RuoYi-Vue-Pro Architecture Overview
AI Large Model

Core Positioning

The AI module is a one-stop AI capability integration platform, connecting to mainstream large models both domestically and internationally, providing 8 major AI capabilities including chat, drawing, knowledge base (RAG), tool calling, workflow, writing, mind mapping, and music creation. Whether you want to add an AI assistant to your system or build an AI-native application, ready-to-use capabilities are all here.

In one sentence: Make your system “able to chat, draw, write articles, and sing.”

Problems Solved

Pain PointHow the AI Module Solves It
Integrating large models requires individual API integrationUnified integration of mainstream domestic and international models, one-time integration
Internal enterprise knowledge cannot be utilized by AIRAG knowledge base lets AI understand your documents
Single AI capability (only chatting)8 major capability modules covering various scenarios
Need to orchestrate AI workflowsIntegrated with Dify/FastGPT/Coze workflow platforms
Complex switching between domestic and international modelsUnified API layer, switch models without changing code

User Roles

graph LR subgraph Roles AI_ADMIN["AI Administrator
Configure model access
Manage knowledge bases
Set permissions"] AI_USER["AI User
Chat conversations
Drawing creation
Knowledge base Q&A
Writing/Mind mapping"] AI_DEV["Developer
Integrate AI capabilities via API
Create custom workflows"] end

8 Major Functional Modules

graph TB AI_CORE["AI Large Model Platform"] AI_CORE --> CHAT["💬 AI Chat
Multi-model conversation
Context memory
Internet search
Thinking reasoning mode"] AI_CORE --> IMAGE["🎨 AI Drawing
Text-to-image
Supports Stable Diffusion
Midjourney"] AI_CORE --> RAG["📚 AI Knowledge Base
RAG retrieval-augmented generation
Upload documents for Q&A"] AI_CORE --> TOOL["🔧 AI Tool Calling
Function Calling
MCP Protocol"] AI_CORE --> WORKFLOW["🔄 AI Workflow
Integrates Dify/FastGPT/Coze
Visual orchestration"] AI_CORE --> WRITE["✍️ AI Writing
Articles/Reports/Emails
Multi-scenario templates"] AI_CORE --> MINDMAP["🧠 AI Mind Mapping
Auto-generate mind maps"] AI_CORE --> MUSIC["🎵 AI Music
Suno AI music creation"]

Supported Models

Domestic Models

ModelProviderFeatures
Tongyi QianwenAlibaba CloudStrong comprehensive capability
DeepSeekDeepSeekHigh cost-effectiveness, strong reasoning
DoubaoByteDanceMultimodal capability
HunyuanTencentGood ecosystem integration
Wenxin YiyanBaiduStrong Chinese comprehension
GLMZhipu AIOpen-source friendly
SparkiFlytekOutstanding voice capability
MoonshotMoonshot AILong context
MiniMaxMiniMaxMultimodal
BaiChuanBaichuan IntelligenceChinese optimized
SiliconFlowSiliconFlowMulti-model aggregation

International Models

ModelProvider
GPTOpenAI
ClaudeAnthropic
LlamaMeta
GeminiGoogle

Core Capability Details

AI Chat

User: Help me analyze this quarter's sales data
  ↓
AI: Based on the data, Q3 sales increased 15% year-over-year, including...
  ↓ (Supports Thinking reasoning mode)
Displays complete reasoning process + conclusions

Supports internet search, enabling AI to access real-time information.

AI Knowledge Base (RAG)

flowchart LR DOC["Upload Documents
PDF/Word/Web Pages"] --> PARSE["Document Parsing
Segmentation/Vectorization"] PARSE --> STORE["Vector Storage"] QUERY["User Question"] --> SEARCH["Retrieve Relevant Passages"] STORE --> SEARCH SEARCH --> LLM["Large Model
Answer with context"] LLM --> ANSWER["Accurate Answer"]

AI Workflow

Integrates Dify, FastGPT, Coze, and other platforms, enabling visual orchestration of complex AI workflows, such as:

  • Intelligent Customer Service: User question → Intent recognition → Knowledge base retrieval → Escalation decision
  • Content Generation: Keywords → Outline → Body → Image generation → Publishing
docs