Start a New Streaming Session

POSThttps://api.konpro.ai/v1/streaming/new

Initialize a new real-time streaming session featuring an Interactive Avatar. This session enables live interaction and communication with customizable settings.

Request History

TIME	STATUS	USER AGENT
Make a request to see history.

0 Requests This Month

Request Body Parameters

Field	Type	Description
quality	string	Streaming quality. Options: "high" (2000 kbps, 720p), "medium" (1000 kbps, 480p, default), "low" (500 kbps, 360p)
avatar_id	string	(Optional) ID of the Interactive Avatar to use. Defaults to "default" if omitted.
voice	object	(Optional) Voice settings for the avatar, including rate, emotion, and third-party configurations.
stt_settings	object	(Optional, beta) Speech-to-text configuration, selecting provider (e.g., deepgram, gladia, assembly_ai) and confidence threshold.
video_encoding	string	Video encoding format. Options: "H264" or "VP8" (default).
knowledge_base	string	(Optional) Prompt for chatbot knowledge base in chat mode.
version	string	(Optional) API version to use. Valid: "v1". Defaults to "v1" if omitted.
knowledge_base_id	string	(Optional, beta) Knowledge base identifier for avatar responses, applicable only when version is "v1".
disable_idle_timeout	boolean	(Optional) Disable the default 2-minute idle timeout. Use with caution to avoid excessive credit consumption. Defaults to false.
activity_idle_timeout	integer	(Optional) Max idle time in seconds before session is deemed inactive. Default: 120s, Min: 30s, Max: 3600s.
livekit_settings	object	(Optional, beta) Custom LiveKit instance connection settings, including room, url, and access token.

Voice Settings Object (Example)

Field	Type	Description
voice_id	string	Voice ID; not all voices support streaming API
rate	float	Speech speed rate, default is 1
emotion	string	Emotion style, e.g., "Excited", "Friendly"
elevenlabs_settings	object	ElevenLabs-specific voice model parameters (optional)

Speech-to-Text Settings Object (Example)

Field	Type	Description
provider	string	STT provider: "deepgram", "gladia", "assembly_ai" (default "assembly_ai" for English)
confidence	float	Confidence threshold, default 0.55

Response Fields

Field	Type	Description
code	integer	Response status code
message	string	Human-readable response message
data.session_id	string	Unique session identifier
data.url	string	WebSocket URL for connecting to the LiveKit room
data.access_token	string	Access token to authenticate with the LiveKit server
data.session_duration_limit	integer	Maximum session duration in seconds
data.is_paid	boolean	Indicates if the session is under a paid plan
data.realtime_endpoint	string	URL for alpha real-time signaling features (optional)
data.livekit_agent_token	string	Token for HeyGen's audio agents, only if using managed LiveKit

cURL Request

shell

curl --request POST \
--url https://api.konpro.ai/v1/streaming/new \
--header 'accept: application/json' \
--header 'content-type: application/json' \
--header 'x-api-key: <your-api-key>' \
--data '{
  "quality": "medium",
  "avatar_id": "default",
  "voice": {
    "rate": 1
  },
  "video_encoding": "VP8",
  "disable_idle_timeout": false,
  "version": "v1",
  "stt_settings": {
    "provider": "deepgram",
    "confidence": 0.55
  },
  "activity_idle_timeout": 120
}'

LANGUAGE

CREDENTIALS

Header

HEADER

RESPONSE

Examples

Choose an example:

application/json

200 - Success

400 - Bad Request

Generate Video from Proofread Start Session