Understand Android SDK Data Structures for IMS - Intelligent Media Services

Read this topic to learn about the data types used in the Android SDK.

Data Structures Overview

Note

The legacy SDK version includes deprecated parameters and methods. Upgrade to the latest SDK version. For more information, see Android User Guide.

Structure Type	Data Type	Description
*Enum*	ARTCAICallAgentType	AI agent type
	ARTCAICallRobotState	Current robot state
	AICallErrorCode	An error occurred during the AI call
	VoicePrintStatusCode	ASR status code
	ARTCAICallSpeakingInterruptedReason	Reason the agent speech was interrupted
	ARTCAICallAudioProfile	Audio encoding configuration
	ARTCAICallAudioScenario	Audio scenario configuration
	ARTCAICallVideoRenderMode	Video rendering mode
	ARTCAICallVideoRotationMode	Video rotation angle
	ARTCAICallVideoRenderMirrorMode	Video rendering mirror mode
	ARTCAICallConnectionStatus	Network connection status during the call
	ARTCAICallTurnDetectionMode	Method to detect when user speech ends
*Class*	ARTCAICallConfig	Call configuration object
	ARTCAICallVideoConfig	Video configuration parameters
	ARTCAICallAudioConfig	Audio configuration parameters
	ARTCAICallAgentTemplateConfig (deprecated)	Configurable parameters for the call
	ARTCAICallChatSyncConfig	Configuration for synchronizing chat agent conversation history
	ARTCAICallAgentInfo	Agent runtime information
	ARTCAICallSendTextToAgentRequest	Request object to send text to the agent
	ARTCAICallVisionCustomCaptureRequest	Request object for custom frame capture by a multimodal large language model
	ARTCAICallVideoCanvas	Video rendering configuration object
	ARTCAICallAgentConfig	Configurable parameters for the call
	ARTCAICallAgentAsrConfig	Speech recognition configuration
	ARTCAICallAgentTtsConfig	Speech synthesis configuration
	ARTCAICallAgentLlmConfig	Large Language Model (LLM) configuration
	ARTCAICallAgentAvatarConfig	Digital human configuration
	ARTCAICallAgentInterruptConfig	Interrupt configuration
	ARTCAICallAgentVoiceprintConfig	Voiceprint denoising configuration
	ARTCAICallAgentTurnDetectionConfig	Turn detection configuration
	ARTCAICallAgentVcrResult	VCR detection result
	FrameMotionResult	Video frame detection result from VCR
	PeopleCountResult	People count detection result from VCR
	EquipmentResult	Electronic device detection result from VCR
	HeadMotionResult	Head motion detection result from VCR
	ARTCAICallAgentVcrConfig	VCR configuration
	ARTCAICallAgentVcrBaseConfig	Base detection configuration for VCR
	ARTCAICallAgentVcrFrameMotionConfig	Video frame detection configuration for VCR
	ARTCAICallExperimentalConfig	Experimental parameters for controlling specific logic policies
	ARTCAICallAgentAmbientConfig	Call environment parameters
	ARTCAICallAgentAutoSpeechContent	Agent speech content for auto-speech scenarios, such as acknowledgments or proactive questions
	ARTCAICallAgentAutoSpeechLlmPending	Auto-speech configuration for the agent when LLM response is delayed
	ARTCAICallAgentAutoSpeechUserIdle	Configuration for agent queries when the user is silent
	ARTCAICallAgentBackChanneling	Configuration module for acknowledgment speech. When enabled, the agent randomly speaks short acknowledgments at specific trigger times

Data Structure Details

Enum

ARTCAICallAgentType

The AI agent type.

Enumeration Name	Description
VoiceAgent	Voice-only call
AvatarAgent	Digital human call
VisionAgent	Visual understanding call
VideoAgent	Video call
ChatBot	Chat message interaction

ARTCAICallRobotState

The current state of the robot.

Enumeration name	Description
Listening	Listening
Thinking	Thinking
Speaking	Speaking

AICallErrorCode

Indicates an error that occurred during the AI call.

