VoiceAgentSessionUpdateConfig interface

安定したリアルタイムセッション設定は session.update クライアントイベントで受け入れられています。

プロパティ

animation

セッションのアニメーション設定。

audio

セッションの入出力オーディオ設定です。

avatar

セッションのアバター設定。

greeting

セッション設定後に積極的なアシスタントの挨拶が始まりました。

include

サービス出力に含める追加のフィールド。

instructions

指示はセッション中ずっと適用されていました。

interim_response

遅延やツール実行のための中間応答設定。

max_output_tokens

1つの応答に対する最大出力トークン数。

metadata

セッションには最大16本の文字列キー・値ペアが接続されます。

output_modalities

出力モダリティはセッションで可能になりました。

parallel_tool_calls

モデルが複数のツールを並列に呼び出しできるかどうか。

reasoning

互換性のあるリアルタイムモデルの推論設定。

temperature

互換性のあるカスケードパイプラインのサンプリング温度。

tools

セッションで利用可能なツール。

tool_choice

セッション中の道具選択行動。

type

セッションの種類。 常に realtime です。

プロパティの詳細

animation

セッションのアニメーション設定。

animation?: VoiceAgentAnimationConfig

プロパティ値

audio

セッションの入出力オーディオ設定です。

audio?: VoiceAgentAudioConfig

プロパティ値

avatar

セッションのアバター設定。

avatar?: VoiceAgentSessionAvatarConfig

プロパティ値

greeting

セッション設定後に積極的なアシスタントの挨拶が始まりました。

greeting?: VoiceAgentGreetingConfigUnion

プロパティ値

include

サービス出力に含める追加のフィールド。

include?: VoiceAgentSessionIncludeOption[]

プロパティ値

instructions

指示はセッション中ずっと適用されていました。

instructions?: string

プロパティ値

string

interim_response

遅延やツール実行のための中間応答設定。

interim_response?: VoiceAgentInterimResponseConfigUnion

プロパティ値

max_output_tokens

1つの応答に対する最大出力トークン数。

max_output_tokens?: VoiceAgentMaxOutputTokens

プロパティ値

metadata

セッションには最大16本の文字列キー・値ペアが接続されます。

metadata?: Record<string, string>

プロパティ値

Record<string, string>

output_modalities

出力モダリティはセッションで可能になりました。

output_modalities?: VoiceOutputModality[]

プロパティ値

parallel_tool_calls

モデルが複数のツールを並列に呼び出しできるかどうか。

parallel_tool_calls?: boolean

プロパティ値

boolean

reasoning

互換性のあるリアルタイムモデルの推論設定。

reasoning?: RealtimeReasoning

プロパティ値

temperature

互換性のあるカスケードパイプラインのサンプリング温度。

temperature?: number

プロパティ値

number

tools

セッションで利用可能なツール。

tools?: VoiceAgentToolUnion[]

プロパティ値

tool_choice

セッション中の道具選択行動。

tool_choice?: VoiceAgentToolChoice

プロパティ値

type

セッションの種類。 常に realtime です。

type: "realtime"

プロパティ値

"realtime"