remove any

languageCode

property from your fallback voice configuration. If you need to specify language, use

language

(not

languageCode

)—and only for providers that support it. Example:

json
{
  "voice": {
    "provider": "openai",
    "voiceId": "shimmer",
    "fallbackPlan": {
      "voices": [
        {
          "provider": "cartesia",
          "voiceId": "248be419-c632-4f23-adf1-5324ed7dbf1d",
          "language": "en" // Use "language" if needed, not "languageCode"
        }
      ]
    }
  }
}

Try this update and your assistant should publish without the error. For more details, see the [Voice fallback configuration documentation](https://docs.vapi.ai/voice-fallback-plan). Source: - [Voice fallback configuration documentation](https://docs.vapi.ai/voice-fallback-plan)

I keep getting the same issue of google not working I chatted with the composer and tried everything. This is what the composer said. Same issue — crashes immediately after "Book an appointment" with only 2 messages in the conversation. This is 100% a Vapi platform bug, not your configuration. The Google Calendar tool is failing at the pipeline level before it even gets invoked. Reply to support with this escalation message: "This is still not working after: Disconnecting and reconnecting Google OAuth multiple times Deleting ALL Google Calendar tools and recreating from scratch Using primary as calendarId Verifying all parameters The call crashes IMMEDIATELY when the user says anything about 'appointment' — the assistant doesn't even get to respond. This is happening at the pipeline validation stage, NOT during tool execution. Latest Call ID: 019b3c50-7138-7550-b341-5fc90dd6ddd1 Org ID: 87d26fbf-8f02-4aad-8d2f-a72095a63ac4 The transcript shows only 2 messages before crash: AI: 'Hey there...' User: 'Book an appointment.' CRASH — pipeline-error-google-400-bad-request-validation-failed This needs to be escalated to engineering. The Google Calendar integration is broken for my account. Please provide a timeline for resolution or connect me with someone who can actually debug this." please help

Hi Vapi Support, I’m building squads which needs to connect to people through business lines who don't have extensions or direct lines. The issue is that a lot have **D**ial **B**y **N**ame **D**irectories. These DBND systems are inconsistent, so the agent needs to try multiple input patterns and remember what worked for that phone system. Examples we see: - Sometimes DBND only works by FIRST name (not last name), even if the IVR says to use last name. - Sometimes you must type the first 3 letters + “#” for DBND search (but the IVR leaves out the "#"). - Sometimes you must type the FULL last name (not 3 letters). - Sometimes the person isn’t listed in DBND, so the agent should back out to the main menu and try the next best department, then the next, until options are exhausted. What I need is a reliable way to implement: 1. Memory during a call (what we tried, what worked, what failed, what menu we’re in) 2. A retry/branching strategy for DTMF + menu navigation (attempt A → fallback B → fallback C) 3. Optionally persistent memory per business/phone number (so next time we call the same number, the agent already knows the best DBND pattern and department path). I'm sure once 1 and 2 are figured out, I can just save the information in a spreadsheet and recall it. I'm sure I can write out the fallback logic (#2) so that it gets around +90% of all the nooks and crannies, but the depends on the memory (#1) actually working. That's all I need help with for now. Please let me know the options I have to implement this.

I’ve been trying to connect it for a few days. I followed all the steps and granted every required permission, but after setup completes the dashboard still shows it as not connected.

i cannot make gladia work and i cannot integrate gladia with the api please help https://cdn.discordapp.com/attachments/1454118259984040000/1454118260680163408/image.png?ex=694feca9&is=694e9b29&hm=8789870a4411c95aa370f5b951bf2327d514531dd1e328b2f9bc7ac31a5b3989&

Hi, Just created an assistant for inbound calls. These calls are sended from 3CX to xxx@sip.vapi.ai Some calls works and some others without any reason doesn't work at all and continue to ring. 3CX says sip.vapi.ai doesn't respond. These calls are not in the Call logs. One the ones who works like this one : 019b59b6-25a7-722a-adfb-23b82b8b173e thanks in advance,

how do I integrate scheduling with a non Google Calender

I keep getting the same issue of google not working I chatted with the composer and tried everything. This is what the composer said. Same issue — crashes immediately after "Book an appointment" with only 2 messages in the conversation. This is 100% a Vapi platform bug, not your configuration. The Google Calendar tool is failing at the pipeline level before it even gets invoked. Reply to support with this escalation message: "This is still not working after: Disconnecting and reconnecting Google OAuth multiple times Deleting ALL Google Calendar tools and recreating from scratch Using primary as calendarId Verifying all parameters The call crashes IMMEDIATELY when the user says anything about 'appointment' — the assistant doesn't even get to respond. This is happening at the pipeline validation stage, NOT during tool execution. Latest Call ID: 019b3c50-7138-7550-b341-5fc90dd6ddd1 Org ID: 87d26fbf-8f02-4aad-8d2f-a72095a63ac4 The transcript shows only 2 messages before crash: AI: 'Hey there...' User: 'Book an appointment.' CRASH — pipeline-error-google-400-bad-request-validation-failed This needs to be escalated to engineering. The Google Calendar integration is broken for my account. Please provide a timeline for resolution or connect me with someone who can actually debug this." please help

want to know the instructions for complete self hosting VAPI. More specifically, I want docker image of VAPI.