Enumeration Name	Description
None	None
InvalidAction	Invalid API call
InvalidParams	Invalid parameter passed to the API
StartFailed	Failed to start the call
AgentSubscriptionRequired	Call initiation failed due to exceeding the daily free trial quota
AgentNotFund	Agent not found
TokenExpired	Call authentication token expired
ConnectionFailed	Connection failed. Call disconnected
KickedByUserReplace	Same user ID joined the same call on different devices, preventing the call from proceeding
KickedBySystem	Kicked out by the system, preventing the call from proceeding
LocalDeviceException	Local device issue prevented the call from proceeding
AgentLeaveChannel	Agent left the channel. Call ended
AgentConcurrentLimit	Digital human agent reached concurrent limit
AgentAudioSubscribeFailed	Failed to subscribe to agent audio
AiAgentAsrUnavailable	Third-party ASR service failed to start
AvatarAgentUnavailable	Digital human service unavailable
ChatLogNotFound	Chat log not found
InternalError	Internal error

VoicePrintStatusCode

The status code for Automatic Speech Recognition (ASR).

Enumeration Name	Description
Disable	Voiceprint denoising VAD disabled and AIVAD disabled
EnableWithoutRegister	Voiceprint denoising VAD enabled but voiceprint registration incomplete
SpeakerRecognized	Voiceprint denoising VAD enabled and speaker recognized
SpeakerNotRecognized	Voiceprint denoising VAD enabled but speaker not recognized
DetectedSpeakerWithAIVad	AIVAD enabled and speaker detected
UndetectedSpeakerWithAIVad	AIVAD enabled but speaker not detected
Unknown	Unknown status

ARTCAICallSpeakingInterruptedReason

The reason why the agent's speech was interrupted.

Enumeration Name	Description
UnKnown	Unknown reason
ByWorks	Specific interrupt words detected
ByVoice	Voice interruption detected
ByInterruptSpeaking	Interrupted by calling the interruptSpeaking() API
BySpeechBroadCast	Interrupted by proactive speech broadcast
ByLlmQuery	Interrupted by proactive LLM query

ARTCAICallAudioProfile

The audio encoding configuration.

Enumeration Name	Description
ARTCAICallAudioLowQualityMode	Low-quality audio mode. Default sample rate: 8000 Hz. Mono. Max bitrate: 12 kbps
ARTCAICallAudioBasicQualityMode	Standard-quality audio mode. Default sample rate: 16000 Hz. Mono. Max bitrate: 24 kbps
ARTCAICallAudioHighQualityMode	(Default) High-quality audio mode. Default sample rate: 48000 Hz. Mono. Max bitrate: 64 kbps
ARTCAICallAudioStereoHighQualityMode	Stereo high-quality audio mode. Default sample rate: 48000 Hz. Stereo. Max bitrate: 80 kbps
ARTCAICallAudioSuperHighQualityMode	Super-high-quality audio mode. Default sample rate: 48000 Hz. Mono. Max bitrate: 96 kbps
ARTCAICallAudioStereoSuperHighQualityMode	Stereo super-high-quality audio mode. Default sample rate: 48000 Hz. Stereo. Max bitrate: 128 kbps

ARTCAICallAudioScenario

The audio scenario configuration.

Enumeration Name	Description
ARTCAICallAudioSceneDefaultMode	Recommended for general real-time communication scenarios
ARTCAICallAudioSceneMusicMode	High-fidelity music audio quality. Recommended for music instruction and other scenarios requiring high music quality

ARTCAICallVideoRenderMode

The video rendering mode.

Enumeration Name	Description
ARTCAICallVideoRenderModeAuto	Automatic mode
ARTCAICallVideoRenderModeStretch	Stretch tiling mode. Stretches input video to match the stream ingest aspect ratio. May distort the image
ARTCAICallVideoRenderModeFill	Cropping mode. Crops input video width or height to match the stream ingest aspect ratio. May lose image content
ARTCAICallVideoRenderModeNoChange	No change

ARTCAICallVideoRotationMode

The video rotation angle.

Enumeration Name	Description
ARTCAICallVideoRotationMode_0	0 degrees
ARTCAICallVideoRotationMode_90	90 degrees
ARTCAICallVideoRotationMode_180	180 degrees
ARTCAICallVideoRotationMode_270	270 degrees

ARTCAICallVideoRenderMirrorMode

The video rendering mirror mode.

