Skip to main content

MCP

πŸ’‘ Tip: This feature is only supported in version V4.1 and above

What is MCP​

MCP (Model Control Plane) is the central module in AI Studio used to connect AI models with external systems.
Its role is to help Agents "access more capabilities," enabling intelligent agents not only to answer questions but also to truly execute tasks, access systems, and call data.
It can be simply understood as: MCP is the key component that transforms AI from "being able to talk" to "being able to do things."

Functional Modules​

Function CategoryMain Purpose
Computing ServicesAddition, subtraction, multiplication, division, modulo, square root, exponentiation
Search ServicesWeb search, academic paper retrieval, data extraction
Data AnalysisChart generation, analytical capabilities
Tool IntegrationMap services, collaboration tool integration

Core Features of MCP​

  1. Unified Connection: Centrally manages models from different sources and external APIs, allowing AI to flexibly utilize multiple intelligent resources.
  2. Permissions and Security: Supports organization-level and individual-level access control to ensure enterprise data security and compliant invocation.
  3. Flexible Expansion: Freely add new models or system interfaces according to business needs, expanding AI capabilities at any time.
  4. Seamless Integration: Closely collaborates with AI Studio’s Agent, Skill, and App modules to easily realize the complete process from model invocation to task execution.

Analogous Example​

You can think of MCP as the "plugin center" or "control room" for AI:

  • When you install a new app on your phone, it gains new functions;
  • Similarly, when an Agent connects to a new MCP, it gains new capabilities.

For example:

  • Connecting to a "Weather Query MCP" enables the Agent to answer weather-related questions;
  • Connecting to a "CRM System MCP" enables the Agent to query customer information and generate sales reports;
  • Connecting to a "Document Analysis MCP" enables the Agent to help extract contract content and key data.

Application Scenarios​

MCP has a wide range of applications, common business scenarios include:

  • Automated Data Processing: Allow Agents to automatically extract data from systems, generate reports, or complete approvals.
  • Intelligent Q&A and Information Retrieval: Connect to enterprise knowledge bases or third-party information sources for precise responses.
  • Cross-System Collaboration: Enable AI to interact with multiple business systems simultaneously to complete complex tasks (such as ticket creation, customer notifications, etc.).
  • Personalized Expansion: Individual users can connect their own tools or APIs through MCP to create a personalized AI assistant.

With MCP, your Agent is no longer just a "question-answering assistant,"
but an intelligent work partner that truly understands business, executes tasks, and connects systems.