GetTask - Get the result of an offline meeting minutes task

Prerequisites

Activate the service and obtain and configure an API key. To prevent security risks from code leaks, configure the API key as an environment variable instead of hard coding it in your code.

To grant temporary access permissions to third-party applications or users, or want to strictly control risky operations such as accessing or deleting sensitive data, you can use a temporary authentication token.Compared to a long-term API key, a temporary authentication token is valid for only a short period (60 seconds) and offers higher security. It is suitable for temporary call scenarios and can effectively reduce the risk of API key leaks.Usage: In your code, replace the API key used for authentication with the obtained temporary authentication token.

Endpoint

https://dashscope.aliyuncs.com/api/v1/services/aigc/multimodal-generation/generation

Request method

POST request

Request headers

Authorization: Bearer {api-key} // Replace {api-key} with your actual API key.
Content-Type: application/json

Request parameters

Name	Type	Required	Description	Sample value
model	string	Yes	The business type. Set the value to `tingwu-meeting`.	`tingwu-meeting`
input	object	Yes	Specify the relevant business parameters.
input.task	string	Yes	The task type. Set the value to `getTask` to get the task result.	getTask
input.dataId	string	Yes	The ID of the offline meeting minutes task to query.	wkb***

Response parameters

Name	Type	Description	Example/Parsing protocol
code	string	The error code.
message	string	The error message.
request_id	string	The request ID.	f97ee37d-0f9c-9b93-b6bf-bd263a232bf9
usage	object	Usage. This parameter is currently empty.	{}
output	object	The business response body.
output.status	int	The task status: 0: Succeeded 1: In progress 2: Failed	0
output.errorCode	string	The error code returned when the task is submitted but fails to run. This is returned when `status` is 2.	TSC.FileError
output.errorMessage	string	The error message returned when the task is submitted but fails to run. This is returned when `status` is 2.	File cannot be read.
output.playbackUrl	string	The URL for audio playback. The URL is valid for 24 hours.	For information about the parsing protocols for the content in the files at the Object Storage Service (OSS) URLs, see: Parsing protocol for output.transcriptionPath Parsing protocol for output.translationsPath Parsing protocol for output.autoChaptersPath Parsing protocol for output.summarizationPath Parsing protocol for output.meetingAssistancePath Parsing protocol for output.pptExtractionPath Parsing protocol for output.textPolishPath Parsing protocol for output.customPromptPath Note: If the URL string is null, the subtask failed. Otherwise, the subtask succeeded.
output.transcriptionPath	string	The OSS URL of the transcription result object.
output.translationsPath	string	The OSS URL of the translation.
output.autoChaptersPath	string	The OSS URL of the chapter overview.
output.summarizationPath	string	The OSS URL of the full-text summary, speaker summary, Q&A review, and mind map.
output.meetingAssistancePath	string	The OSS URL of the action items and keywords.
output.pptExtractionPath	string	The OSS URL of the extracted PPT and summary.
output.textPolishPath	string	The OSS URL of the polished text.
output.customPromptPath	string	The OSS URL of the custom prompt.

Parsing protocol for output.transcriptionPath

The outermost layer is an Object. Its fields are as follows:

Name	Type	Description	Example
audioInfo	object	The audio information object.
audioInfo.Size	long	The audio size in bytes.	3335756
audioInfo.Duration	long	The audio duration in milliseconds. For real-time speech recognition, this field does not represent the actual audio duration.	104006
audioInfo.SampleRate	int	The audio sampling rate.	16000
audioInfo.Language	string	The audio language.	cn
audioSegments	array<array<int>>	The range of valid audio segments.
audioSegments[i][0]	int	The start time of the valid audio segment in milliseconds.	1690
audioSegments[i][1]	int	The end time of the valid audio segment in milliseconds.	10704
paragraphs	array	A list of paragraphs.
paragraphs[0].paragraphId	String	Paragraph ID	16987422100275*****
paragraphs[0].speakerId	String	The speaker ID.	1
paragraphs[0].words	array	A list of words.
paragraphs[0].words[0].id	int	The word ID.	10
paragraphs[0].words[0].sentenceId	int	The sentence ID.	1
paragraphs[0].words[0].start	int	The start time in milliseconds.	4970
paragraphs[0].words[0].end	int	The end time in milliseconds.	5560
paragraphs[0].words[0].text	String	The word text.	Hello,

Example JSON:

