Endpoint
PUT /user/assistant/{id}
Path parameters
Unique identifier of the assistant to update.
Request
All body parameters are optional. Provide only the fields you want to update.Core fields
Name of the assistant. Maximum 255 characters.
Voice ID. Use the Get Voices endpoint with the
mode parameter to retrieve compatible voices.Primary language ID.
Assistant type:
inbound or outbound.Engine mode:
pipeline, multimodal, or dualplex.Timezone identifier (e.g.,
America/New_York, Europe/Bucharest).Greeting message spoken at the start of a call. Maximum 200 characters.
System prompt defining the assistant’s behavior.
Mode-specific fields
LLM model ID. Applies to
pipeline mode only.Multimodal model ID. Applies to
multimodal and dualplex modes.Fallback LLM for tool calls in multimodal/dualplex modes. Set to
null to remove.Turn detection sensitivity for multimodal/dualplex modes. Range: 0–1. Set to
null for auto.Secondary languages
Additional language IDs. Replaces all existing secondary languages. Pass
[] to remove all.Knowledge base
ID of the knowledge base to attach. Set to
null to remove.How the assistant uses the knowledge base:
function_call or prompt.Phone number
ID of the phone number to assign. Set to
null to unassign.Custom mid-call tools
Custom mid-call tool IDs. Replaces all existing tool assignments. Pass
[] to remove all tools.Built-in tools
Built-in tools. Replaces all existing built-in tools. Pass
[] to remove all.Voice and TTS settings
Voice stability. Range: 0–1.
Voice similarity. Range: 0–1.
Speech speed multiplier. Range: 0.7–1.2.
LLM temperature. Range: 0–1.
Whether to enable emotional TTS synthesis.
Custom TTS provider ID. Set to
null to use language default.Custom STT provider ID. Pipeline mode only. Set to
null to use language default.Call behavior
Whether the caller can interrupt the assistant.
Whether to use filler audio. Pipeline mode only.
Custom filler phrases per category. Replaces existing filler config.
Whether to record calls.
Whether to enable noise cancellation.
Whether the assistant waits for the customer to speak first.
Timing
Maximum call duration in seconds. Range: 20–1200.
Maximum silence duration before re-engagement, in seconds. Range: 1–360.
Maximum silence at call start before ending, in seconds. Set to
null to disable.Maximum ringing time in seconds. Range: 1–60.
Re-engagement
Seconds between re-engagement attempts. Range: 7–600.
Custom re-engagement message prompt. Maximum 1000 characters. Set to
null to use default.Voicemail
Whether to end the call when voicemail is detected.
Message to leave on voicemail. Maximum 1000 characters. Set to
null to disable.Endpoint detection
Voice activity detection type:
vad or ai.Endpoint sensitivity. Range: 0–5.
Interrupt sensitivity. Range: 0–5.
Minimum words before an interruption is processed. Range: 0–10. Set to
null to disable.Ambient sound
Background ambient sound:
off, office, city, forest, crowded_room, cafe, or nature.Ambient sound volume. Range: 0–1.
Webhook configuration
Whether webhook notifications are enabled.
Webhook URL for post-call notifications. Set to
null to remove.Whether webhooks are only sent for completed calls.
Whether to include the recording URL in webhook payloads.
Post-call evaluation
Whether to enable AI post-call evaluation.
Post-call data extraction schema. Replaces the existing schema. See Create assistant for field details.
Variables
Custom variables. Replaces all existing variables.
Conversation ended settings
Minutes of inactivity before the conversation ends. Range: 1–1440. Set to
null to disable.Whether to allow re-triggering after inactivity ends the conversation.
Webhook URL for conversation ended events. Set to
null to remove.Response
Confirmation message.
Example
200 Success
404 Not found