Enumeration Name	Description
ARTCAICallVideoRenderMirrorModeOnlyFront	Mirror only front camera preview. No mirroring for others
ARTCAICallVideoRenderMirrorModeAllEnabled	Mirror all
ARTCAICallVideoRenderMirrorModeAllDisable	Do not mirror any

ARTCAICallTurnDetectionMode

The method used to detect the end of a user's speech.

Enumeration Name	Description
ARTCAICallTurnDetectionNormalMode	Normal mode. Does not use AI for semantic analysis. Uses ASR silence duration to detect speech end
ARTCAICallTurnDetectionSemanticMode	Semantic mode. Uses AI to analyze context and semantics to detect speech end

ARTCAICallConnectionStatus

The network connection status during the call.

Enumeration Name	Value	Description
ARTCAICallConnectionStatusInit	0	Initialization complete
ARTCAICallConnectionStatusDisconnected	1	Network connection disconnected
ARTCAICallConnectionStatusConnecting	2	Establishing network connection
ARTCAICallConnectionStatusConnected	3	Network connected
ARTCAICallConnectionStatusReconnecting	4	Re-establishing network connection
ARTCAICallConnectionStatusFailed	5	Network connection failed

Class

ARTCAICallConfig

The call configuration object.

Parameter	Type	Description
agentId	String	Agent ID
agentType	ARTCAICallAgentType	Agent type. Must match the type of agentId. Otherwise, agent startup fails
agentUserId	String	User ID for the agent. If empty, the agent service assigns a UID
region	String	Region where the agent service resides. Must match the region of agentId. Otherwise, agent startup fails
userData	String	User-defined information. Passed to the agent
enableAudioDelayInfo	boolean	Enable audio delay statistics. Enabled by default
agentConfig	ARTCAICallAgentConfig	agentConfig parameter used to start the call
audioConfig	ARTCAICallAudioConfig	Local audio configuration
videoConfig	ARTCAICallVideoConfig	Local video configuration. Applies only to VisionAgent or VideoAgent
chatSyncConfig	ARTCAICallChatSyncConfig	Associated chat agent configuration. If set, call history syncs to the chat agent during the call
mAiCallVideoConfig	ARTCAICallVideoConfig	Video-related configuration (deprecated starting with version 2.5)
mAliCallAudioConfig	ARTCAICallAudioConfig	Audio-related configuration (deprecated starting with version 2.5)
mAiCallAgentTemplateConfig	ARTCAICallAgentTemplateConfig (deprecated)	Configurable parameters for AI audio-video calls (deprecated starting with version 2.5)
mAiCallChatSyncConfig	ARTCAICallChatSyncConfig	Chat synchronization configuration. Integrates call agent and chat agent conversation content into one session (deprecated starting with version 2.5)

ARTCAICallVideoConfig

The video configuration parameters.

Parameter	Type	Meaning
useHighQualityPreview	boolean	Use local high-definition preview
useFrontCameraDefault	boolean	Start front camera by default
cameraCaptureFrameRate	int	Camera capture frame rate
useSurfaceView	boolean	Use SurfaceView for rendering. If false, use TextureView
videoEncoderWidth	int	Video encoding width
videoEncoderHeight	int	Video encoding height
videoEncoderFrameRate	int	Video encoding frame rate
videoEncoderBitRate	int	Video encoding bitrate
videoEncoderKeyFrameInterval	int	Keyframe interval in milliseconds
isCameraMute	boolean	Mute local video. Not muted by default

ARTCAICallAudioConfig

The audio configuration parameters.

Parameter	Type	Description
audioProfile	ARTCAICallAudioProfile	Audio encoding configuration. Default: ARTCAICallAudioHighQualityMode. Set this parameter to specify sample rate and number of sound channels
audioScenario	ARTCAICallAudioScenario	Audio scenario configuration. Default: ARTCAICallAudioSceneMusicMode. Use ARTCAICallAudioSceneDefaultMode for Bluetooth audio capture
enableSpeaker	boolean	Play audio through speaker or earpiece. Speaker enabled by default. Set to false to switch to earpiece
isMicrophoneOn	boolean	Enable microphone. Enabled by default

ARTCAICallAgentTemplateConfig (deprecated)

The configurable parameters for the call.

