Update one or more settings of an existing AI assistant. All fields are optional - only include the fields you want to modify.
Path Parameters
The unique identifier of the assistant to update
Request Body
Core Fields
Assistant display name (maximum 255 characters)
Voice identifier to use (must exist in available voices list)
Language name for the assistant (maximum 100 characters)
Large Language Model name to use (maximum 100 characters)
Call direction configuration. Options: receive, make
Engine type selection. Options: pipeline, multimodal
Timezone setting (example: “America/New_York”)
Opening message the assistant will speak at call start
System prompt that defines the assistant’s behavior, personality, and instructions
Optional Configuration
Phone number identifier to assign (set to null to remove assignment)
Array of mid-call tool identifiers to synchronize with the assistant. This replaces all existing tool assignments. Pass an empty array to remove all tools. Each ID must be a valid tool belonging to your account.
Voice activity detection method. Options: vad, ai
Endpoint sensitivity level (range: 0-5)
Interruption sensitivity level (range: 0-5)
Ambient sound volume level (range: 0-1)
Enable AI-powered call evaluation after call completion
send_webhook_only_on_completed
Send webhooks only for successfully completed calls
include_recording_in_webhook
Include call recording URL in webhook payload
Webhook Configuration
Enable webhook notifications
Webhook endpoint URL for post-call notifications (can be set to null to remove)
Interruption Settings
Enable minimum interrupt words configuration
Minimum words required before interruption is allowed (range: 0-10)
Variables and Schema
Custom key-value variable pairs for the assistant
Schema definition for AI data extraction after calls Show post_call_schema properties
Schema field name (alphanumeric characters and underscores only)
Data type. Options: string, number, bool
post_call_schema[].description
Description explaining what this field represents
end_call_tool.description
Description for end call tool functionality (maximum 500 characters)
Voice and Speech Settings
LLM temperature configuration (range: 0-1)
Voice stability configuration (range: 0-1)
Voice similarity configuration (range: 0-1)
Speech speed multiplier (range: 0.7-1.2)
Call Behavior Settings
Allow customers to interrupt the assistant
Use filler audio responses during processing
Re-engagement interval in seconds (range: 7-600)
Maximum call length in seconds (range: 20-1200)
Maximum silence duration before re-engagement in seconds (range: 1-120)
Automatically end call when voicemail is detected
Enable noise cancellation during calls
Party that speaks first. Options: AI assistant, Customer
Response
Success confirmation message
Unique identifier of the updated assistant
Assistant type (inbound or outbound)
Engine mode (pipeline or multimodal)
Error Responses
Error when the assistant doesn’t exist or belongs to another account
Error message describing validation failure
Detailed field-level validation errors
500 Internal Server Error
Error message indicating assistant update failure
200 Success Response
404 Not Found
422 Validation Error
500 Internal Server Error
{
"message" : "Assistant updated successfully" ,
"data" : {
"id" : 789 ,
"name" : "Updated Sales Assistant" ,
"status" : "active" ,
"type" : "outbound" ,
"mode" : "pipeline"
}
}
Notes
Only the fields included in your request body will be updated
All validation rules are identical to assistant creation
To remove a phone number assignment, set phone_number_id to null
To disable webhooks, set webhook_url to null or is_webhook_active to false
The assistant must belong to your account to be updated