DeepTutor: open-source, agent-native personalized tutor trending again
HKUDS/DeepTutor, the University of Hong Kong's open-source personalized learning assistant, is trending on GitHub again, up about 1,600 stars in a day to roughly 47,200. It builds a persistent learner profile across sessions, grounds teaching in uploaded documents through RAG with source-linked answers, and ships as a web app via PyPI or Docker with guided learning paths, quizzes targeting weak areas, and multi-user server support; v1.6.15 landed 19 Oct with bounded tutoring memory and native Gemini speech.
Why it matters
What you could build with it
Does it hold up?
Mature and deployable: 47k stars, monthly releases through v1.6.15, and documented PyPI/Docker installs with multi-user support, though setup needs Python plus Node and a configured LLM provider.
Built with DeepTutor
- DeepTutor: Open-Source, Multi-agent AI Learning Assistantarticle · Install guide: PyPI, Docker, and source installs, the localhost:3782 web UI, CLI mode, and LightRAG/GraphRAG extras.
- Open-Source AI Learning Assistant | DeepTutor for Privacy-Conscious Learnersarticle · Self-hosting pitch: Apache 2.0, persistent memory, autonomous TutorBots, and a unified workspace for chat, research, and quizzes.
- HKUDS/DeepTutor v1.6.15article · Release notes for 19 Oct: revised learning paths, independently graded question groups, bounded tutoring memory, and native Gemini speech.
Learn more
First spotted on github: source.
More AI agents
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.