Config

This verb is used to update speech settings, configure barge-in, and collect DTMF or speech input during an active call.

The config verb allows developers to change the default speech settings during a session or to collect speech or DTMF input in the background while other verbs run. This verb is non-blocking, so the specified settings are changed immediately and the application proceeds with the next verb.

{
  "verb": "config",
  "synthesizer": {
    "vendor": "microsoft",
    "language": "de-DE",
    "voice": "de-DE-KillianNeural"
  },
  "recognizer": {
    "vendor": "google",
    "language": "de-DE"
  },
  "bargeIn": {
    "enable": true,
    "sticky": true,
    "input": ["speech", "digits"],
    "actionHook": "/userInput",
    "partialResultHook": "/partialInput",
    "finishOnKey": "#",
    "numDigits": 5,
    "minDigits": 1,
    "maxDigits": 5,
    "interDigitTimeout": 3000,
    "dtmfBargein": true,
    "minBargeinWordCount": 2
  },
  "fillerNoise": {
    "enable": true,
    "url": "https://example.com/filler.wav",
    "startDelaySecs": 2
  },
  "vad": {
    "enable": true,
    "voiceMs": 250,
    "silenceMs": 500,
    "strategy": "adaptive",
    "mode": 3
  },
  "speechFallback": {
    "type": "dial",
    "reason": "recognizerFailure",
    "dial": {
      "number": "[phone]"
    },
    "refer": {
      "uri": "sip:[e-mail]"
    }
  },
  "actionHookDelayAction": {
    "enabled": true,
    "noResponseTimeout": 5000,
    "noResponseGiveUpTimeout": 15000,
    "retries": 2,
    "actions": [
      {
        "verb": "say",
        "text": "Waiting for response..."
      }
    ],
    "giveUpActions": [
      {
        "verb": "say",
        "text": "No response received. Moving on."
      }
    ]
  },
  "boostAudioSignal": "+3dB",
  "listen": {
    "startTimeout": 5000,
    "stopTimeout": 2000
  },
  "notifyEvents": true,
  "onHoldMusic": "https://example.com/hold.mp3",
  "referHook": "/sipRefer",
  "reset": ["recognizer", "synthesizer"],
  "record": {
    "action": "startCallRecording",
    "siprecServerURL": "sip:[e-mail]",
    "recordingID": "call12345",
    "headers": {
      "X-Custom-Header": "value"
    }
  },
  "sipRequestWithinDialogHook": "/sipRequest",
  "amd": true
}

Configuration

The following table lists the available parameters:

