I am trying to add an api call in the "Create Custom tools" and given the POST method api url and given all the information for the payload or body(which the api call takes as the input) and whenever i am trying to add the headers of this format: -H "Authorization: Bearer YOUR_ACCESS_TOKEN" \ -H "Content-Type: application/vnd.amadeus+json" \ -H "X-HTTP-Method-Override: GET" \ and saving it, it is showing that the changes are saved, but while creating the tool it is showing("The tool cannot be created") and i have tried to create the tool with out the headers, then the tool is getting created , so once the tool is created i tried to save the header, then it is showing that the "changes are saved", but when i go to Phone numbers (user interaction button) and come to the tools button, the headers are getting disaapeared like there is no headers saved. how to over come this problem?

When I try hitting get call endpoint through postman its working, when I try the same through my server, its throwing 520 error, did anyone faced this issue?

I would like to know how reliable is this if I plan to use this on my system prompt? The assistant should be able to say the date along with what day is it today correctly in a consistent manner. https://cdn.discordapp.com/attachments/1243025299361370176/1331328089984335912/image.png?ex=67913767&is=678fe5e7&hm=704c63373ba218be75f3a5292141b3d1aa45d5ef4d4bb1cfceff615eaabe55ea&

Did you try it?

The best way is to connect worldtimeapi.org api with the time zone through make. Using an http get request. And when ever youll ask the time from the assistant, it will run the tool and get the exact time in response.

How to restrict call access for particular phone number or assistnatid using API?

as a test-demo, i am trying to emulate the front page of the vapi.ai website: i want to press the mic button and have a conversation - in my own website. is this possible? if so, is there something in the docs or a tutorial to help with this? also, how do i get an API key? i asked the vapi AI helper, but it did not know!

i see in the docs (https://dashboard.vapi.ai/assistants/1d7df586-e231-4937-9aaa-274f26df1aa3): const apiKey = "51662f41-59c2-4d46-b173-80b3a5918959"; const assistantId = "1d7df586-e231-4937-9aaa-274f26df1aa3"; i presume these are dummy API keys - but are they in fact what we are supposed to use?

Hi, are there any plans to also incorporate Google STT and TTS in VAPI in the (near) future?

Hello Everyone. I am new to this channel. I know there is a way to up load documents to teach Vapi how to handle calls. I have needed to translate the transcription of the calls to feed it to Vapi. Is there a way to feel it .wav files for it to learn. It says it will accept it, however when trying to add it, it does not permit me to.

This step you mentioned will be in the prompt right? Also will the prompt be in two languages or just one language alone? For example, i need an AI voice agent who speaks english and Portuguese will the prompt go this way... You are an assistant that speaks English and Portuguese, Liten actively and detect the user language then. If its Portuguese, do i need to write some prompts and script in portuguese to make that work. Anyone that has an answer to my request kindly tag me. Appreciate it a lot

No, you don't need to write the prompt in multiple languages. Only English is enough

Alright thanks but i got it in that format right? Also does the transcriber and voice matters in scenario like this?

Yes, you have to use multi

HI, I am currently trying to work on a multi-lingual voice agent; when the user talks in French, it should respond in French, same for English. Using azure multi-lingual model. I madeit in the prompting also clear that it only should speak these two languages, but somehow it keeps talking backon occassion in Portugees or Japanese. Anyone who has also experienced this? And was able to fix it correctly with prompting

Anyone has an idea where to get a turkish number for voice agent?

Twilio.com

hi guys, so what's the best available english voice according to you? I find most of them not authentic enough

Hello, has anyone managed to use squads via websockets without relying on a phone number ? Everytime I try, I do not receive any sound and my audio isn't sent neither and the call gets closed down 5-10 seconds after I start the websocket connection

I've been using GHL interfacing with VAPI via Make for a while, testing with real clients, and I can confidently say that the latest 11Labs - Eleven Flash V2.5 - is a total game-changer. Previously, calls felt unnatural, with awkward interruptions, but I truly believe TODAY MARKS THE INFLECTION POINT I was waiting for. Honestly, my Voice Virtual Assistant now outperforms our real Appointment Setters. There's always room for improvement, but this result is nothing short of incredible. Give it a listen and judge for yourself! https://cdn.discordapp.com/attachments/1243025299361370176/1335347315707220049/1fae3f9e-64c4-4846-9143-5d04ba75953e-1738345273969-2677bf52-6ee3-4280-b8ca-2ba6560375c8-mono.wav?ex=679fd699&is=679e8519&hm=09d113b88c8492a2dc210cfd21fb8ac0a71e41fe40eddfd5ebdc12fc35f3ea62&

11Labs - Ellie Tender Young... listen in my post.

hello please i need your help hello, please i need your help fixing this. i can not add a play.ht custom voice. https://cdn.discordapp.com/attachments/1243025299361370176/1335777489913512067/Screenshot_2025-02-03_at_1.40.51_AM.png?ex=67a1673b&is=67a015bb&hm=62b5c58b567688fc53a87c95e7dc3221c2a488ed540d7f2328524afe46befe55&

Hi there, do you guys whether there is some kind of documentation regarding the video recording feature? If not, any additional info regarding this is appreciated. https://cdn.discordapp.com/attachments/1243025299361370176/1335890817390215219/image.png?ex=67a1d0c6&is=67a07f46&hm=7f06e71a6ac803caa4da5531e9fa860be44d8362369e74a13afc2edadc8987ed&

when i train hindi agent on my script that agent can use typical hindi insted of my script what to do ?

Hi, did you figure this out?

For some reason I am noticing none of the knowledge bases are connecting to the Voice Assistants. When clicking on knowledge base it is showing no files. Yet in files they are all there. Is anyone experiencing the same?

did you see the update going around about this

if not I can share

Please do

Is there a work around to resolve this?