Implementation:Langgenius Dify CreateApp Workflow

Knowledge Sources	Dify
Domains	Workflow, DAG, Frontend
Last Updated	2026-02-12 00:00 GMT

Overview

Description

The createApp function is the frontend service method responsible for creating new applications in the Dify platform. When the mode parameter is set to AppModeEnum.WORKFLOW or AppModeEnum.ADVANCED_CHAT, the resulting application is a DAG-based workflow that can be visually composed in the Workflow Builder. The function issues a POST request to the /apps endpoint with the application metadata and mode discriminator.

Usage

To create a workflow application:

• Import createApp from web/service/apps.ts.
• Pass an object with name (required), mode set to AppModeEnum.WORKFLOW (for pure DAG) or AppModeEnum.ADVANCED_CHAT (for conversational DAG), and optional fields icon_type, icon, icon_background, description, and config.
• The function returns a Promise<AppDetailResponse> containing the created application's details including its unique id.

Code Reference

Source Location

web/service/apps.ts, lines 23-41.

Signature

export const createApp = ({
  name,
  icon_type,
  icon,
  icon_background,
  mode,
  description,
  config,
}: {
  name: string
  icon_type?: AppIconType
  icon?: string
  icon_background?: string
  mode: AppModeEnum
  description?: string
  config?: ModelConfig
}): Promise<AppDetailResponse> => {
  return post<AppDetailResponse>('apps', {
    body: { name, icon_type, icon, icon_background, mode, description, model_config: config },
  })
}

Import

import { createApp } from '@/service/apps'
import { AppModeEnum } from '@/types/app'

I/O Contract

Inputs

Parameter	Type	Required	Description
name	string	Yes	Display name for the new application
icon_type	AppIconType	No	Type of icon (emoji or image)
icon	string	No	Icon value (emoji character or image URL)
icon_background	string	No	Background color for the icon
mode	AppModeEnum	Yes	Application mode; use AppModeEnum.WORKFLOW ('workflow') for DAG pipelines or AppModeEnum.ADVANCED_CHAT ('advanced-chat') for conversational DAG workflows
description	string	No	Human-readable description of the application
config	ModelConfig	No	Optional model configuration for chat-based modes

Outputs

Field	Type	Description
(return)	Promise<AppDetailResponse>	Resolves to the full application detail object, including the generated id, name, mode, and other metadata

Usage Examples

import { createApp } from '@/service/apps'
import { AppModeEnum } from '@/types/app'

// Create a pure DAG workflow application
const workflowApp = await createApp({
  name: 'Document Processing Pipeline',
  mode: AppModeEnum.WORKFLOW,
  description: 'Extracts and transforms document content',
})

// Create a conversational DAG workflow application
const advancedChatApp = await createApp({
  name: 'Customer Support Bot',
  mode: AppModeEnum.ADVANCED_CHAT,
  description: 'Multi-step chat workflow with knowledge retrieval',
})

Related Pages

• Principle:Langgenius_Dify_Workflow_Application_Creation