Important

This type is deprecated starting with version 2.5. Use ARTCAICallAgentConfig instead.

Parameter	Type	Description
aiAgentId	String	Agent ID. Set this field when initiating an agent call using the AICallKit call interface. Note Agent ID must be set and cannot be empty.
aiAgentRegion	String	Region where the agent service resides. Default: cn-shanghai. Note Set the region based on your agent.
aiAgentUserId	String	User ID associated with the agent. If not set, the AI server generates a random user ID
userExtendData	String	Business extension information. Must be a JSON string. Passed to the LLM.
aiAgentGreeting	String	Agent greeting. Spoken by the AI agent when the user joins the call
aiAgentUserOnlineTimeout	int	Time in seconds before the agent closes the task if the user does not join the call. Default: 60 seconds
aiAgentUserOfflineTimeout	int	Time in seconds before the agent closes the task after the user leaves the call. Default: 5 seconds
aiAgentWorkflowOverrideParams	String	Workflow override parameters. Default: none
aiAgentBailianAppParams	String	Bailian application center parameters. For more information, see Passing User Information to Bailian.
aiAgentAsrMaxSilence	int	Maximum silence duration for speech recognition in milliseconds. Range: 200–1200 ms. Default: 400 ms
aiAgentVolume	int	Agent speaking volume. Range: -1 to 100. Default: -1. If unset, uses Alibaba Cloud's recommended adaptive volume mode
enableVoiceInterrupt	boolean	Enable voice interruption. Default: true
enableIntelligentSegment	boolean	Enable intelligent sentence segmentation. Default: true
enableVoicePrint	boolean	Enable voiceprint recognition. Default: false. To enable, set enableVoicePrint to true and provide voiceprintId
voiceprintId	String	Voiceprint ID. If enableVoicePrint is true and voiceprintId is non-empty, voiceprint denoising is enabled. Empty means voiceprint denoising is disabled
aiAgentVoiceId	String	Agent voice ID
aiAgentMaxIdleTime	int	Maximum idle time in seconds before the agent goes offline automatically. Set to -1 to disable automatic logout. Default: 600 seconds
llmHistoryLimit	int	Maximum number of LLM/multimodal LLM conversation history turns to retain. Less than zero uses the server default: 10 turns
aiAgentGracefulShutdown	boolean	Enable graceful shutdown. Default: false Note Graceful shutdown means the agent finishes speaking its current sentence before stopping. Maximum duration: 10 seconds
enablePushToTalk	boolean	Enable push-to-talk mode. Default: false
aiAgentAvatarId	String	Digital human model ID. Specify when the agent type is AvatarAgent
asrLanguageId	String	ASR language ID. Empty uses the agent's default configuration. Options: zh_mandarin: Chinese en: English zh_en: Chinese-English mixed es: Spanish jp: Japanese
wakeUpQuery	String	Wake-up phrase. Spoken by the user before the call starts. The agent responds immediately after the call starts. Example: "What's the weather like today?"
llmSystemPrompt	String	LLM system prompt. Example: "You are a friendly and helpful assistant focused on providing accurate information and advice." Note Not supported for LLM nodes configured as Bailian workflows.
interruptWords	List<String>	Trigger words for interrupting the conversation. Examples: "Hold on", "I know"
aiAgentLlmHistoryLimit	int	Maximum number of LLM/multimodal LLM conversation history turns to retain. Default: 10 turns
aiAgentVadLevel	int	VAD sensitivity setting for AIVAD. Higher values increase resistance to voice interference. Range: [0–11]. Default: 11 0 disables VAD. 1–10: Higher numbers make interruption harder. 11 significantly improves voice preservation and noise resistance.

ARTCAICallChatSyncConfig

The configuration for synchronizing the chat agent's conversation history.

Parameter	Type	Description
sessionId	String	Unique identifier for a user-agent conversation. Default: empty
chatBotAgentId	String	Agent ID for the associated chat bot
receiverId	String	Receiver ID for the chat bot agent. This is the user's user ID

ARTCAICallAgentInfo

The agent's runtime information.

