jeecgboot/JeecgBoot
Developer ToolsA Java/Vue low-code platform that generates CRUD code and whole backend systems from prompts, aimed at enterprise Java teams.
JeecgBoot is worth a look if you are building repetitive enterprise CRUD inside a Spring Boot/Vue3 stack and want code generation plus an AI assistant inside one platform; if you want a small, minimal stack, the breadth works against you.
The problem
A Java team receives a new business module: tables, list pages, edit forms, permissions, menus, import/export, maybe a report. Before, someone would hand-write entity, mapper, service and controller layers, then repeat the same work in Vue views, then wire menu and button permissions. With JeecgBoot, the user describes the table or module in natural language or the online form designer, and the platform produces the schema SQL, backend CRUD and frontend pages with menu and permission entries, so the first runnable version exists before anyone opens an IDE.
What's different
The distinguishing claim is the workflow, not a single generator: Skills generate, then configuration happens online, code is generated, merged by hand, and further edited with AI. That sequence matters because generated code is treated as a starting point that developers keep and extend rather than a black box that must be regenerated. The generator covers single-table, tree, one-to-one and one-to-many models and ships multiple templates, including a uni-app template for mobile, and generated pages include query filters and Excel import/export. The AI side is integrated rather than bolted on: model management, knowledge base, chat assistant, workflow orchestration, MCP plugins and prompt management live in the same admin UI, and langchain4j with support for ChatGPT, DeepSeek, Qwen and local models avoids locking you to one vendor.
The trade-off is scope. This is not a library you add to an existing app; it is an opinionated full-stack platform with its own module structure, permission model, online form engine and microservice option. Adopting it for one screen means adopting its conventions, and customizing deep behavior means working inside those conventions. It also bundles commercial components such as JimuReport and JimuBI, so check licensing and edition differences before assuming everything shown is available under Apache-2.0.
Who should try it
Good fit: an enterprise Java team with a large internal-admin backlog, already comfortable with Spring Boot, MyBatis-Plus and Vue3, that wants a standard scaffold plus generation for routine modules. A team that needs AI chat, a knowledge base and MCP tools on the same admin platform may also prefer one integrated product to assembling several.
Poor fit: a solo developer building a small service, or anyone whose stack is not Java/Vue, will spend more time removing platform assumptions than writing code. Teams needing a narrowly scoped codegen CLI should also look elsewhere; the platform is the product here.
Requirements and limitations
Frontend needs Node 20.19+ or 22.12+ and pnpm 9+; backend defaults to JDK 17 with JDK 21 and 24 listed, Maven, Lombok in the IDE, and MySQL 5.7+ scripts by default. Other databases are listed as supported, including Oracle, SQL Server, PostgreSQL, Dameng and Kingbase, but the README says only MySQL scripts ship and others require a separate conversion guide, so porting to another database is work you own database conversion notes.
Ongoing care is the bigger constraint: this is a full-stack platform with Redis, a security stack based on Shiro and JWT, and an optional microservice profile with Nacos, Gateway, Sentinel and SkyWalking. Upgrades touch generated code and templates, so plan for merge discipline rather than treating generation as a one-time action.
HotRepo's take
HotRepo recorded a single daily-board observation on 2026-10-28, when the repository had 48,306 stars and gained 7 that day. One dated sample cannot establish whether interest is rising or fading, so persistence cannot be judged from this history.
The concrete thing to verify before committing is the boundary between community and commercial editions: several components in the feature list are named separately, and the README excerpts do not spell out which parts are covered by the Apache-2.0 license.
Written by AI from the project's README, updated October 28, 2026. The project's own docs are the source of truth.