REST: session.updated
HTTP request
POST https://ai.api.cloud.yandex.net/v1/realtime/server-events/session.updated
Body parameters
Request schema: application/json
{
"type": "string",
"event_id": "string",
"session": {
"type": "string",
"object": "string",
"id": "string",
"model": "string",
"output_modalities": [
"string"
],
"instructions": "string",
"tools": [
{}
],
"turn_detection": {},
"input_audio_transcription": {},
"voice": "string | null",
"audio": {}
}
}
Returned when a session is updated with a session.update event, unless there is an error.
|
Field |
Description |
|
type |
enum Required field. The event type, must be
|
|
event_id |
string Required field. The unique ID of the server event. |
|
session |
Required field. Realtime session configuration object. |
RealtimeSession
Realtime session configuration object.
|
Field |
Description |
|
type |
enum Required field. The type of session. Always
|
|
object |
string Object type identifier (e.g., realtime.session). |
|
id |
string Unique identifier of the Realtime session. |
|
model |
string Model used to generate responses during the session. Currently, only gpt://{YANDEX_CLOUD_FOLDER_ID}/speech-realtime-250923 is supported. |
|
output_modalities[] |
string |
|
instructions |
string Default system instructions prepended to model calls. Used to guide response content, format, and audio behavior. |
|
tools[] |
|
|
turn_detection |
Server-side turn detection (VAD) configuration for audio input. |
|
input_audio_transcription |
Configuration for transcribing input audio. |
|
voice |
string | null Voice identifier for audio output. |
|
audio |
Configuration for input and output audio. |
Response
HTTP Code: 200
OK