Parameter	Type	Description
agentId	String	Current agent ID
agentType	ARTCAICallAgentType	Agent type
agentUserId	String	User ID for the agent in the RTC channel
channelId	String	RTC channel ID where the agent resides
instanceId	String	Instance ID for the current agent. Assigned by the system when the agent starts. Used to identify and track the agent's full lifecycle and runtime status
requestId	String	Request ID for the current agent

ARTCAICallSendTextToAgentRequest

The request object for sending text to the agent.

Parameter	Type	Description
text	String	Text message sent to the agent. Example: "What is this?"

ARTCAICallVisionCustomCaptureRequest

The request object for custom frame capture by a multimodal large language model.

Parameter	Type	Description
text	String	Text parameter for the multimodal large language model request
enableASR	boolean	Include ASR results from user speech as input to the large language model. If true, use ASR results and captured frames. If false, use the text field and captured frames Note Applies only during continuous frame capture
isSingle	boolean	When set to true, initiates a single-frame capture and immediately exits the custom frame-capture state. A value of false enables continuous frame capturing, which automatically stops after the specified duration.
eachDuration	int	Frame capture interval in seconds
num	int	Number of images captured per frame
duration	int	Duration for continuous frame capture in seconds. Ignored if isSingle=true. Custom frame capture ends when this duration expires
userData	String	JSON string containing custom business information. Sent to the large language model along with text and captured frames for enterprise processing

ARTCAICallVideoCanvas

The video rendering configuration object.

Parameter	Type	Description
renderMode	ARTCAICallVideoRenderMode	Rendering mode. Default: ARTCAICallVideoRenderModeAuto
mirrorMode	ARTCAICallVideoRenderMirrorMode	Mirror mode. Default: ARTCAICallVideoRenderMirrorModeOnlyFront
rotationMode	ARTCAICallVideoRotationMode	Rotation angle. Default: ARTCAICallVideoRotationMode_0
zOrderOnTop	boolean	Set whether SurfaceView should appear above all other windows. Default: true
zOrderMediaOverlay	boolean	Set whether SurfaceView should appear above MediaPlayer, Camera, and similar windows. Default: true

ARTCAICallAgentConfig

The configurable parameters for the call.

Parameter	Type	Description
agentGreeting	String	Agent greeting. Empty uses the agent's configuration value. Maximum length: 100 characters
wakeUpQuery	String	Instruction spoken by the user before the call starts. The agent responds immediately after the call starts
agentMaxIdleTime	int	Maximum idle time in seconds before the agent goes offline automatically. Default: 600 seconds
userOnlineTimeout	int	Time in seconds before the agent closes the task if the user does not join the call. Default: 60 seconds
userOfflineTimeout	int	Time in seconds before the agent closes the task after the user leaves the call. Default: 5 seconds
enablePushToTalk	boolean	Enable push-to-talk mode
agentGracefulShutdown	boolean	Enable graceful shutdown: finish speaking the current sentence before stopping
volume	int	Agent speaking volume. Range: 0–400. Default: 100
workflowOverrideParams	String	Workflow override parameters
enableIntelligentSegment	boolean	Enable intelligent sentence segmentation
asrConfig	ARTCAICallAgentAsrConfig	Speech recognition configuration
ttsConfig	ARTCAICallAgentTtsConfig	Speech synthesis configuration
llmConfig	ARTCAICallAgentLlmConfig	Large Language Model (LLM) configuration
avatarConfig	ARTCAICallAgentAvatarConfig	Digital human configuration
interruptConfig	ARTCAICallAgentInterruptConfig	Interrupt configuration
voiceprintConfig	ARTCAICallAgentVoiceprintConfig	Voiceprint denoising configuration
turnDetectionConfig	ARTCAICallAgentTurnDetectionConfig	Turn detection configuration
experimentalConfig	ARTCAICallExperimentalConfig	Non-production customization configuration
ARTCAICallAgentVcrConfig	ARTCAICallAgentVcrConfig	VCR configuration
preConnectAudioUrl	String	Audio effect played after connection and before the greeting. Supports URL input. Greeting still plays after the audio effect
ambientConfig	ARTCAICallAgentAmbientConfig	Ambient sound configuration
backChannelingConfig	List<ARTCAICallAgentBackChanneling>	Configuration module for acknowledgment speech. After configuration, the system randomly speaks short acknowledgments at specific trigger times
autoSpeechForLlmPendingConfig	ARTCAICallAgentAutoSpeechLlmPending	Auto-speech configuration for the agent when LLM response is delayed
autoSpeechForUserIdleConfig	ARTCAICallAgentAutoSpeechUserIdle	Configuration for agent queries when the user is silent

ARTCAICallAgentAsrConfig

The speech recognition configuration.

