kUML

kUML is a Kotlin-based modeling environment for teams that want diagrams, specifications, and generated artifacts to live in a software workflow. It is strongest for developers who are comfortable treating models as source code and build inputs. The broad notation coverage and open-source licensing are persuasive, while script execution makes workspace trust an important condition.

Overview

kUML expresses UML 2.x, SysML 2, C4, BPMN 2.0, and ERM through a type-safe Kotlin DSL. It covers all 14 UML 2.x diagram types: class, object, package, component, composite structure, deployment, profile, use case, activity, state machine, sequence, communication, timing, and interaction overview. That range suits projects that move between architecture, behavior, deployment, and data modeling instead of limiting themselves to one diagram style.

The Kotlin approach changes who benefits most. Compile-time validation and IDE support can catch modeling errors in the same kind of feedback loop developers use for application code. Reverse engineering and deployment support extend the workflow beyond drawing. Readers looking for a purely visual editor with no code-oriented model layer may find the approach less natural.

kUML connects to CLI and Gradle workflows, IntelliJ, Obsidian, Asciidoctor, and VS Code. Structurizr DSL round-trip supports exchange with Structurizr-based work, while the optional kuml-io-emf module provides XMI interoperability. An online handbook covers installation, usage, and features, and the Apache 2.0 license permits open-source use.

Key features

Modeling and validation

The DSL spans software and systems notations in one environment. Compile-time checks and IDE assistance are a practical advantage when diagrams are maintained beside Kotlin projects or reviewed as code. The trade-off is that modeling requires comfort with Kotlin syntax and tooling rather than only manipulating diagram shapes.

AI and generated artifacts

An MCP server supports agent introspection, and the kuml ai subcommand enables AI-assisted modeling. Built-in transformations can produce JPA, OpenAPI 3.0 YAML, Kubernetes manifests, and Dockerfiles, making models useful as inputs to implementation and deployment work. These outputs are most valuable when a team wants a repeatable model-to-artifact path; they do not replace the need to review generated assets.

Distribution and security

Installation spans Homebrew on macOS and Linux, Chocolatey and MSI on Windows, native DEB, RPM, and DMG packages, Docker, and Maven Central library modules. A Compose Multiplatform desktop app is distributed as a macOS cask. The MCP authoring server validates scripts with KumlScriptGuard and applies the same sandboxing used by CLI render and validate commands. Even so, rendering Kotlin scripts executes their source code, so only trusted scripts or workspaces should be run.

Pricing

kUML is free and open source under Apache 2.0. The Free / open source plan costs free and includes the software; no paid plans or feature limits are stated. That removes seat or quota budgeting for teams able to operate the project themselves, while commercial support terms are not part of this offering.

Platforms

kUML supports Windows, macOS, Linux, web browsers, APIs, and self-hosted deployments. Package options and Docker support accommodate local, library-based, and hosted setups, with the desktop distribution specifically offered through a macOS cask.

Who it's for

Choose kUML for development teams that want type-safe models in Kotlin, compile-time checking, integrations with editors and build tools, and generation of persistence, API, container, or Kubernetes artifacts. It also fits organizations needing several modeling standards and open-source licensing. A conventional diagramming audience, or a team unwilling to execute model scripts, should look elsewhere.

Pros and cons

Pros

• Wide notation coverage: UML 2.x, SysML 2, C4, BPMN 2.0, and ERM support can consolidate varied modeling work.
• Developer feedback: Kotlin typing, compile-time validation, and IDE integration bring familiar checks to models.
• Useful automation: JPA, OpenAPI, Kubernetes, and Dockerfile transformations connect models to delivery assets.
• Open distribution: Apache 2.0 licensing and packages for major desktop and build environments reduce adoption barriers.

Cons

• Code-first learning curve: The Kotlin DSL favors developers and is less direct than a shape-based editor for occasional diagram authors.
• Execution risk: Rendering evaluates Kotlin source, so untrusted scripts or workspaces are unsafe to run despite sandboxing.
• Operational responsibility: Self-hosted, CLI, library, and container options give flexibility but leave setup choices to the team.

Alternatives

Sparx Enterprise Architect is the better choice when a freemium product with a free trial and a multi-user professional modeling environment is preferred.

ArgoUML suits readers seeking a free, open-source UML tool for Linux, macOS, and Windows with Java 5+ support.

BESSER is another free, open-source option spanning API, desktop, web, and self-hosted use.

BridgePoint fits users who specifically want an Apache 2.0 editor, verifier, and model compilers.

Eclipse Papyrus offers a free open-source modeling environment across desktop, web, and self-hosted platforms.

Horusec is a free Apache 2.0 project with CLI and platform components.

UMLet is preferable for a free UML tool available as stand-alone desktop software, an Eclipse plugin, web app, and VS Code extension.

Violet UML Editor is a completely free, cross-platform UML editor.

See the broader UML Modeling Software category for more options.

Verdict

kUML is a strong recommendation for developer-led teams that want rigorous Kotlin models, broad standards support, and direct generation of implementation and deployment files. Its main reason to choose it is the connection between validated models and everyday build tooling. Look elsewhere if you need a purely visual editor or cannot accept the security responsibility of executing modeling scripts.