{
  "audioInfo": {
    "duration": 300000,
    "language": "cn",
    "sampleRate": 16000,
    "size": 9605546
  },
  "audioSegments": [
    [
      2960,
      6908
    ],
    [
      7310,
      12879
    ]
  ],
  "paragraphs": [
    {
      "paragraphId": "1758265418316500000",
      "speakerId": "1",
      "words": [
        {
          "end": 3895,
          "id": 10,
          "sentenceId": 1,
          "start": 2960,
          "text": "At this moment"
        },
        {
          "end": 4030,
          "id": 20,
          "sentenceId": 1,
          "start": 3895,
          "text": "I"
        }
      ]
    }
  ]
}

Parsing protocol for output.translationsPath

The outermost layer is an Object with the following fields:

Parameter	Type	Description
paragraphs	list[]	A collection of translation results organized by paragraph. This corresponds to the speech recognition result message.
paragraphs.paragraphId	string	The paragraph ID. This corresponds to the ParagraphId in the speech recognition result.
paragraphs.sentences	list[]	A collection of translated sentences.
paragraphs.sentences[i].sentenceId	long	The sentence ID.
paragraphs.sentences[i].start	long	The start time of the segment relative to the beginning of the audio. This is a relative timestamp in milliseconds.
paragraphs.sentences[i].end	long	The end time of the segment relative to the beginning of the audio. This is a relative timestamp in milliseconds.
paragraphs.sentences[i].text	string	The translated text. This corresponds to the speech recognition result message.

Example JSON:

{
  "paragraphs": [
    {
      "paragraphId": "1758274598863500000",
      "sentences": [
        {
          "end": 6908,
          "sentenceId": 1,
          "start": 2960,
          "text": "At the moment, I will systematically tell you some things before we leave. "
        },
        {
          "end": 12879,
          "sentenceId": 2,
          "start": 7310,
          "text": "First of all, our car is equipped with this kind of remote control key, which can be directly controlled in APP. "
        },
        {
          "end": 14576,
          "sentenceId": 3,
          "start": 13850,
          "text": "This has two advantages. "
        },
        {
          "end": 22510,
          "sentenceId": 4,
          "start": 14580,
          "text": "First, if you are out of town and your car is now, for example, on the side of the road, you want to park your car in your garage, but you are not there, what should you do at this time? "
        },
        {
          "end": 36466,
          "sentenceId": 5,
          "start": 22820,
          "text": "Sharing one is to use this APP to share a key and remote key with your friend or driver. Let him tilt in the car prohibited by this APP. Just park the car and let him sleep again. It is to give him a temporary key. This is the first advantage. "
        },
        {
          "end": 45255,
          "sentenceId": 6,
          "start": 36740,
          "text": "The second advantage is that you can see the car locks and windows. If you are not sure whether the car doors are closed or not, you can watch and use them at home. "
        },
        {
          "end": 48408,
          "sentenceId": 7,
          "start": 47210,
          "text": "No, as long as there's a match for Wang. "
        }
      ]
    }
  ]
}

Parsing protocol for output.autoChaptersPath

The outermost layer is an array with the following fields:

Parameter	Type	Description
[i].id	int	The ordinal number of the chapter.
[i].start	long	The start time of the chapter relative to the beginning of the audio. This is a relative timestamp in milliseconds.
[i].end	long	The end time of the chapter relative to the beginning of the audio. This is a relative timestamp in milliseconds.
[i].headline	string	A one-sentence title for the chapter.
[i].summary	string	The chapter summary.

Example JSON:

[
    {
        "end": 245450,
        "headline": "Smart Car Remote Control and Parking Features",
        "id": 1,
        "start": 2960,
        "summary": "Introduces the remote control features of a smart car, including sharing temporary keys through an app, remotely controlling car locks and windows, starting the vehicle, and assistance features for automatic parking and maneuvering into and out of parking spaces. It also mentions the in-car voice control system, which allows both front and rear passengers to issue commands without a wake-word."
    },
    {
        "end": 299930,
        "headline": "Analysis of City NGP and Driving Assistance Features",
        "id": 2,
        "start": 248440,
        "summary": "Discusses the conditions for using City NGP, noting that it can be tried on enabled road sections. On sections where it is not enabled, only basic assistance features like ACC are available. It highlights the dangers of manual driving in poor lighting conditions and the behavioral differences of the NGP system with and without support for specific road sections."
    }
]

Parsing protocol for output.summarizationPath

The root element is an object that contains the following fields:

