To create a voice configuration, first create a YAML file that defines your voice settings. This file includes the voice name, speech-to-text and text-to-speech provider settings, the primary language, and optional configurations for advanced settings.
Deepgram WebSocket API URL (e.g., “wss://api.deepgram.com/v1/listen”)
string
required
Deepgram API key
string
required
Speech recognition model (e.g., “nova-2”, “nova-3”). Only nova-2 and nova-3 models (and their variations like nova-2-finance, nova-3-medical) are officially supported.
string
Language code (e.g., “en-US”)
string
Language hint for Flux multilingual model to improve transcription accuracy. See Flux Multilingual & Language Prompting for supported languages and usage. Do not set this parameter when using the flux-general-en model.
boolean
Convert numbers to numerals
array
List of keywords to boost recognition accuracy (e.g., [“help”, “support”, “Mr. Smith”]). Only supported for nova-3 models and their variations. Using this parameter with nova-2 models will result in a validation error.
array
List of keywords to detect with optional boost/suppression values (e.g., [“snuffleupagus:5”, “monster”, “cookie:-2”]). Only supported for nova-2 models and their variations. Using this parameter with nova-3 models will result in a validation error.
boolean
Opt out of model improvement program
integer
Number of audio channels
boolean
Enable speaker diarization
boolean
Enable dictation mode
integer
Endpointing silence duration in seconds
array
Extra parameters to pass to Deepgram (e.g., [“custom_param1:value1”])
boolean
Enable interim results
boolean
Transcribe each audio channel independently
boolean
Filter profanity
boolean
Add punctuation and capitalization
array
Redact sensitive information (e.g., [“pci”, “pii”])
Search for specific terms (e.g., [“action item”, “follow up”])
boolean
Apply smart formatting to the transcript
array
Tags for the request (e.g., [“customer_service”, “sales_call”])
integer
How long Deepgram will wait to send UtteranceEnd message after word has been transcribed
boolean
Enable Deepgram’s voice activity detection events
string
API version (e.g., “latest”)
float
Only for v2 endpoint. End-of-turn confidence required to fire an eager EOT event (0.3-0.9)
float
Only for v2 endpoint. End-of-turn confidence required to finish a turn (0.5-0.9)
integer
Only for v2 endpoint. Time (ms) after speech to finish turn regardless of EOT confidence
Deepgram TTS Configuration Parameters:
Show detailed parameter descriptions
string
required
Deepgram API key
string
Deepgram API URL
string
Language code (e.g., “en”)
string
Voice model (e.g., “aura-2-thalia-en”)
boolean
Opt out of model improvement program
Available Deepgram models:Deepgram provides two generations of text-to-speech (TTS) models. Both work with the deepgram_tts provider.
Aura 2 (for example, aura-2-thalia-en) is the existing model.
Flux TTS (for example, flux-afsana-bn) is a newer model.
Set the model parameter to the model name listed in the tables. The model name follows the pattern flux-<voice>-<language>.Speech to text (Flux):
Model name
Languages
Multilingual
flux-general-en
English (en, en-AU, en-CA, en-GB, en-IN, en-NZ, en-US)
No
flux-general-multi
English (en, en-AU, en-CA, en-GB, en-IN, en-NZ, en-US), German (de), Spanish (es), French (fr), Hindi (hi), Italian (it), Japanese (ja), Dutch (nl), Portuguese (pt), Russian (ru)
Yes
Text to speech (Flux TTS):
Show Flux TTS voices (language, model name, and style)
ElevenLabs API key. You do not need to provide this key when you use ElevenLabs as an OEM provider.
string
default:"scribe_v2_realtime"
ElevenLabs STT model ID such as scribe_v2_realtime.
string
ElevenLabs STT service URL such as https://api.elevenlabs.io/v1/speech-to-text.
string
Language code for speech recognition such as en.
boolean
default:"false"
Include word-level timestamps in transcription results.
boolean
default:"false"
Enable server-side logging for the session.
boolean
default:"false"
Include detected language in transcription results.
string
Commit strategy to use. Set to vad to commit based on voice activity detection, or set to manual to commit manually.
float
Seconds of silence before VAD commits an utterance. Use this parameter only when you set commit_strategy to vad.
float
VAD speech probability threshold from 0.0 to 1.0. Use this parameter only when you set commit_strategy to vad.
integer
Minimum speech duration in milliseconds for VAD to consider valid speech, between 50 and 2000. Use this parameter only when you set commit_strategy to vad.
integer
Minimum silence duration in milliseconds for VAD to consider a speech break, between 50 and 2000. Use this parameter only when you set commit_strategy to vad.