Parameter	Type	Description	Required
actionHookDelayAction.enabled	boolean	Enables or disables the delayed action hook behavior. When enabled, the system waits for the configured action hook to respond before executing any delay or give-up actions.	No
actionHookDelayAction.noResponseTimeout	number	The timeout in milliseconds to wait for a response from the action hook before executing the delay actions, such as prompting the user with a say verb.	No
actionHookDelayAction.noResponseGiveUpTimeout	number	The timeout in milliseconds to wait before executing the give-up actions if the action hook never responds.	No
actionHookDelayAction.retries	number	The number of retry attempts to call the action hook before giving up.	No
actionHookDelayAction.actions	array	An array of verbs to execute while waiting for the action hook response, such as a say verb to provide feedback to the user.	No
actionHookDelayAction.giveUpActions	array	An array of verbs to execute if the action hook never responds, for example, a say verb to inform the user and continue the application.	No
amd	boolean	Enables Answering Machine Detection (AMD) to distinguish whether the call is answered by a human or a machine.	No
bargeIn.enable	boolean	Enables background listening for speech or DTMF input while other verbs execute. If disabled, stops any background listening tasks currently running.	No
bargeIn.sticky	boolean	If both bargeIn.enable and bargeIn.sticky are true, another background gather automatically starts after detecting speech or DTMF input, allowing continuous input collection.	No
bargeIn.actionHook	string	A webhook URL to invoke when user input is collected by the background gather. The default is voice.	No
bargeIn.partialResultHook	string	A webhook URL to receive interim transcription results during background gathering. Useful for providing real-time feedback or logging partial input.	No
bargeIn.input	array	Specifies allowed input types: ['digits'], ['speech'], or ['digits', 'speech'].	Yes
bargeIn.finishOnKey	string	The DTMF key that signals the end of input in the background gather.	No
bargeIn.numDigits	number	The exact number of DTMF digits expected to gather.	No
bargeIn.minDigits	number	The minimum number of DTMF digits expected to gather. The default is 1.	No
bargeIn.maxDigits	number	The maximum number of DTMF digits expected to gather.	No
bargeIn.interDigitTimeout	number	The time in milliseconds to wait between DTMF digits after reaching the minimum number of digits.	No
bargeIn.dtmfBargein	boolean	Enables DTMF barge-in so that entering a DTMF tone can interrupt audio playback during background gathering.	No
bargeIn.minBargeinWordCount	number	The minimum number of words the user must speak before triggering barge-in. Helps prevent accidental interruptions during speech prompts.	No
boostAudioSignal	string | number	Specifies the number of decibels to increase or decrease the outgoing audio signal level (for example, -6 dB or +3 dB). Default is 0 dB.	No
fillerNoise.enable	boolean	Enables or disables filler noise played while waiting for user input or during processing.	Yes (if used)
fillerNoise.url	string	The URL to the MP3 or WAV audio file to play as filler noise.	No
fillerNoise.startDelaySecs	number	The delay in seconds before starting filler noise playback.	No
listen	object	A nested listen verb that streams session audio to a remote server via WebSocket.	No
notifyEvents	boolean	Enables event notifications over WebSocket connections. Verbs must include an id property to use this feature.	No
onHoldMusic	string	The URL to an audio file to play when the session is placed on hold.	No
recognizer	object	Contains configuration options for the speech recognition engine. This includes language selection, hints, diarization, and other advanced settings.	No
recognizer.vendor	string	The speech recognition provider to use, for example, Google, Amazon, or Azure. The vendor determines transcription quality, supported languages, and feature availability.	Yes
recognizer.label	string	A custom label to identify this recognizer instance in logs or dashboards. Helpful when multiple recognizers are configured.	No
recognizer.language	string	The primary language code for transcription, for example, en-US for English, fr-FR for French. Determines how speech is interpreted.	No
recognizer.hints	array	An array of words or phrases that may appear in the audio and should be recognized more accurately. Useful for domain-specific terms, names, or technical vocabulary.	No
recognizer.hintsBoost	number	A numeric value specifying how strongly the recognizer should prioritize the hint words. Higher numbers give stronger emphasis, improving accuracy for key terms.	No
recognizer.altLanguages	array	An array of additional language codes that the recognizer can use for multilingual audio. Allows recognition of mixed-language content.	No
recognizer.profanityFilter	boolean	If true, the recognizer will automatically remove or mask profanity from the transcription output.	No
recognizer.interim	boolean	If true, returns partial transcription results as the audio is being processed. Useful for live captions or real-time feedback.	No
recognizer.punctuation	boolean	If true, punctuation marks, for example, periods or commas, are included in the transcription to improve readability.	No
recognizer.diarization	boolean	If true, enables speaker diarization, which assigns segments of the transcript to individual speakers.	No
recognizer.diarizationMinSpeakers	number	The minimum number of speakers expected in the audio. Helps the diarization algorithm distinguish between speakers accurately.	No
recognizer.diarizationMaxSpeakers	number	The maximum number of speakers expected in the audio. Prevents the algorithm from splitting speech unnecessarily.	No
recognizer.vad	object	Voice Activity Detection settings. Determines how the system detects when someone is speaking vs. silence, improving transcription timing and accuracy.	No
recognizer.fallbackVendor	string	Specifies an alternative transcription vendor to use if the primary vendor fails. Ensures reliability in critical workflows.	No
recognizer.fallbackLanguage	string	Language code to use for the fallback vendor. Must match a language supported by the fallback provider.	No
referHook	string	A webhook URL to invoke when a SIP REFER is received in the session.	No
reset	string | array	Resets either the recognizer or synthesizer to default application settings.	No
record.action	string	The call recording action: startCallRecording, stopCallRecording, pauseCallRecording, or resumeCallRecording.	Yes
record.siprecServerURL	string | array	The SIP URI(s) for the SIPREC server. Required if record.action is startCallRecording.	Conditional
record.recordingID	string	A user-defined identifier for the recording.	No
record.headers	object	SIP headers to include in the SIPREC request.	No
sipRequestWithinDialogHook	string	A webhook to invoke when a SIP request (such as INFO, NOTIFY, REFER) is received within a dialog.	No
speechFallback.type	string	The type of fallback action (for example, dial or refer).	Yes (if used)
speechFallback.reason	string	The reason for executing the fallback (for example, recognizerFailure).	No
speechFallback.dial	object	A dial verb to execute as fallback if speech recognition fails.	No
speechFallback.refer	object	A sip:refer verb to execute as fallback if speech recognition fails.	No
synthesizer	object	Session-level text-to-speech settings. See Synthesizer Properties for details.	No
synthesizer.vendor	string	The TTS provider to use, for example, google, aws, microsoft, deepgram, elevenlabs, nuance, or custom:<provider-name>. The vendor determines the available voices, languages, and engine options. See supported speech vendors.	Yes
synthesizer.label	string	A custom label to identify this synthesizer instance. Useful when multiple TTS configurations from the same vendor are configured. Must match a label defined in the Voice Gateway Application.	No
synthesizer.language	string	The language code for the speech output, for example, en-US or de-DE. Required if a vendor is defined.	No
synthesizer.voice	string | object	The specific voice to use. Can be a string representing the vendor-specific voice name (for example, en-US-Wavenet-F for Google TTS) or an object with advanced properties. Defaults to the Application-level TTS voice if not provided.	No
synthesizer.engine	string	The TTS engine type. Options are: standard — the default engine neural — a high-quality natural voice generative — an experimental AI voice long-form — optimized for long text	No
synthesizer.gender	string	The desired voice gender: MALE, FEMALE, or NEUTRAL. Used for vendors that support gender selection.	No
synthesizer.options	object	A vendor-specific TTS options object. Common options include speakingRate (0.25–4.0), pitch (-20–20), and volumeGainDb (-96–16). These control the speech speed, pitch, and volume.	No
synthesizer.fallbackVendor	string	An alternative TTS vendor to use if the primary vendor fails or returns an error.	No
synthesizer.fallbackLabel	string	A label for the fallback TTS instance. Must match a label defined in the Voice Gateway Application.	No
synthesizer.fallbackLanguage	string	The language code for the fallback synthesizer. Defaults to the primary language if not provided.	No
synthesizer.fallbackVoice	string | object	The voice for the fallback synthesizer. Can be a string or object.	No
transcribe	object	A nested transcribe verb for background transcription of audio.	No
vad.enable	boolean	Enables or disables Voice Activity Detection (VAD).	Yes (if used)
vad.voiceMs	number	Duration in milliseconds of voice required to trigger detection.	No
vad.silenceMs	number	Duration in milliseconds of silence required to end detection.	No
vad.strategy	string	The VAD detection strategy (for example, adaptive or fixed).	No
vad.mode	number	Numeric value representing VAD sensitivity mode.	No

More Information

• Listen
• Gather