LlmInterimResponseConfig interface

Package:

@azure/ai-voicelive

Configuration for LLM-based interim response generation. Uses LLM to generate context-aware interim responses when any trigger condition is met.

Extends

InterimResponseConfigBase

Properties

instructions	Custom instructions for generating interim responses. If not provided, a default prompt is used.
maxCompletionTokens	Maximum number of tokens to generate for the interim response.
model	The model to use for LLM-based interim response generation. Default is gpt-4.1-mini.
type	The discriminator possible values: static_interim_response, llm_interim_response

Inherited Properties

latencyThresholdInMs	Latency threshold in milliseconds before triggering interim response. Default is 2000ms.
triggers	List of triggers that can fire the interim response. Any trigger can activate it (OR logic). Supported: 'latency', 'tool'.

Property Details

instructions

Custom instructions for generating interim responses. If not provided, a default prompt is used.

instructions?: string

Property Value

string

maxCompletionTokens

Maximum number of tokens to generate for the interim response.

maxCompletionTokens?: number

Property Value

number

model

The model to use for LLM-based interim response generation. Default is gpt-4.1-mini.

model?: string

Property Value

string

type

The discriminator possible values: static_interim_response, llm_interim_response

type: "llm_interim_response"

Property Value

"llm_interim_response"

Inherited Property Details

latencyThresholdInMs

Latency threshold in milliseconds before triggering interim response. Default is 2000ms.

latencyThresholdInMs?: number

Property Value

number

Inherited From InterimResponseConfigBase.latencyThresholdInMs

triggers

List of triggers that can fire the interim response. Any trigger can activate it (OR logic). Supported: 'latency', 'tool'.

triggers?: string[]

Property Value

string[]

Inherited From InterimResponseConfigBase.triggers