TwiML → VobizXML: verb-by-verb

If your Twilio app answers calls by returning TwiML, migrating to Vobiz is mostly a verb-for-verb rename. The document shape is identical: a root element wrapping ordered voice verbs. Twilio’s <Response> becomes Vobiz’s <Response>, <Say> becomes <Speak>, and the builder - twilio.twiml.voice_response.VoiceResponse() - becomes vobizxml.ResponseElement(). This page maps each voice verb and its attributes, then shows before/after IVRs you can copy.

The builder swap

Twilio’s helper libraries expose a VoiceResponse builder; Vobiz ships the equivalent ResponseElement in the bundled vobizxml module. Nesting, chaining, and serialization all mirror what you already do.

from twilio.twiml.voice_response import VoiceResponse

response = VoiceResponse()
response.say('Hello from Twilio.')
print(str(response))

from vobiz import vobizxml

response = vobizxml.ResponseElement()
response.add_speak('Hello from Vobiz.')
print(response.to_string())

const { twiml } = require('twilio');

const response = new twiml.VoiceResponse();
response.say('Hello from Twilio.');
console.log(response.toString());

import { vobizxml } from '@vobiz/sdk';

const response = new vobizxml.ResponseElement();
response.addSpeak('Hello from Vobiz.');
console.log(response.toString());

Verb-by-verb mapping

TwiML verb	Builder (Twilio)	VobizXML verb	Builder (Vobiz)	Notes
`<Say>`	`response.say(text, voice=, language=, loop=)`	`<Speak>`	`add_speak(text, voice=, language=, loop=)`	Voice names `man`/`woman` map to Vobiz `MAN`/`WOMAN`; `language` and `loop` carry over 1:1.
`<Play>`	`response.play(url, loop=)`	`<Play>`	`add_play(url, loop=)`	Plays MP3/WAV from a URL; `loop="0"` repeats until the call moves on.
`<Gather>`	`response.gather(input=, timeout=, num_digits=, finish_on_key=, speech_timeout=, action=)`	`<Gather>`	`add_gather(input_type=, execution_timeout=, num_digits=, finish_on_key=, speech_end_timeout=, action=)`	`input`→`inputType`, `timeout`→`executionTimeout`, `speechTimeout`→`speechEndTimeout`; `numDigits`/`finishOnKey` keep their names. Vobiz adds `digitEndTimeout` for inter-digit timing.
`<Dial>`	`response.dial(caller_id=, timeout=, time_limit=, hangup_on_star=, action=)`	`<Dial>`	`add_dial(caller_id=, timeout=, time_limit=, hangup_on_star=, action=)`	`callerId`, `timeout`, `timeLimit`, `hangupOnStar`, `action` all carry over by name.
`<Number>` (in `<Dial>`)	`dial.number('+1…')`	`<Number>`	`add_number('+1…')`	PSTN destination nested in `Dial`.
`<Client>` / `<Sip>` (in `<Dial>`)	`dial.client('agent')` / `dial.sip('sip:…')`	`<User>`	`add_user('sip:…')`	Reach a SIP endpoint or softphone; supports `sendDigits` and `sipHeaders`.
`<Record>`	`response.record(timeout=, finish_on_key=, max_length=, play_beep=, action=)`	`<Record>`	`add_record(timeout=, finish_on_key=, max_length=, play_beep=, file_format=, action=)`	`timeout` is the silence limit on both; Vobiz adds `fileFormat` (mp3/wav) and `startOnDialAnswer`.
`<Conference>` (in `<Dial>`)	`dial.conference('Room 1234')`	`<Conference>`	`add_conference('Room 1234')`	Named room by text content. On Vobiz `Conference` is a top-level verb - return it directly from `Response`.
`<Pause>`	`response.pause(length=)`	`<Wait>`	`add_wait(length=)`	Silent pause; Vobiz `Wait` adds optional `silence`/`minSilence` for beep detection.
`<Redirect>`	`response.redirect(url, method=)`	`<Redirect>`	`add_redirect(url, method=)`	Hand control to another answer URL that returns fresh XML.
`<Hangup>`	`response.hangup()`	`<Hangup>`	`add_hangup()`	Ends the call.
`<Stream>`	`response.start().stream(url=)` / `response.connect().stream(url=)`	`<Stream>`	`add_stream(url, audio_track=, bidirectional=)`	Fork call audio to a `wss://` server; `track`→`audioTrack`, and `bidirectional="true"` gives two-way media.

Vobiz also exposes add_dtmf (send digits) and add_preanswer (run verbs during early media) on the same ResponseElement, plus a first-class Gather with combined dtmf speech input.