Voice ID from ElevenLabs (e.g., “CwhRBWXzGAHq8TQ4Fs17”)
string
Language code (e.g., “en”, “ja” for Japanese)
string
Text normalization mode (“auto”, “on”, “off”)
integer
Optimize streaming latency (0-4)
boolean
Apply language-specific text normalization. Currently only supported for Japanese (ja). Using this parameter with other languages will result in a validation error.
array
List of pronunciation dictionary locators with pronunciation_dictionary_id and version_id
integer
Seed for deterministic audio generation
object
Advanced voice customization settings
float
Speech speed (default: 1.0)
float
Voice stability - how stable the voice is and the randomness between each generation (0.0-1.0, default: 0.5)
float
Style exaggeration - the higher the value, the more computational resources are used (0.0-1.0, default: 0.0)
float
Similarity boost - how closely the AI should adhere to the original voice (0.0-1.0, default: 0.75)
Speech pitch adjustment (-20.0 to 20.0, default: 0.0)
float
Volume gain in dB (-96.0 to 16.0, default: 0.0)
integer
Audio sample rate in Hertz (e.g., 24000)
string
Audio encoding format (e.g., LINEAR16, MP3, OGG_OPUS)
Advanced Voice Settings
Voice configurations support advanced features for enhanced call handling:
Voice Activity Detection (VAD)
Automatically detect when the user is speaking:
YAML
vad: enabled: true provider: "silero_vad" silero_vad_config: confidence: 0.7 # Speech detection confidence threshold (0.0-1.0) start_seconds: 0.2 # Time before transitioning to SPEAKING state stop_seconds: 0.8 # Time before transitioning to QUIET state min_volume: 0.6 # Minimum audio volume threshold (0.0-1.0)
DTMF Input (Dual-Tone Multi-Frequency)
Enable keypad input during voice calls:
YAML
dtmf_input: inter_digit_timeout_ms: 2500 # Time to wait for next digit (ms) termination_key: "#" # Key to signal end of input maximum_count: 10 # Maximum number of digits ignore_speech: true # Disable speech recognition during DTMF
User Idle Handler
Handle situations when the user stops responding:
YAML
user_idle_handler: enabled: true idle_timeout: 7 # Timeout in seconds idle_max_reprompts: 2 # Max reprompt attempts idle_timeout_message: "Are you still there?" # Message to play on idle idle_hangup_message: "" # Message to play before hanging up repeat_previous_message: true
User Idle Handler Configuration Parameters:
Show detailed parameter descriptions
boolean
Enable idle handling (default: false)
integer
Seconds of user inactivity before triggering the handler (default: 7)
integer
Number of reprompt attempts before ending the call (default: 2)
string
Message played to the user on idle timeout (e.g., "Are you still there?")
string
Message played before hanging up after all reprompts are exhausted
boolean
Repeat the last agent message before the idle prompt. When false, only the idle_timeout_message is played (default: true)
Agent Idle Handler
Provide feedback when the agent is processing:
YAML
agent_idle_handler: typing_enabled: true # Enable typing sound typing_duration_seconds: 5 # Typing sound duration audio_clip_id: "guitar_1" # Hold Audio hold_audio_seconds: 15 # Hold audio duration pre_hold_message: "Please wait a moment..." # Message before hold audio pre_hold_messages_additional: [] hold_message: "Still processing your request..." # Periodic hold message hold_messages_additional: [] long_running_task_seconds: 2
Agent Idle Handler Configuration Parameters:
Show detailed parameter descriptions
boolean
Show a typing indicator while the agent is processing (default: true)
integer
Duration (seconds) to show the typing indicator (default: 5)
string
ID of the hold music audio clip to play (e.g., "guitar_1")
integer
Seconds to play hold audio before switching to the hold message (default: 15)
string
Message played before hold audio begins (e.g., "One moment please")
array
Additional pre-hold messages randomly selected alongside pre_hold_message (e.g., ["Just a moment...", "Bear with me..."])
string
Message played periodically while the agent is processing (e.g., "Just a few more seconds, thanks for your patience")
array
Additional hold messages randomly selected alongside hold_message (e.g., ["Almost there...", "Thank you for waiting..."])
integer
Seconds of processing time before the pre-hold message is played (default: 2)
Path to the YAML file with the voice configuration.
Note:
The import command creates a new voice configuration or updates an existing one based on the name. If a voice configuration with the same name exists, the command updates it using the new configuration.
Listing voice configurations
To list all available voice configurations:
BASH
orchestrate voice-configs list
Show command flags
Show full details of all voice configurations in JSON format.
Getting voice configuration details
To retrieve details of a specific voice configuration:
BASH
orchestrate voice-configs get --name <config_name>
Show command flags
string
Voice configuration ID to retrieve. Either ID or name is required.
string
Voice configuration name to retrieve. Either ID or name is required.
Voice configuration ID to remove. Either ID or name is required.
string
Voice configuration name to remove. Either ID or name is required.
Note:
If both --id and --name are provided, the ID takes precedence.
Important:
When you remove a voice configuration, agents using this configuration will no longer have voice capabilities. Make sure you want to remove the configuration before you proceed.