curl --request POST \
--url https://dev.voice.ai/api/v1/agent/init-agent \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '
{
"agent_template": "<string>",
"name": "<string>"
}
'{
"agent_template": {
"prompt": "<string>",
"greeting": "<string>",
"llm_temperature": 0.7,
"llm_model": "gemini-2.5-flash-lite",
"tts_min_sentence_len": 20,
"tts_params": {
"voice_id": "<string>",
"temperature": 1,
"top_p": 0.8,
"response_format": "mp3"
},
"min_silence_duration": 0.55,
"min_speech_duration": 0.1,
"user_silence_timeout": 10,
"max_call_duration_seconds": 900,
"allow_interruptions": true,
"allow_interruptions_on_greeting": false,
"min_interruption_words": 1,
"auto_noise_reduction": true,
"allow_agent_to_end_call": false,
"allow_agent_to_skip_turn": false,
"min_endpointing_delay": 0.5,
"max_endpointing_delay": 3,
"vad_activation_threshold": 0.6,
"phone_number": "<string>",
"mcp_servers": [
{}
],
"category": "business",
"playground_visibility": "unlisted"
},
"available_types": [
"<string>"
],
"description": "Pre-configured agent template based on selected type"
}Initialize Agent From Template
Create an agent from a saved template.
curl --request POST \
--url https://dev.voice.ai/api/v1/agent/init-agent \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '
{
"agent_template": "<string>",
"name": "<string>"
}
'{
"agent_template": {
"prompt": "<string>",
"greeting": "<string>",
"llm_temperature": 0.7,
"llm_model": "gemini-2.5-flash-lite",
"tts_min_sentence_len": 20,
"tts_params": {
"voice_id": "<string>",
"temperature": 1,
"top_p": 0.8,
"response_format": "mp3"
},
"min_silence_duration": 0.55,
"min_speech_duration": 0.1,
"user_silence_timeout": 10,
"max_call_duration_seconds": 900,
"allow_interruptions": true,
"allow_interruptions_on_greeting": false,
"min_interruption_words": 1,
"auto_noise_reduction": true,
"allow_agent_to_end_call": false,
"allow_agent_to_skip_turn": false,
"min_endpointing_delay": 0.5,
"max_endpointing_delay": 3,
"vad_activation_threshold": 0.6,
"phone_number": "<string>",
"mcp_servers": [
{}
],
"category": "business",
"playground_visibility": "unlisted"
},
"available_types": [
"<string>"
],
"description": "Pre-configured agent template based on selected type"
}Authorizations
Bearer token authentication. Use your API key as the bearer token. Format: Authorization: Bearer
Response
Successful Response
Show child attributes
Show child attributes
Show child attributes
Show child attributes
Voice ID to use for text-to-speech generation
Sampling temperature (0.0-2.0). Higher values make output more random.
0 <= x <= 2Nucleus sampling parameter (0.0-1.0). Controls diversity of output.
0 <= x <= 1Audio format: mp3, wav, or pcm
Whether to allow interruptions during the greeting message
Minimum words required for interruption (0=any speech, 1+=word count)
0 <= x <= 10Enable automatic noise reduction based on environment
Whether agent can end calls via tool or timeout
Whether agent can skip turns and yield conversation control
Phone number in E.164 format (supports multiple providers). Empty strings will be converted to None.
MCP servers configuration. List of MCP server configs with name, description, authentication and connection settings. Transport type is automatically detected from URL path.
business, playground public, unlisted