- Product Overview
- Server API Integration
- Avatar aPaas API Calling Methods
- Avatar Image Customization and Voice Clone API Documentation - v 0.2.1
- Video Generation Service API Documentation - v 4.9.9
- Overview
- Digital Human aPaaS API Calling Methods
- Audio Production API
- Video Production API - Basic Edition
- Audio and Video Production Progress Query API
- Video Production API - Advanced Version
- Customer Resource Query Anchor API
- Querying All Images of a Specific Anchor
- Querying the Supported Timbres for VirtualmanKey (to Be Deprecated)
- Querying the Supported Actions for VirtualmanKey
- Appendix
- API Integration FAQs
- Interactive Digital Human Service API Documentation - v 5.0.3
- Overview
- Digital Human aPaaS API Calling Methods
- Live Session Management
- Creating a New Live Stream Session
- Using the Personal Asset Image to Create a Stream
- Querying the Session Status
- Starting Session
- Closing Live Session
- Querying the Conversation List under the Digital Human Project
- Querying the Conversion List under the Personal Asset Image
- Querying the Session List under UIN
- Instruction-driven
- Other APIs
- Error Code Description
- Personal Asset Management API Documentation - v 1.1.2
- Overview
- Digital Human aPaaS API Calling Methods
- Querying for Avatar List by Pagination API
- Querying Supported Timbres for Avatars (to be Deprecated)
- Querying Customer Service Asset Information
- Querying Timbre Lists by Pagination
- Querying Image Asset Information - Query Anchor
- Querying Image Asset Information - Querying all Avatars under the Anchor
- Querying the List Of Actions Supported by the Avatar
- Appendix 1 - Service Asset Type
- Appendix 2 - Emotional Style
- Appendix 3 - Digital Human Type
- API Integration FAQs
- FAQs
- Product Overview
- Server API Integration
- Avatar aPaas API Calling Methods
- Avatar Image Customization and Voice Clone API Documentation - v 0.2.1
- Video Generation Service API Documentation - v 4.9.9
- Overview
- Digital Human aPaaS API Calling Methods
- Audio Production API
- Video Production API - Basic Edition
- Audio and Video Production Progress Query API
- Video Production API - Advanced Version
- Customer Resource Query Anchor API
- Querying All Images of a Specific Anchor
- Querying the Supported Timbres for VirtualmanKey (to Be Deprecated)
- Querying the Supported Actions for VirtualmanKey
- Appendix
- API Integration FAQs
- Interactive Digital Human Service API Documentation - v 5.0.3
- Overview
- Digital Human aPaaS API Calling Methods
- Live Session Management
- Creating a New Live Stream Session
- Using the Personal Asset Image to Create a Stream
- Querying the Session Status
- Starting Session
- Closing Live Session
- Querying the Conversation List under the Digital Human Project
- Querying the Conversion List under the Personal Asset Image
- Querying the Session List under UIN
- Instruction-driven
- Other APIs
- Error Code Description
- Personal Asset Management API Documentation - v 1.1.2
- Overview
- Digital Human aPaaS API Calling Methods
- Querying for Avatar List by Pagination API
- Querying Supported Timbres for Avatars (to be Deprecated)
- Querying Customer Service Asset Information
- Querying Timbre Lists by Pagination
- Querying Image Asset Information - Query Anchor
- Querying Image Asset Information - Querying all Avatars under the Anchor
- Querying the List Of Actions Supported by the Avatar
- Appendix 1 - Service Asset Type
- Appendix 2 - Emotional Style
- Appendix 3 - Digital Human Type
- API Integration FAQs
- FAQs
After you Creating a Long Connection Channel, the digital human server will return the real-time digital human driving status information to the client.
Name | Type | Required | Description |
Type | int | Yes | Downstream message types. Other return types for Avatar can be ignored. 3: broadcast status. 4: content returned by the large model. 9: driving failed. |
SessionId | String | Yes | Unique identifier for the video stream session |
ReqId | String | Yes | Unique identifier for a single drive session |
Seq | int | Yes | Unique sequence number of the streaming clause |
SpeakStatus | String | No | Initial: initial status WaitingTextStart: waiting for text broadcast to start TextStart: text broadcast in progress WaitingTextOver: waiting for text broadcast to end TextOver: Text broadcast ended. WaitingAudioStart: waiting for audio broadcast to start AudioStart: Audio broadcast started. WaitingAudioOver: waiting for audio broadcast to end AudioOver: Audio broadcast ended. Error: Driving failed (similar to TextOver and AudioOver, indicating the end state of a drive session). This only indicates the most recent drive failure and does not affect the ability to continue sending drive commands. SentenceNext: In streaming clause mode, this status is returned, indicating to the client that it can send the next clause upon receiving this status.
SentenceStart: In streaming clause mode, this status is returned, indicating to the client that the current clause playback has started upon receiving this status.
SentenceOver: In streaming clause mode, this status is returned, indicating to the client that the current clause playback has ended upon receiving this status. |
FinalType | int | No | When SpeakStatus returns AudioOver, it has a value indicating the source of the Final packet that marks the end of audio playback： 1 - Client input; 2 - Server-generated packet due to timeout |
Text | String | No | This field is present when Type is 2 or 4, indicating a question prompt. |
TextPro | String | No | This field is present when the Type is 2 or 4, containing the broadcast text with SSML tags. |
TextDisplay | String | No | This field is present when Type is 2 or 4, containing the broadcast text for display on the client. |
ContentType | String | No | This field is present when the Type is 4, used to distinguish the content type returned by the large model:
Image: Image
Hyperlink: Hyperlink
OrderedList: Ordered list
UnOrderedList: Unordered list
List: Table |
TtsSupport | bool | No | This field is present when the Type is 4, indicating whether the content returned by the large model needs to be broadcast via tts. |
Final | bool | No | This field is present when the Type is 4, indicating whether the content returned by the large model is the final clause. |
IsHighLight | bool | No | This field is present when the Type is 4, indicating whether the content returned by the large model needs to be highlighted on the client side. |
Uninterrupt | bool | No | This field is present when the Type is 2 or 4, indicating whether the current broadcast sentence can be interrupted:
true: Cannot be interrupted
false: Can be interrupted |
Muted | bool | No | This field is present when the Type is 2 or 4, indicating whether the current broadcast text has the microphone muted. |
InteractionType | String | No | This field is present when the Type is 2, and can be used for platform customization. Common types include popup, image, etc. |
InteractionContent | String | No | This field is present when the Type is 2, used for special messages such as pop-ups, images, and other non-text content. |
ErrorCode | int | Yes | Error code, with a value of 0 indicating normal operation. A non-zero value indicates an error, such as an invalid request body, incorrect timing of drive commands, or excessive request rate. See Section 7 for the error code list. |
ErrorMessage | String | No | Error Description |
CloudAiExtra | String | No | Extended fields returned by Tencent's LLM Knowledge Engine. Response example:
{"record_id":"b24e6505-be78-4ce9-ad9e-7075dd227994","references":[{"id":"1780434945714421760","type":2,"url":"https://oaqbot.qidian.qq.com/#/preview/md?id=1780227052790087680","name":"Client meeting","doc_id":"11296"}]} |
Note:
When websocket reconnects, the cloud will proactively push the playback status triggered by the last three drive requests to prevent the client from losing corresponding drive feedback events due to network instability during the reconnection process.
Yes
No
Was this page helpful?