The CreateTask API creates an offline meeting minutes task by using a POST request-Alibaba Cloud Model Studio(Model Studio)-阿里云帮助中心

API reference for creating 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.

NoteTo 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 URL

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

Request method

This operation uses the POST method.

Request headers

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

Create a task

Creates an offline meeting minutes task.

Request parameters

Name	Type	Required	Description	Example
model	string	Yes	Defines the business type. Set the value to `tingwu-meeting`.	tingwu-meeting
input	object	Yes	Pass the relevant business parameters
input.fileUrl	string	One of three	The OSS URL of the file to analyze.	https://.oss-cn-hangzhou.aliyuncs.com/%E8%AF%95%E9%A9%BE%E6%A1%88%E4%BE%8Bsmall.wav?OSSAccessKeyId=YOUR_ACCESS_KEY_ID&Expires=&Signature=YOUR_SIGNATURE
input.text	string		The text to analyze.	Each line must be in the following format: ${Speaker Name}: ${Speaker Content} Example: `Speaker1: I want to buy an SUV. Speaker2: What is your price range?`
input.dataId	string		Associates a task that has been uploaded and parsed.	j0oS1XbbT***
input.appId	string	Yes	The application ID.	tw_YrN4Ss9o5***
input.task	string	Yes	Defines the task type. Set the value to `createTask` to create a task.	createTask
input.type	string	Yes	The type of meeting minutes task to create. For an offline meeting minutes task, set the value to `offline`.	offline
input.phraseId	string	Yes	The ID of the hotword list. You can obtain this ID from the Hotwords page in the console. This parameter is required when you create a meeting task but is optional when you end the meeting and generate minutes.	z4p5EVXof***
parameters	object	No	The control parameters for the offline meeting minutes.
parameters.transcription	object	Yes	The parameters required for transcription.
parameters.transcription.model	string	Yes	The transcription model: `paraformer-v2` (Chinese and English): `cn` `paraformer-v2` (Chinese, English, Japanese, Korean, and Cantonese): `multilingual` `paraformer-v2` (English): `en` `paraformer-v2` (Cantonese): `yue` `paraformer-v2` (Japanese): `ja` `paraformer-v2` (Korean): `ko` `fun-asr (Chinese and English)` `Education realm model` (Chinese and English): `domain-education` `Automotive realm model` (Chinese, English, and Cantonese): `domain-automotive`	cn
parameters.transcription.languageHints	string[ ]	No	This parameter is valid only when `model` is set to `multilingual`. The selectable languages. Pass the selected languages from the multi-selection box: Chinese (cn), English (en), Cantonese (yue), Japanese (ja), Korean (ko), German (de), French (fr), and Russian (ru). Example: ["cn", "en", "yue", "ko"]	["cn"]
parameters.transcription.diarizationEnabled	boolean	Yes	Specifies whether to enable speaker diarization.	true
parameters.transcription.diarizationSpeakerCount	int	No	The number of speakers to separate when speaker diarization is enabled. `0`: The number of speakers is indefinite. `2`: The number of speakers is two. The default value is `0`.
parameters.transcription.translationEnabled	boolean	Yes	Specifies whether to enable translation.	true
parameters.transcription.translationTargetLang	string[ ]	No	If translation is enabled, set the target language: Chinese (cn), English (en), Japanese (ja), Korean (ko), German (de), French (fr), or Russian (ru). Only a single language is supported. Example: ["cn"]. Note that the value of this parameter must be different from the value of the `model` parameter. Otherwise, the task fails to be created.	["cn"]
parameters.audio	object	No	The parameters related to audio input.	?
parameters.audio.audioChannelMode	string	No	The multi-channel audio settings. "": Telephone recording or single-channel audio with a sample rate of 16 kHz or higher. mix: Multi-channel audio with a sample rate of 16 kHz or higher (recorded by an in-car device). If you do not specify this parameter, the default value "" is used.	""
parameters.analysis	object	Yes	The LLM analysis parameters.
parameters.analysis.model	string	Yes	The global analysis model: `Default` -> default `tingwu-plus` -> tingwu-plus `qwen-plus` -> qwen-plus `qwq` -> qwq	default
parameters.analysis.keyInformationEnabled	boolean	Yes	Specifies whether to enable keyword extraction.	true
parameters.analysis.actionsEnabled	boolean	Yes	Is the To-Do List enabled?	true
parameters.analysis.fullSummaryEnabled	boolean	Yes	Specifies whether to enable full text summarization.	true
parameters.analysis.fullSummaryFormat	string	No	The format of the full text summary: Paragraph summary: Default `markdown`: A summary in Markdown format.	default
parameters.analysis.conversationalEnabled	boolean	Yes	Specifies whether to enable speaker summaries.	true
parameters.analysis.questionsAnsweringEnabled	boolean	Yes	Enable review of Q&A pairs	true
parameters.analysis.mindMapEnabled	boolean	Yes	Specifies whether to enable mind map generation.	true
parameters.analysis.mindMapFormat	string	No	The format of the mind map: Timestamp: timestamp `plain`: Does not include timestamps.	timestamp
parameters.analysis.pptExtractionEnabled	boolean	Yes	Specifies whether to enable PPT extraction.	true
parameters.analysis.autoChaptersEnabled	boolean	Yes	Enable chapter preview?	true
parameters.analysis.autoChapterGranularity	string	No	The granularity of chapters. `Coarse`: Approximately 4 chapters per hour of audio or video. `General`: Approximately 6 chapters per hour of audio or video. `Meticulous`: Approximately 12 to 15 chapters per hour of audio or video.	General
parameters.analysis.autoChapterTitleLengthLevel	string	No	The length of chapter titles. `Short`: 6 to 25 characters, with an average of 13 characters. `Normal`: 10 to 28 characters, with an average of 16 characters. `Long`: 10 to 30 characters, with an average of 20 characters.	Short
parameters.analysis.textPolishEnabled	boolean	Yes	Specifies whether to convert spoken language to written language.	true
parameters.analysis.customPromptEnabled	boolean	Yes	Specifies whether to enable custom prompts.	true
parameters.analysis.customPromptModel	string	No	If custom prompts are enabled, select a model: tingwu-turbo tingwu-plus qwen-max	tingwu-turbo
parameters.analysis.customPromptTransType	string	No	The format of the text that replaces the `{Transcription}` placeholder in the prompt. The protocols are as follows: `default`: Transcription text only. Example: What is the weather like in Beijing? What is the weather like in Shanghai? `chat`: Speaker and transcription text. Example: Speaker 1:What is the weather like in Beijing?\nSpeaker 2:What is the weather like in Shanghai?\n `sentence-chat`: Sentence ID, speaker, and text. Example: [1] Speaker 1:What is the weather like in Beijing?\n[2] Speaker 2:What is the weather like in Shanghai?\n	default
parameters.analysis.customPromptContent	string	Yes	The custom prompt entered by the user. Note: Place the `{Transcription}` placeholder where the original text is needed in the prompt. When the prompt runs, the placeholder is replaced with the transcribed text in the format defined by `parameters.analysis.customPromptTransType`.	Please summarize the following content: {Transcription}

