Umple is a suite of technologies for:
- Modeling software using data models (class diagrams), state models and others; the models can be edited textually or graphically
- Analysing models (finding errors, linting)
- Embedding action code in models using Java, Python, PhP, C++ and Ruby
- Organizing code using a variety of separation of concerns mechanisms (traits, mixins, mixsets, aspects)
- Generating code in any of the above languages to create libraries or complete systems
- Generating diagrams, examples, and many other artifacts
- Ensuring that when AI generates code it does so via first generating a model that humans can understand
The main web page of Umple is https://www.umple.org
The following are the core repositories:
- umple: This contains the compiler and UmpleOnline, the online version of umple accessible live at https://try.umple.org
- umple-lsp: The language server protocol for Umple, used by IDEs and text editors. It uses the compiler above to analyze models
- umpleonline: A new UmpleOnline front end created mostly using AI. It uses the same compiler. it is accessible live at https://newumpleonline.umple.org
- umple.vscode: VSCode plugin for Umple (newly updated in 2026 to use umple-lsp).
- umple.zed: Zed plugin for Umple (new as of early 2026).
- umple-skills A collection of AI skills to allow AI aghents to work with Umple
- More key repos to be added here later
The Umple user manual is at https://manual.umple.org
For comprehensive information such as the issue tracker, wiki, governance and contribution guidelines, see the main Umple repository
Umple is mostly developed by students (mostly 4th year, Masters and PhD students).
Umple is supported by grants from NSERC and infrastructure from the Digital Research Alliance of Canada. Umple has also received corporate support over the years from various sources including IBM.