Parameter	Type	Description
asrLanguageId	String	ASR language ID. Empty uses the agent's configuration value
asrMaxSilence	int	Speech segmentation threshold. Speech ends if silence exceeds this duration. Default: 400 ms. Range: 200–1200 ms
asrHotWords	List<String>	ASR hot word list. Limit: upto 500 words. Each word: up to 10 characters
vadLevel	int	VAD sensitivity setting for AIVAD. Default: 11. Range: [0–11] 0 disables VAD. 1–10: Higher numbers make interruption harder. 11 differs significantly from previous versions, reducing voice distortion during preprocessing and providing greater resistance to interference.
customParams	String	Runtime parameters for custom ASR integration. Use URL parameter format. Example: "mode=fast&sample=16000&format=wav"
vadDuration	int	Minimum duration threshold for voice activity detection. Controls interruption sensitivity. Default: 0 (disabled). Valid range: 200–2000 ms. Common values: [200, 500] (1–4 characters)
asrMaxSilence	Int32	Speech segmentation threshold. Speech ends if silence exceeds this duration. Range: 200–1200 ms. Default: -1 (uses agent's default configuration from the console)

ARTCAICallAgentTtsConfig

The speech synthesis configuration.

Parameter	Type	Description
agentVoiceId	String	Agent voice ID. Empty uses the agent's configuration value
pronunciationRules	List	Pronunciation rule array. Supports up to 20 rules. If nil or empty, no rules apply. Example: `[ { "Word": "overlap", // Target word "Pronunciation": "chǒng dié", // Replacement pronunciation "Type": "replacement" // Homograph rule }, { "Word": "action", "Pronunciation": "háng dòng", "Type": "replacement" } ]`
speechRate	double	TTS speech rate. Applies to all TTS types. Range: [0.5, 2.0]. Default: 1.0. Values less than 0 use console configuration
languageId	String	TTS language code. Valid only for MiniMax TTS type
emotion	String	TTS emotion type. Valid only for MiniMax TTS type
modelId	String	TTS model ID. Currently supports only minimax. Options: speech-01-turbo / speech-02-turbo
speechRate	Double	TTS speech rate. Applies to all TTS types. Range: [0.5, 2.0]. Default: -1. Values less than 0 use agent's default configuration (console configuration)

ARTCAICallAgentLlmConfig

The Large Language Model (LLM) configuration.

Parameter	Type	Description
llmHistoryLimit	int	Maximum number of LLM/multimodal LLM conversation history turns to retain. Default: -1. Values less than 0 use agent's default configuration (console configuration)
llmSystemPrompt	String	LLM system prompt
bailianAppParams	String	Bailian application center parameters
llmCompleteReply	boolean	Send complete LLM results Note When enabled, the full LLM result is returned via the onLLMReplyCompleted event callback after LLM generation completes
openAIExtraQuery	String	Additional OpenAI protocol LLM query parameters Note Parameters must use key=value format. Separate multiple parameters with &. All values must be strings
outputMinLength	int	Minimum text output length in characters. Text shorter than this is cached for concatenation. Range: [0, 100]. Values ≤ 0 mean no limit. Default: no limit
outputMaxDelay	int	Maximum text output delay in milliseconds. Force output cached text after this time. Range: [1000, 10000]. Values ≤ 0 mean no limit. Default: no limit
historySyncWithTTS	boolean	Sync large model message history with TTS playback content. Default: false. When enabled, message history matches TTS playback content with minor tolerance Note When the user interrupts the agent, the next message sent to the large model inserts the `<ims_agent_interrupted>` tag at the interruption point. Example: `[ {"role": "user", "content": "Tell me a story."}, {"role": "assistant", "content": "Sure, I'll tell you a story from Romance of the Three Kingdoms. Do you<ims_agent_interrupted> want to hear it?"}, {"role": "user", "content": "Let's pick another one."} ]`

ARTCAICallAgentAvatarConfig

The digital human configuration.

