OpenMAIC - Tsinghua University's open-source multi-agent AI classroom platform

OpenMAIC is an open source multi-agent AI classroom platform developed by the Tsinghua University team. It can transform any topic or document into an immersive interactive course with one click. The platform supports voice teaching by AI teachers, discussions by AI classmates raising their hands, real-time whiteboard drawing, and can generate a variety of teaching scenarios such as slides, tests, interactive simulations, and project-based learning. The platform has been connected OpenClaw, users can directly generate courses in chat applications such as Feishu, and the course preparation cost is only one thousandth of that of traditional MOOCs. The system is based on two years of real classroom data, allowing everyone to have a 24/7 AI classroom.

Main functions of OpenMAIC

• One-click generation : The platform supports inputting topics or uploading documents, and AI automatically generates a complete course including slides, quizzes, interactive simulations and project-based learning.
• Multi-agent classroom : AI teachers give voice lessons and operate whiteboards, and AI students take the initiative to ask questions and discuss, creating an immersive learning atmosphere.
• OpenClaw integration : Directly call the AI assistant to generate courses in chat applications such as Feishu and Slack without opening a web page.
• Flexible export : Courseware can be exported as editable PPTX or self-contained interactive HTML, supporting multiple languages and voice interaction.

Technical principles of OpenMAIC

• content generation layer : The platform adopts a two-stage pipeline architecture: in the first stage, the Plan agent analyzes user input and generates a structured course outline. In the second stage, each scenario agent converts outline nodes into specific teaching content in parallel, including slide layout, test question design, interactive web page code generation, etc.
• Classroom playback engine : Based on the Agentic architecture, the AI teacher acts as an autonomous decision-making agent and drives the explanation process through a state machine. It can determine in real time when to give voice explanations, when to draw a whiteboard, and when to initiate discussions, and completes teaching with visual effects such as TTS speech synthesis and spotlights.
• Multi-agent orchestration layer : Using LangGraph as the core framework, a Director Graph is constructed to coordinate multiple agent roles, manage the turn switching of scenarios such as teacher teaching, students raising hands, round table debates, etc., to ensure the orderliness and coherence of multi-Agent interactions.
• action execution engine : The platform encapsulates more than 28 kinds of atomic operations, covering voice broadcast, whiteboard drawing (supporting formulas, charts, flow charts), laser pointer pointing, scene switching, etc., providing a standardized classroom interaction capability interface for upper-level agents.

OpenMAIC project address

• Project official website ：https://open.maic.chat/
• GitHub repository ：https://github.com/THU-MAIC/OpenMAIC

Application scenarios of OpenMAIC

• higher education : College students can quickly generate final review courses, such as letting AI explain difficult knowledge points such as bubble sorting and Taylor’s formula, and professors can efficiently produce professional courseware such as “Introduction to Artificial Intelligence”.
• K12 Education : Generate a Mathematical Olympiad thinking training camp for primary school students to cultivate mathematical thinking through interactive games such as chicken and rabbit in the same cage and tree planting problems, or help younger students intuitively understand abstract scientific concepts such as general relativity.
• lifelong learning : The platform supports 30-minute quick Python programming for zero-level users, which can instantly analyze DeepSeek’s latest papers, interpret stock market trends and other personalized knowledge needs.
• Corporate training : The platform can quickly produce standardized courses such as product knowledge and compliance training, and new employees can learn through role-playing project systems to simulate real business scenarios. ©