Response parameters

Name	Type	Description	Example
output	object
output.dataId	string	The task ID.	lsA***
usage	object	The usage. The return value is an empty object: `{}`.	{}
code	string	The error code.	InvalidParameter
message	string	The error message.	Agent Input text format error.
request_id	string	The request ID.	f97ee37d-0f9c-9b93-b6bf-bd263a232bf9

Request example

Replace sk-*** in the Authorization request header with your Alibaba Cloud Model Studio API key. For more information about how to obtain an API key, see Prerequisites.

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": {
        "task": "createTask",
        "type": "offline",
        "appId": "tw_AxlVSM**",
        "fileUrl": "",
        "phraseId": "z4p5EVXof***"
    },
    "parameters": {
        "transcription": {
            "model": "cn",
            "diarizationEnabled": true,
            "diarizationSpeakerCount": 0,
            "translationEnabled": true,
            "translationTargetLang": [
                "en"
            ]
        },
        "analysis": {
            "model": "default",
            "keyInformationEnabled": true,
            "actionsEnabled": true,
            "fullSummaryEnabled": true,
            "fullSummaryFormat": "markdown",
            "conversationalEnabled": true,
            "questionsAnsweringEnabled": true,
            "mindMapEnabled": true,
            "mindMapFormat": "timestamp",
            "pptExtractionEnabled": true,
            "autoChaptersEnabled": true,
            "autoChapterGranularity": "Coarse",
            "autoChapterTitleLengthLevel": "Short",
            "textPolishEnabled": true,
            "customPromptEnabled": true,
            "customPromptModel": "tingwu-turbo",
            "customPromptTransType": "default",
            "customPromptContent": "{Transcription}Guess how many people are in the preceding conversation."
        }
    }
}'

Response example

{
    "output": {
        "dataId": "z4p***"
    },
    "usage": {},
    "request_id": "1e92069c-e1e6-4a5a-97aa-19c3d880140d"
}