AI资讯大全AIPPXP.CN搜索 ↗
工具介绍 · 附安装包

DeepTutor: 52,000 stars self-hosted AI private education, turning your materials into a learning platform that will ask questions, ask questions, and keep an eye on progressDeepTutor:5.2 万星自托管 AI 私教,把你的资料变成会出题、会追问、会盯进度的学习平台

A collection of essays and textbooks you've never read? This open source project from HKUDS (Hong Kong Big Data Intelligence Laboratory) has made "learning" a complete product: multi-agent collaboration plus rag, turn your documents into a knowledge base, generate structured learning "Book", with quizzes, mastery tracking and learning paths, and remember where you have learned throughout the memory layer. With one-click self-hosting with Docker, your data is all yours.

收藏了一堆论文和教材,从来没读完过?这个来自 HKUDS(港大数据智能实验室)的开源项目把「学习」这件事做成了完整产品:多代理协作加 RAG,把你的文档变成知识库,生成结构化的学习「Book」,配测验、掌握度追踪和学习路径,还有贯穿始终的记忆层记住你学到哪了。可以 Docker 一键自托管,数据全在自己手里。

2026-10-31 更新 · 免费
DeepTutor:5.2 万星自托管 AI 私教,把你的资料变成会出题、会追问、会盯进度的学习平台

What's the difference between it and "throwing PDFs to ChatGPT"它和「把 PDF 扔给 ChatGPT」差在哪

The difference is in the action of "teaching". Ordinary questions and answers are what you say. DeepTutor will first build a knowledge base for your materials, and then generate a structured learning manual around it. Each concept is accompanied by explanations, follow-up questions, and quizzes. If you make a mistake, you will come back and talk about it.

It has built-in three layers of memory: remember this conversation in the short term, track your learning progress and weaknesses in the medium term, and accumulate your knowledge portrait in the long term. The next time it opens, it knows which step of the Fourier transform you were stuck in last time.

Mastery gates are a very practical design: the current concept is not up to standard, and the learning path will not let you go to the next chapter.

中文

差在「教」这个动作上。普通问答是你说一句它答一句;DeepTutor 会先给你的材料建知识库,再围绕它生成一本结构化的学习手册,每个概念配讲解、追问和测验题,做错了还会绕回来重讲。

它内置三层记忆:短期记住本次对话,中期追踪学习进度和薄弱点,长期积累你的知识画像。下次打开,它知道你上回卡在傅里叶变换的哪一步。

掌握度门槛(mastery gates)是个很实用的设计:当前概念没达标,学习路径不会放你进入下一章。

它和「把 PDF 扔给 ChatGPT」差在哪

Core features at a glance核心功能一览

Knowledge base: Upload documents are automatically segmented, indexed, and retrieved by rag, and multiple libraries are supported for separate management.

Study Book: Automatically generate a chapterized study manual, which can be read while highlighting and generating supporting quizzes.

Conversational teaching: You can specify the teaching mode in the chat interface, and the answer will be broken down into small steps and take the initiative to check the understanding.

Calendar and Taskpad: Schedule learning plans for review with Mastery Data Reminders.

中文

知识库:上传文档自动切分、索引、做 RAG 检索,支持建多个库分开管理。

学习 Book:自动生成章节化的学习手册,可以边读边划重点、生成配套测验。

对话式教学:聊天界面里可以指定教学模式,回答会拆成小步并主动检查理解程度。

日历与任务板:把学习计划排进日程,配合掌握度数据提醒复习。

核心功能一览

Deployment: Docker on-demand or pip on-demand部署:Docker 一条龙或 pip 按需装

The most troublesome thing is Docker: the warehouse is equipped with compose.yaml, and there is a complete web interface and service side when pulled up, which is suitable for users who want to run on home servers or NAS for a long time.

To embed your own Python project, directly pip install the core module: pip install deeptutor (requires Python 3.11–3.14), and then connect your own model API according to the documentation.

The model access takes the OpenAI compatible interface, OpenAI, DeepSeek, local Ollama can be connected, the key is configured in .env, and does not depend on any specific vendor.

中文

最省事的是 Docker:仓库里配好了 compose.yaml,拉起来就有完整的 Web 界面和服务端,适合想在家庭服务器或 NAS 上长期跑的用户。

想嵌入自己的 Python 项目,直接 pip 安装核心模块:pip install deeptutor(需要 Python 3.11–3.14),再按文档接上你自己的模型 API。

模型接入走 OpenAI 兼容接口,OpenAI、DeepSeek、本地 Ollama 都能接,密钥配置在 .env 里,不依赖任何特定厂商。

部署:Docker 一条龙或 pip 按需装

Download instructions下载说明

The download button provides the PyPI official core module wheel (1.68 MB, deeptutor 1.7.0, the same version number as the latest release of GitHub), including the core engine, multi-agent orchestration and rag modules, Apache-2.0 is licensed; the web interface and Docker orchestration files are available on the GitHub repository.

Self-hosting requires a self-hosted model API (or local model), and both the learning data and the vector index are stored on your own machine.

中文

下载按钮提供 PyPI 官方核心模块 wheel(1.68MB,deeptutor 1.7.0,与 GitHub 最新 Release 同版本号),含核心引擎、多代理编排与 RAG 模块,Apache-2.0 许可;Web 界面与 Docker 编排文件请在 GitHub 仓库获取。

自托管需要自备模型 API(或本地模型),学习数据与向量索引都存在你自己的机器上。

Use reminder使用提醒

The AI-generated explanations and quizzes suggest self-checking key knowledge points; the project iterates quickly, and the deployment steps are based on the warehouse README. This article is compiled from the public information of the project.

中文

AI 生成的讲解与测验建议自行核对关键知识点;项目迭代很快,部署步骤以仓库 README 为准。本文整理自项目公开资料。

资源下载 · Download

⬇ Download · 点击下载:DeepTutor 核心模块 wheel(1.68MB)

取自 PyPI 官方源 deeptutor 1.7.0,与 GitHub HKUDS/DeepTutor v1.7.0 同版本,Apache-2.0,含核心引擎与多代理/RAG 模块。

0阅读0 条评论

阅读与点赞数据保存在你的浏览器本地,欢迎留下你的想法。

评论 文明发言,让讨论更有价值

正能量公益广告今日正能量学一点,用一点;今天种下的种子,会长成明天的能力。去免费下载专区 →广告