Parameter	Type	Meaning
agentAvatarId	String	Digital human model ID. Empty uses the agent's configuration value

ARTCAICallAgentInterruptConfig

The interrupt configuration.

Parameter	Type	Description
enableVoiceInterrupt	boolean	Enable intelligent interruption
interruptWords	List<String>	Specific words or phrases that trigger conversation interruption
noInterruptMode	String	ASR text handling strategy when the agent speaks and intelligent interruption is disabled. Options: cache: Cache ASR text. Process it after the current turn ends discard: Discard ASR text immediately Other values (including empty): Use server default configuration

ARTCAICallAgentVoiceprintConfig

The voiceprint denoising configuration.

Parameter	Type	Description
useVoiceprint	boolean	Is voiceprint denoising detection used in current sentence segmentation?
voiceprintId	String	Voiceprint ID. Non-empty means voiceprint denoising is enabled

ARTCAICallAgentTurnDetectionConfig

The turn detection configuration.

Parameter	Type	Description
turnEndWords	List<String>	Words that signal speech end. Examples: "Done", "I'm finished"
mode	ARTCAICallTurnDetectionMode	Method to detect when user speech ends. Default: semantic mode. Uses AI to analyze context and semantics to detect speech end. Default value: ARTCAICallTurnDetectionSemanticMode
semanticWaitDuration	int	Custom wait time for semantic speech end detection in milliseconds. Range: [0, 10000]. Default: -1 (AI selects optimal wait time) Note Ignored if mode is ARTCAICallTurnDetectionNormalMode
eagerness	[String]?	Applies only when `mode = "Semantic"`. Overrides `SemanticWaitDuration`. Controls how quickly AI responds after detecting user pause: Low: Wait up to 6 seconds. Reduces false interruptions Medium: Wait up to 4 seconds. Balanced for most scenarios High: Wait up to 2 seconds. Faster interaction but higher false interruption risk Other values (including empty): Use server default configuration

ARTCAICallAgentVcrResult

The VCR detection result.

Parameter	Type	Description
resultData	Object	All VCR detection results returned by the agent
stillFrameMotionResult	FrameMotionResult	VCR still frame detection results
invalidFrameMotionResult	FrameMotionResult	VCR Invalid Frame Detection Results
peopleCountResult	PeopleCountResult	VCR real-time people count detection result
equipmentResult	EquipmentResult	VCR electronic device detection result
headMotionResult	HeadMotionResult	VCR head motion detection result
lookAwayResult	LookAwayResult	VCR gaze deviation detection result

LookAwayResult

The VCR gaze deviation detection result.

Property Name	Type	Description
count	int	Total number of gaze deviations detected up to the current frame
duration	int	Total duration of gaze deviation up to the current frame in milliseconds

FrameMotionResult

The VCR video frame detection result.

Parameter	Type	Meaning
duration	int	Time elapsed since sending this frame in milliseconds

PeopleCountResult

The VCR people count detection result.

Parameter	Type	Description
count	int	Number of people detected by VCR

EquipmentResult

The VCR electronic device detection result.

Parameter	Type	Description
mobilePhoneCount	int	Total number of devices
watchCount	int	Number of watches
headPhoneCount	int	Number of headphones

HeadMotionResult

The VCR head motion detection result.

Parameter	Type	Description
nodDetected	boolean	OK
shakeDetected	boolean	Shake your head

ARTCAICallAgentVcrConfig

The VCR configuration.

Parameter	Type	Description
data	JSONObject	JSON object passed by the user. Cached for later use when generating JSON strings. Enables custom extension
stillFrameMotion	ARTCAICallAgentVcrFrameMotionConfig	VCR Silent Frame Detection Configuration
invalidFrameMotion	ARTCAICallAgentVcrFrameMotionConfig	VCR invalid frame detection configuration
peopleCount	ARTCAICallAgentVcrBaseConfig	VCR real-time people count detection configuration
equipment	ARTCAICallAgentVcrBaseConfig	VCR electronic device detection configuration
headMotion	ARTCAICallAgentVcrBaseConfig	VCR head motion detection configuration
lookAway	ARTCAICallAgentVcrBaseConfig	VCR gaze deviation detection configuration

ARTCAICallAgentVcrBaseConfig

The VCR base detection configuration.