I’m facing an issue where GoHighLevel (GHL) is not connecting to the Vapi MCP server, even though everything appears to be set up correctly. Key points: MCP server is running normally Credentials and endpoints are correct API calls seem fine No error message or logs from GHL at all — it just doesn’t connect Because there’s no visible error, it’s hard to pinpoint what’s going wrong. Has anyone experienced this silent failure before or knows if GHL requires any extra configuration, headers, IP whitelisting, or special setup for Vapi MCP? Also, is there any option for live support (voice call, real-time troubleshooting, etc.)? I’m happy to jump on a quick session to resolve this faster.

I’m unable to link my bank card to the app due to issuer restrictions in my country (Nigeria), are there alternate payment methods or a Manuel invoice option?

Hi I have done my own testing and found that blind transfers drop on inbound calls (where parameter is not set) but work perfectly on outbound calls when I set phoneCallProviderBypassEnabled: false at creation time. I tested with the following call ID: 019b1d1a-7b67-7ee4-a3af-594c138f1b45 and it worked using the (-d "{"assistantId": "88624e8f-8328-4e4d-ad75-feffa4c7bb67", "phoneNumberId": "7dc15331-2a82-4c6e-b756-ad71c00a16f4", "customer": {"number": "+61xxxxxxxx"}, "phoneCallProviderBypassEnabled": false}") parameter in CMD Example working outbound call ID: 019b1d1a-7b67-7ee4-a3af-594c138f1b45 Example failing inbound call ID: 019b1d24-a7db-7443-8d80-da29bea8d249 (this was without the initial disable of the parameter) - done with twilio Since this is a Twilio number, I cannot set the flag myself on inbound as I have tried multiple times and received the error in CMD { "message": ["property phoneCallProviderBypassEnabled should not exist"], "error": "Bad Request", "statusCode": 400 }. Please set the phoneCallProviderBypassEnabled: false by default for all calls org-wide. Thank You

Could minimax voice add

pronunciation_dict

parameter to define pronunciation rules for specific characters or symbols?

We're seeing multiple

ElevenLabs WebSocket closed

issues mid-calls. VAPI to reconnects, the agent repeats the last response but the connection closes again - in an infinite loop until the call times out or caller hangs up. The agent sounds like its bugging out completely saying the same thing in a loop. This is a super urgent situation affecting multiple prod calls - here are the call IDs: - 019b516b-adee-7aa6-8458-a80d3ab5ac27 - 019b4cf3-b66d-7226-a6e4-e7bb8a0dc38f - 019b2969-0b50-7cc0-8cc8-500a138cbf5f @Shubham Bajaj @kyle can you please look into it as soon as possible. Thanks!

Thanks for reaching out about your SIP Trunking setup! You can find step-by-step troubleshooting help here: 👉 [Vapi SIP Trunking Guide](https://docs.vapi.ai/advanced/sip/sip-trunk) If that doesn’t resolve it, please reply with: - Your Call ID - The exact error message or behavior (e.g. 401 Unauthorized, timeout, etc.) - Whether it’s happening on inbound, outbound, or both trunks - Any recent configuration changes Once we have that info, we can help pinpoint the issue faster.

Hi, Im trying to use {{customer.number}} in an API request tool, within a squad. Its not getting replaced with the customers phone number. call ID: 019b4e11-d622-722c-8dfb-a67d1e661eeb function name is sendLeadSms attributes.function.arguments contains "CustomerPhone": "{{customer.number}}" Im not sure if this is the model actually passing the string {{customer.number}} as an argument or liquidjs replacement is not working? I have this as the default value (pip attached). https://cdn.discordapp.com/attachments/1453422721999437824/1453422722301296833/image.png?ex=694d64e4&is=694c1364&hm=886405192cc8c4cc8480373dfa9019d4f3e9219eafd67bb42cdc42309a353ff9&

