OpenMAIC
Multi-agent platform that turns a topic or document into an interactive AI classroom with slides and quizzes.
Slides, quizzes, simulations and project-based learning delivered by AI teachers and classmates.
OpenMAIC (Open Multi-Agent Interactive Classroom) turns a topic or uploaded material into a lesson with slides, quizzes, interactive HTML simulations and project-based learning. It suits educators and developers who can host a Node and PostgreSQL app and bring their own LLM keys.
What it is
You describe a topic or attach documents and it builds an outline and scenes, with AI teachers and AI classmates who lecture, discuss, use a whiteboard and speak via text-to-speech. Classrooms export to PPTX, a single offline HTML file or a ZIP, and models are configured per task in an openmaic.yml file covering OpenAI, Anthropic, Gemini, DeepSeek, Ollama and many others.
Who it's for
It suits teachers, course builders and ed-tech developers who want an editable, self-hosted classroom generator. It does not suit non-technical users who want a hosted product with support, since setup involves a database, environment files and API keys.
Key features
One-click lesson generation
Builds a lesson from a topic or attached materials.
Multi-agent classroom
AI teachers and peers lecture, discuss and answer in real time.
Varied scene types
Slides, quizzes, interactive HTML simulations and PBL scenes.
Whiteboard and speech
Agents draw diagrams, write formulas and speak with TTS.
Export options
Save as PPTX, offline HTML or a classroom ZIP.
Per-task model slots
Assign different providers and models to outline, content, TTS and search.
Up and running in 3 steps
Our honest verdict
What's great
- MIT licence permits commercial use, which the README states directly
- Supports a long provider list including local options such as Ollama, Lemonade and FunASR
- Server-side persistence means generation survives closed tabs, according to the changelog
- Agent skill integration with OpenClaw, Codex and messaging apps
Where it falls short
- Needs Node.js 22.19 or newer, pnpm 10 and PostgreSQL 16, or a Docker Compose stack of two containers
- You supply and pay for LLM and optional media API keys, and the README recommends turning on illustrations, narration and video for best results
- The Compose setup is single-user and listens on localhost only until you add your own protection
- It is changing fast with frequent releases, one-way upgrades and a release candidate, so plan for migration work
Pricing & license
Free under the MIT licence. The README shows no paid hosted plan; your costs are the server, the PostgreSQL database and whatever LLM, speech and media APIs you connect.
Open remote AI-training roles
873 open remote AI-training roles, updated November 15, 2026; most pay $20–$170 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
These reviewed AI tools do the same job instantly — from our scored directory.
Use it, or have us set it up
FAQ
Can I try it before installing?
Yes, the README links a live demo at open.maic.chat.
Is it free for commercial use?
The README says it is MIT-licensed, so commercial use is permitted free of charge.
Which AI models does it use?
You configure them. The README lists OpenAI, Anthropic, Gemini, DeepSeek, Qwen, Kimi, Ollama and any OpenAI-compatible API.