Parameter	Type	Description
paragraphSummary	string	The full-text summary result. The at sign and hash (`@#`) symbol separates the summary into two parts: Before `@#`: Begin with a brief summary of the article. Use `@#` to render the key points as an unordered list in Markdown format.
conversationalSummary	list[]	A list of speaker summary results.
conversationalSummary[i].speakerId	string	The speaker ID.
conversationalSummary[i].speakerName	string	The speaker name.
conversationalSummary[i].summary	string	The summary for the speaker.
questionsAnsweringSummary	list[]	A list of Q&A review summary results.
questionsAnsweringSummary[i].question	string	The question.
questionsAnsweringSummary[i].sentenceIdsOfQuestion	list[]	A list of SentenceIds from the original Automatic Speech Recognition (ASR) transcription that correspond to the question.
questionsAnsweringSummary[i].answer	string	The answer to the question.
questionsAnsweringSummary[i].sentenceIdsOfAnswer	list[]	A list of SentenceIds from the original ASR transcription that correspond to the answer.
mindMapSummary	list[]	A list of mind map results.
mindMapSummary[i].title	string	The text content of a single node in the mind map.
mindMapSummary[i].beginTime	int	The timestamp of a single mind map node relative to the start time, in milliseconds. This field exists only in leaf nodes.
mindMapSummary[i].topic	list[]	A list of child nodes for a single mind map node.

The following is a JSON example:

{
  "conversationalSummary": [
    {
      "speakerId": "2",
      "speakerName": "Speaker 2",
      "summary": "Discussed the transition between autonomous driving and manual control of the vehicle."
    },
    {
      "speakerId": "1",
      "speakerName": "Speaker 1",
      "summary": "Detailed the preparations before vehicle departure."
    }
  ],
  "mindMapSummary": [
    {
      "title": "Introduction to Vehicle Features Before Departure",
      "topic": [
        {
          "title": "Remote Key and App Control",
          "topic": [
            {
              "beginTime": 14580,
              "title": "Remotely control the vehicle, such as sharing a temporary key with a friend or valet to park in the garage",
              "topic": []
            },
            {
              "beginTime": 36740,
              "title": "Check the door lock status to ensure security",
              "topic": []
            }
          ]
        },
        {
          "title": "Remote Start and Horn Location",
          "topic": [
            {
              "beginTime": 54080,
              "title": "Locate the vehicle in an underground garage by sounding the horn",
              "topic": []
            }
          ]
        }
      ]
    }
  ],
  "paragraphSummary": "This record provides a detailed introduction and demonstration of the remote control and autonomous driving features of a smart car, including remote keys, parking space maneuvering, voice control, and the working principles of the City NGP driving assistance system.@#\n\n- **Introduction to Remote Control Features[2]**\n  1. **Remote Key Sharing[4]**\n  2. **Confirmation of Door and Window Status[6]**",
  "questionsAnsweringSummary": [
    {
      "answer": "This system has two benefits. First, if you are away and want a friend or valet to park your car in the garage, you can share the remote key from the app to let them park the car using the app. Second, you can check and control the car locks and windows from home using the app, and even perform functions like remote start and sounding the horn.",
      "question": "What are the benefits of this remote key system?",
      "sentenceIdsOfAnswer": [
        1,
        2,
        3,
        4,
        5,
        6,
        7
      ],
      "sentenceIdsOfQuestion": [
        1,
        2,
        3,
        4,
        5,
        6,
        7
      ]
    }
  ]
}

Parsing protocol for output.meetingAssistancePath

The root element is an object with the following fields:

Parameter	Type	Description
keywords	list[]	The extracted keywords.
actions	list[]	A collection of action items and action item summaries.
actions[i].id	long	The ordinal number of the action item.
actions[i].sentenceId	long	The ID of the corresponding sentence in the original ASR transcription.
actions[i].start	long	The start time relative to the beginning of the audio. This is a relative timestamp in milliseconds.
actions[i].end	long	The end time relative to the beginning of the audio. This is a relative timestamp in milliseconds.
actions[i].text	string	The content of the action item.
classifications	object	The scenario classification. Currently, there are three scenario categories.
classifications.interview	float	The confidence score for the interview scenario.
classifications.lecture	float	The confidence score for the lecture scenario.
classifications.meeting	float	The confidence score for the meeting scenario.

Example JSON:

{
  "actions": [
    {
      "end": 0,
      "id": 1,
      "sentenceId": 5,
      "start": 0,
      "text": "Provide feedback to the courier department about the damaged package and request repackaging or compensation."
    }
  ],
  "classifications": {
    "interview": 0.0,
    "lecture": 0.22630735,
    "meeting": 0.2158773
  },
  "keywords": [
    "Courier",
    "Package Damage"
  ]
}

Parsing protocol for output.pptExtractionPath

The root element is an object. Its fields are described below:

Parameter	Type	Description
availableForSummary	boolean	Indicates whether a PPT summary text exists. A value of false means no PPT summary result is available. A value of true means at least one is available.
keyFrameList	list[]	A list of PPT results, including PPT images and summary text.
keyFrameList[i].id	long	The ordinal number.
keyFrameList[i].start	long	The start time relative to the beginning of the audio. This is a relative timestamp in milliseconds.
keyFrameList[i].end	long	The end time relative to the beginning of the audio. This is a relative timestamp in milliseconds.
keyFrameList[i].fileUrl	string	The generated PPT image result. This is an HTTP download link.
keyFrameList[i].summary	string	The corresponding PPT summary.