A one-key menu. input/timeout/numDigits become inputType/executionTimeout/numDigits, and nested say becomes add_speak.

from twilio.twiml.voice_response import VoiceResponse

response = VoiceResponse()
gather = response.gather(
    input='dtmf',
    num_digits=1,
    timeout=10,
    action='/menu-choice',
    method='POST',
)
gather.say('Press 1 for sales, 2 for support, or 0 for an operator.')
response.say("We didn't receive your input. Goodbye.")
response.hangup()
print(str(response))

from vobiz import vobizxml

response = vobizxml.ResponseElement()
gather = response.add_gather(
    input_type='dtmf',
    num_digits=1,
    execution_timeout=10,
    action='https://yourapp.com/menu-choice',
    method='POST',
)
gather.add_speak('Press 1 for sales, 2 for support, or 0 for an operator.')
response.add_speak("We didn't receive your input. Goodbye.")
response.add_hangup()
print(response.to_string())

import { vobizxml } from '@vobiz/sdk';

const response = new vobizxml.ResponseElement();
const gather = response.addGather({
  inputType: 'dtmf',
  numDigits: 1,
  executionTimeout: 10,
  action: 'https://yourapp.com/menu-choice',
  method: 'POST',
});
gather.addSpeak('Press 1 for sales, 2 for support, or 0 for an operator.');
response.addSpeak("We didn't receive your input. Goodbye.");
response.addHangup();
console.log(response.toString());

The XML Vobiz emits is verb-for-verb recognizable:

<?xml version="1.0" encoding="UTF-8"?>
<Response>
    <Gather action="https://yourapp.com/menu-choice" method="POST" inputType="dtmf" numDigits="1" executionTimeout="10">
        <Speak>Press 1 for sales, 2 for support, or 0 for an operator.</Speak>
    </Gather>
    <Speak>We didn't receive your input. Goodbye.</Speak>
    <Hangup/>
</Response>

Before / after: forward a call with caller ID

Twilio’s <Dial> + <Number> maps straight to Vobiz’s <Dial> + <Number>; a <Client>/<Sip> target becomes <User>. callerId and timeout keep their names.

from twilio.twiml.voice_response import VoiceResponse

response = VoiceResponse()
response.say('Please hold while we connect you.')
dial = response.dial(caller_id='+14155551234', timeout=20, action='/dial-status')
dial.number('+14165553434')
print(str(response))

from vobiz import vobizxml

response = vobizxml.ResponseElement()
response.add_speak('Please hold while we connect you.')
dial = response.add_dial(caller_id='+14155551234', timeout=20,
                         action='https://yourapp.com/dial-status')
dial.add_number('+14165553434')
print(response.to_string())

import { vobizxml } from '@vobiz/sdk';

const response = new vobizxml.ResponseElement();
response.addSpeak('Please hold while we connect you.');
const dial = response.addDial({ callerId: '+14155551234', timeout: 20, action: 'https://yourapp.com/dial-status' });
dial.addNumber('+14165553434');
console.log(response.toString());

To reach a softphone or SIP endpoint (Twilio <Client>/<Sip>), nest a <User> instead of <Number>:

Vobiz · Python

dial = response.add_dial(caller_id='+14155551234')
dial.add_user('sip:john1234@registrar.vobiz.ai')

Before / after: fork audio to a WebSocket

Twilio’s start().stream() (one-way) and connect().stream() (two-way) both map to Vobiz’s <Stream> - track becomes audioTrack, and bidirectional="true" requests two-way media.

from twilio.twiml.voice_response import VoiceResponse

response = VoiceResponse()
connect = response.connect()
connect.stream(url='wss://example.com/audio')
print(str(response))

from vobiz import vobizxml

response = vobizxml.ResponseElement()
response.add_stream(
    'wss://example.com/audio',
    audio_track='both',
    bidirectional=True,
)
print(response.to_string())

Key differences

Same document, renamed verbs. The root stays <Response>; only leaf names change (Say→Speak, Pause→Wait, Client/Sip→User). Nesting, ordering, and the “return XML from your answer URL” model are identical.
Gather gives you two silence timers. Twilio’s single timeout becomes Vobiz’s executionTimeout, and speechTimeout becomes speechEndTimeout. Vobiz adds a dedicated digitEndTimeout for inter-digit pacing, so DTMF and speech end conditions are tuned independently. Full attribute list on the Gather reference.
One Gather handles digits and speech together. Set inputType="dtmf speech" and whichever the caller does first is posted to your action URL - one verb for menus and open-ended intent capture.
User covers both Client and Sip. A single <User> element dials SIP endpoints and softphones, carries sendDigits for extensions, and sets custom X-VH- SIP headers via sipHeaders.
Conference is a top-level verb. Where Twilio nests <Conference> inside <Dial>, Vobiz returns <Conference> directly from <Response>; the room is created on first join and named by the element’s text content.
Transfers are declarative. Instead of an imperative “modify live call” step, return a <Redirect> (or new <Dial>) from your answer URL and Vobiz continues the call with the fresh XML - the same webhook-driven pattern your TwiML app already uses.
Streaming is built in. <Stream> forks audio over wss:// with audioTrack selection and bidirectional two-way media for AI voice-agent pipelines. See the Stream reference for codecs and reconnect options.

Twilio → Vobiz overview - the full migration order and at-a-glance matrix.
Gather XML reference - every input attribute and webhook parameter.

AI Voice Platforms

Official SDKs

Browser & WebRTC

Business Phone Systems

Migrations

TwiML → VobizXML: verb-by-verb

The builder swap

Verb-by-verb mapping

Before / after: a DTMF IVR menu

Before / after: forward a call with caller ID

Before / after: fork audio to a WebSocket

Key differences

​The builder swap

​Verb-by-verb mapping

​Before / after: a DTMF IVR menu

​Before / after: forward a call with caller ID

​Before / after: fork audio to a WebSocket

​Key differences

​Related

The builder swap

Verb-by-verb mapping

Before / after: a DTMF IVR menu

Before / after: forward a call with caller ID

Before / after: fork audio to a WebSocket

Key differences

Related