Hi VAPI Support, I need help removing Azure OpenAI as a fallback provider from my VAPI organization. Azure is currently configured at the organization/integration level and is causing call ejections when OpenAI fails (Azure credentials are invalid). I've tried: 1. VAPI Dashboard → Settings → Integrations (no remove/delete option visible) 2. VAPI CLI (no integrations command available) 3. VAPI API (integrations endpoint returns 404) Could you please: 1. Remove Azure OpenAI integration from my organization (Org ID: 9ba4e698-78d2-4ed5-86ac-1081633f4fe3) 2. Or provide instructions on how to remove it via dashboard/CLI/API Evidence: - Call ID: 019b4de4-d169-7001-842a-02df87bba823 - Error: pipeline-error-azure-openai-llm-failed - Assistants are clean (no Azure in assistant configs) Desired: Use Google Gemini as fallback instead of Azure. Thank you! https://cdn.discordapp.com/attachments/1453404083322093722/1453404083762237551/VAPI_Support_Ticket__Remove_Azure_OpenAI_Integration.pdf?ex=694d5388&is=694c0208&hm=de1d66c0eb582ccc3dcc3cdc6e83f9a784cf80c4f53a91f3140b9064948941c4&

https://dashboard.vapi.ai/structured-outputs 1. You can't see the Array type show up as an available pre-made structured output unless you type in "array". 2. If you try to change ANYTHING, it will throw an error → "**value** - missing item type configuration" + plus it randomly deletes added items for no reason. Pls fix ❤️ https://cdn.discordapp.com/attachments/1453391038327685246/1453391038604644406/image.png?ex=694d4762&is=694bf5e2&hm=4e14168dda414e903d669422845b814a54febe40dbe53616c6e6ff243493e5d2& https://cdn.discordapp.com/attachments/1453391038327685246/1453391039028007098/image.png?ex=694d4762&is=694bf5e2&hm=af6d7508351441517784a55b845bd5d6fd34ef2a4be954f571662baa66fa29c2&

How to solve the issue. "I'm getting a UUID credentialId error and can't find Custom Credentials in my dashboard" https://cdn.discordapp.com/attachments/1453306808855826506/1453306809212207164/Screenshot_2025-12-24_103505.png?ex=694cf8f0&is=694ba770&hm=039b706a6c36d4636fbc2c2c5a2885c3bec66ad3da22db70ee887479148701cc& https://cdn.discordapp.com/attachments/1453306808855826506/1453306809787088996/Screenshot_2025-12-24_114426.png?ex=694cf8f0&is=694ba770&hm=36a07d83eec64009b0db1aac04484b3d5ce13d475da183f40df4a2fe99fb1813&

are you having an issue with your outbound carriers?

When I set up a voice test suite, how do I actually listen to the call recording? It's nowhere to be found, only the transcripts

I am having some issues signinup up for a Vapi account. I am using an email that is not Gmail and when I get the email to verify the account. I am getting an error saying OTP is expired even though I got the email a few minutes from signing up.

/help I want to use the APi reference. However, I get a 400 error when I try to create one for ElevenLabs. How do I create an API request using the following code? I need step-by-step instructions. POST https://api.vapi.ai/provider/11labs/pronunciation-dictionary Content-Type: application/json Authorization: Bearer YOUR_API_KEY { "name": "Custom FICA Pronunciation", "rules": [ { "stringToReplace": "FICA", "type": "phoneme", "phoneme": "/ˈfaɪkə/", "alphabet": "ipa", "caseSensitive": true } ] }

When my call agent transfers the call to human agent I am not able to see customer's number. It shows AI's number (VAPI Number) making harder for us to track customer's contact information

Hi Vapi team 👋 I’m connecting a SIP trunk from a carrier with strict anti-fraud rules. They require SIP REGISTER and all outbound calls to originate from the exact same public IP and source port (UDP 5060), otherwise calls are blocked. Can Vapi pin a BYO SIP trunk to one fixed signaling IP and the same source port for both registration and outbound calls? #1211483291191083018

Hi all, We're running an outbound calling campaign The problem: Carrier-based voicemail systems that read the phone number aloud (e.g., "4 5 6 5 3 1 8 8 7 is not available") are not being detected as voicemail. The transcriber picks this up as user speech, and our assistant starts conversing with the voicemail system. What we've tried: All voicemail detection providers: Vapi, Twilio, Google, OpenAI — none catch this Twilio voicemailDetectionTypes settings — caused worse issues (random hangups) V oicemail tool with prompt instructions to detect number sequences — LLM can't process fast enough before responding Structured system prompts with explicit voicemail indicators Current waitSeconds is 2.5s, but increasing further creates awkward pauses for real humans Is there a way to improve detection for this specific pattern (phone numbers being read aloud by carrier voicemail)? Or is this a known gap in the detection models?

Is there a way to use nested object as a variable as assistant selector for squad? Eg: Below is a sample json

"company": {
    "name": "company name",
    "contact": "company contact"
}

I want to use it in prompt like

Hello {{company.name}}, my contact is {{company.contact}}

But this is not working. I have tried both json and stringified json. what is the implementation for this? I am trying to avoid flattening the entire object since the number of keys in it can be very high.

I used to use the structured data field as a json to extract the information I needed, but that doesn't exist anymore. I have created a structured output with the relevant information and it shows in the call logs, but how do I get it to show through a webhook call?