The following is a JSON example.

{
    "availableForSummary": true,
    "keyFrameList": [
        {
            "end": 17470,
            "fileUrl": "https://prod-tingwu-paas-common-beijing.oss-cn-beijing.aliyuncs.com/***",
            "id": 1,
            "start": 30
        }
    ]
}

Parsing protocol for output.textPolishPath

The root element is an array that contains the following fields:

JSON field	Type	Description
[i].formalParagraphText	string	The polished text result, converted from spoken to written language.
[i].sentenceIds	list	A list of SentenceIds corresponding to the polished content.
[i].paragraphId	string	The paragraph ID of the text result. This corresponds to the paragraph ID in the speech recognition result.
[i].start	long	The start time of the text result in the original audio. This is a relative timestamp in milliseconds.
[i].end	long	The end time of the text result in the original audio. This is a relative timestamp in milliseconds.

Example JSON:

[
    {
        "end": 32392,
        "formalParagraphText": "Hello everyone, I am Li Zhe, the product manager for Tongyi Tingwu. Today, I will introduce the Tongyi Tingwu public cloud API. First, let's get an overview of Tingwu. Tingwu is a work and study assistant focused on audio and video content. It can accurately analyze, extract, and summarize audio and video content to help enterprise customers quickly grasp key information and improve content utilization efficiency. The built-in large model supports various functions such as summarization, quality inspection, and translation, and also has speech recognition and visual understanding capabilities.",
        "paragraphId": "1758527947000500001",
        "sentenceIds": [
            1,
            2,
            3,
            4,
            5
        ],
        "start": 30
    }
]

Parsing protocol for output.customPromptPath

The root element is an array that contains the following fields:

JSON field	Type	Description
[i].name	string	The dataId corresponding to the task.
[i].result	string	The result returned by the large model.
[i].truncated	boolean	Indicates whether the result was truncated.

Example JSON:

[
    {
        "name": "fgVnGvkqj5xA",
        "result": "This question seems to be missing some specific background information. For example, does 'how many people' refer to a specific scenario or the number of people in a specific group?",
        "truncated": false
    }
]

Request example

curl --location 'https://dashscope.aliyuncs.com/api/v1/services/aigc/multimodal-generation/generation' \
--header 'Authorization: Bearer sk-***' \
--header 'Content-Type: application/json' \
--data '{
  "model": "tingwu-meeting",
  "input": {
    "dataId": "***",
    "task": "getTask"
  }
}'

Response example

{
    "output": {
        "autoChaptersPath": "https://oss.example.com/...",
        "customPromptPath": "https://oss.example.com/...",
        "meetingAssistancePath": "https://oss.example.com/...",
        "playbackUrl": "https://oss.example.com/...",
        "pptExtractionPath": "https://oss.example.com/...",
        "status": 0,
        "summarizationPath": "https://oss.example.com/...",
        "textPolishPath": "https://oss.example.com/...",
        "transcriptionPath": "https://oss.example.com/...",
        "translationsPath": "https://oss.example.com/..."
    },
    "usage": {},
    "request_id": "ac5e2400-5aaf-44a6-8cf0-481bab4cefe3"
}

​Prerequisites

​Endpoint

​Request method

​Request headers

​Request parameters

​Response parameters

​Parsing protocol for output.transcriptionPath

​Parsing protocol for output.translationsPath

​Parsing protocol for output.autoChaptersPath

​Parsing protocol for output.summarizationPath

​Parsing protocol for output.meetingAssistancePath

​Parsing protocol for output.pptExtractionPath

​Parsing protocol for output.textPolishPath

​Parsing protocol for output.customPromptPath

​Request example

​Response example

Prerequisites

Endpoint

Request method

Request headers

Request parameters

Response parameters

Parsing protocol for output.transcriptionPath

Parsing protocol for output.translationsPath

Parsing protocol for output.autoChaptersPath

Parsing protocol for output.summarizationPath

Parsing protocol for output.meetingAssistancePath

Parsing protocol for output.pptExtractionPath

Parsing protocol for output.textPolishPath

Parsing protocol for output.customPromptPath

Request example

Response example