HKUDS

DeepTutor: Personalized Tutoring with LlamaIndex RAG

DeepTutor is an agent-native tutoring workspace for personalized learning, research, problem solving, and practice. It uses LlamaIndex for its RAG pipeline and runs as a web app, CLI, or Docker deployment.

ActivePythonApache-2.0AI AgentsEducationWeb App
Stars
41.8k
Forks
5.3k
Stars, last 30 days
+18,247
Last commit
1 day ago
Contributors
100+

Our take

Activeregular commits and releases

Health score

69/100

How it is scored
Activity
87
Community
73
Issues
66
Pull requests
50

+18,247 stars in the last 30 days

  • Releases more often than 91% of the projects we track
  • Merges more pull requests than 91% of the projects we track
  • More contributors than 78% of the projects we track
  • Slower to merge pull requests than 91% of the projects we track

DeepTutor is a feature-rich, actively maintained tutoring platform with broad deployment choices, but its extensive configuration and evolving feature set make it a better match for hands-on adopters than for teams seeking a minimal, turnkey tutor.

Good fit if

  • You want one workspace for personalized tutoring, study materials, practice, and research.
  • You need a self-hostable web app or CLI with configurable LlamaIndex-based retrieval and other knowledge engines.
  • You value an actively developed project with regular releases and contributions from a broad contributor base.

Look elsewhere if

  • You need a narrowly scoped tutoring tool with little setup or few configuration choices.
  • Your environment cannot provide Python 3.11–3.14, or Node.js 20+ for the packaged PyPI web app.
  • You need a proprietary license rather than Apache-2.0.
All health signals
Last commit2026-10-18 (1 day ago)
Commits, last 90 days100+
Releases, last 12 months87 (latest v1.6.15, 2026-10-18)
Contributors100+ (top contributor: 62% of commits)
Issues closed, last 90 days40+ (typically closed in 7 days)
Pull requests merged, last 90 days87 (typically merged in 6 days)
Project age10 months

Checked on 2026-10-20 with the GitHub API.

Project history

Daily GitHub snapshots since Oct 7, 2026. History covers stars and forks; health signals are not tracked historically.

Selected snapshot

Stars

59,825

Forks

7,242

Change since Nov 16, 2026: +296 stars·+36 forks

Overview

DeepTutor is an agent-native workspace for lifelong personalized tutoring. It brings chat, research, problem solving, quizzes, visualizations, and guided practice into a shared learning context, helping learners work with their own materials and carry context between activities. Its knowledge-base RAG pipeline uses LlamaIndex, alongside other selectable retrieval engines.

The project offers a web app, a CLI, and Docker deployment. It is aimed at learners who want a customizable AI study environment, and at developers or educators who want to configure its tools, models, and knowledge sources rather than use a single-purpose tutoring chatbot.

Key Features

  • Shared agent runtime for chat, question answering, quizzes, research, visualization, and study workflows.
  • Personal learning context that can connect knowledge bases, books, notes, question banks, and memory.
  • LlamaIndex RAG plus other supported knowledge engines, including PageIndex, GraphRAG, and LightRAG.
  • Tools and skills that extend the tutor with web and paper search, code execution, MCP services, and community skills.
  • Learning workflows for courses, mastery practice, reading, and video study with transcript search.
  • CLI commands for chat, capability runs, knowledge-base management, memory, and configuration.
  • Local deployment options through PyPI, a source checkout, or Docker.

Use Cases

  • Learners can ask questions grounded in their own textbooks and documents, then reuse that context in practice or research.
  • Students can turn study materials into quizzes and structured mastery practice, and track ongoing study tasks.
  • Researchers can use agent-assisted web and paper search to investigate a topic and produce cited reports.
  • Developers can run the CLI or self-host the web app and configure providers, retrieval engines, skills, and tools.

What you need

Detected in the repository

  • Python >=3.11,<3.15 (from pyproject.toml)
  • A Dockerfile, so it can run in a container; a Compose file sets up the related services
  • A test suite and automated checks on GitHub Actions

License in plain words

Apache-2.0permissive

  • Commercial use: yes
  • Modify and redistribute: yes
  • You must keep: the license, the NOTICE file and a note of your changes
  • Share your changes: no
  • Includes an explicit patent grant from the contributors:

A summary, not legal advice: the LICENSE file is what applies.

Getting Started

For a quick Docker trial:

docker run --rm --name deeptutor \
  -p 127.0.0.1:3782:3782 \
  -v deeptutor-data:/app/data \
  ghcr.io/hkuds/deeptutor:latest

Open http://127.0.0.1:3782 and configure a model provider in Settings. See the README for PyPI, source, CLI, and deployment instructions.

Considerations

  • The project is actively changing, with frequent commits and releases. Teams adopting it should expect ongoing updates and verify that changes suit their deployment.
  • The full local web app needs Python 3.11–3.14 and Node.js 20+; source development calls for Node.js 22 LTS. Docker is an alternative when installing those runtimes directly is undesirable.
  • Model-provider credentials and, for knowledge-base features, an embedding configuration must be supplied. The README does not require a GPU or a database for a basic setup.
  • Code execution uses a runner sidecar or Linux bubblewrap when available; the local restricted-subprocess fallback is described as best effort. Review the sandbox configuration before enabling generated-code workflows.
  • Apache-2.0 permits broad use, subject to its license terms and notices.

Found this useful?

Share it with someone who would like DeepTutor.

Comparisons