Parameter	Type	Description
enable	boolean	Do you want to enable this?

ARTCAICallAgentVcrFrameMotionConfig

The VCR video frame detection configuration.

Parameter	Type	Description
callbackDelay	int	Delay in milliseconds before triggering the callback

ARTCAICallExperimentalConfig

The experimental parameters for controlling specific logic policies.

Parameter	Type	Description
rtcSdkParams	JSONObject	RTC SDK parameters
commonParams	JSONObject	Common parameters

ARTCAICallAgentAmbientConfig

The call environment parameters.

Property Name	Type	Description
volume	int	Ambient sound volume. Default: 100
resourceId	String	Resource ID registered for ambient sound in the console. Empty string disables ambient sound

ARTCAICallAgentAutoSpeechContent

The agent speech content for auto-speech scenarios, such as acknowledgments or proactive questions.

Property Name	Type	Description
probability	Double	Trigger probability. Range: 0.0–1.0
text	String	Prompt text. UTF-8 encoded. Example: "Are you still there?" Maximum length: 20 characters for acknowledgment speech. Maximum length: 100 characters for auto-reply scenarios

ARTCAICallAgentAutoSpeechLlmPending

The auto-speech configuration for the agent when the LLM response is delayed.

Property Name	Type	Description
waitTime	int	Wait threshold in milliseconds. Trigger prompt after exceeding this duration. Range: 500–10000 ms. Required
messages	List<ARTCAICallAgentAutoSpeechContent>	Collection of waiting prompts. Maximum: 10 items. Each item ≤ 100 characters. Total probability must equal 1.0

ARTCAICallAgentAutoSpeechUserIdle

The configuration for agent queries when the user is silent.

Property Name	Type	Description
waitTime	int	Silence threshold in milliseconds. Trigger query after exceeding this duration. Range: 5000–600000 ms. Recommended: 10000 ms
messages	List<ARTCAICallAgentAutoSpeechContent>	Collection of waiting prompts. Maximum: 10 items. Each item ≤ 100 characters. Total probability must equal 1.0
maxRepeats	int	Maximum number of queries. Range: 0–10. Recommended: 5. After exceeding this, stop querying and end the call

ARTCAICallAgentBackChanneling

The configuration module for acknowledgment speech.

Property Name	Type	Description
enable	boolean	Is the Echo Feature enabled?
triggerStage	String	Timing of Associated Triggers
probability	double	Trigger probability. Range: 0.0–1.0
words	List<ARTCAICallAgentAutoSpeechContent>	Collection of acknowledgment speech content. Maximum: 10 items. Each item ≤ 20 characters. Total probability must equal 1.0

IARTCAICallService Details

generateAIAgentShareCall

Requests to start a shared agent call.

/**
 * Request to start a shared agent call
 * @param userId User ID of the currently logged-in user
 * @param aiAgentId Agent ID
 * @param aiAgentType Agent type
 * @param artcaiCallConfig Agent configuration
 * @param callback Callback for the request
 */
void generateAIAgentShareCall(String userId, String aiAgentId, ARTCAICallEngine.ARTCAICallAgentType aiAgentType, ARTCAICallEngine.ARTCAICallConfig artcaiCallConfig, IARTCAICallServiceCallback callback);

ARTCAIAgentUtil Details

parseAiAgentShareInfo

Parses shared agent information.

/**
 * Parse shared agent information
 * @param shareInfoText Shared agent information text
 * @return Structured configuration for the shared agent
 */
public static ARTCAIAgentShareInfo parseAiAgentShareInfo(String shareInfoText);

parseAiAgentInfo

Parses agent startup response information.

/**
 * Parse agent startup response information
 * @param jsonObject Agent startup response information
 * @return Structured information for the agent startup response
 */
public static ARTCAIAgentInfo parseAiAgentInfo(JSONObject jsonObject);