Skip to main content

Build with Camunda 8.9

Go from zero to a running workflow in under two minutes.

Choose how you want to get started:

FASTEST START

Free trial on SaaS

Get your own fully managed Camunda cluster. Nothing to install, just sign up and start modeling today.

  • Zero setup - runs in Camunda's cloud
  • Web Modeler included
  • AI agent skills built in
  • 30 days free, fully featured
Create free account ~60 seconds to your first cluster
OR
FULL CONTROL

Run locally with the CLI Camunda 8.9+

Use the c8ctlCamunda CLI tool. Run locally and scale seamlessly to enterprise clusters with the same runtime and stack, deploying anywhere with one command.

  • Works on macOS, Linux, Windows
  • Desktop Modeler for modeling
  • AI agent skills built in
  • Your data stays on your machine
Install now ~2 minutes to a running engine

Get started with c8ctl Camunda 8.9+

Install c8ctl from npm and spin up a Self-Managed Camunda cluster on your machine.

c8ctl uses c8run, which requires OpenJDK 21–25 .

Terminal
$ npm install @camunda8/cli -g
$ c8ctl cluster start

# Camunda is now running! Deploy your first process:
$ git clone https://github.com/camunda/camunda-8-get-started.git
$ cd camunda-8-get-started/1-rocket-launch/
$ c8ctl deploy .
$ c8ctl run rocket-launch.bpmn --variables='{"fuelLevel":90}'

# Open Operate at http://localhost:8080/operate to see your process instance running. Log in with the credentials `demo/demo`. 

No npm? Download Camunda 8 Run directly and start a local Camunda cluster without the CLI.

Download Camunda 8 Run

Requires OpenJDK 21–25. Extract the archive and run:

Terminal
# macOS / Linux
./start.sh

# Windows
.\c8run.exe start

Installation guide

Download Desktop Modeler to design and edit BPMN diagrams, DMN decisions, and forms.

Download Desktop Modeler

On macOS, you can install the Desktop Modeler via Homebrew:

Terminal
$ brew install --cask camunda-modeler

Or download and install manually:

Version 5.45 · Installation guide

Everything from your terminal Camunda 8.9+

c8ctl gives you a single CLI for the full Camunda lifecycle with no browser required.

Manage your clusters

Terminal
$ c8ctl cluster start 8.9.0-alpha5
$ c8ctl cluster stop

Deploy and start processes

Terminal
$ c8ctl deploy ./rocket-launch.bpmn plot-destination.dmn
$ c8ctl create pi --id=rocket-launch --variables='{"fuelLevel":90}'
$ c8ctl list pi

Interact with processes

Terminal
$ c8ctl list jobs --type=launch-approval
$ c8ctl activate jobs launch-approval
$ c8ctl complete job 2251799813685252

Monitor and debug

Terminal
$ c8ctl list inc --state=ACTIVE
$ c8ctl get inc 2251799813685251
$ c8ctl resolve inc 2251799813685251

Teach your AI agent Camunda Camunda 8.9+

Give your coding assistant Camunda superpowers. Deploy processes, manage clusters, and query instances right from your AI workflow.

Add Camunda skills as Claude plugin Coming soon

Terminal
$ claude plugin add camunda/camunda-ai-dev-kit

Available skills:
  /new-project  — scaffold a new Camunda project
  /new-process  — generate a BPMN process
  /new-agent    — generate an agentic AI process (ad-hoc sub-process + AI Agent connector)
  /new-dmn      — generate a DMN decision table
  /new-form     — generate a Camunda Form
  /new-worker   — generate a job worker
  /deploy       — deploy resources to Camunda
  /start        — start a process instance
  /status       — check instance/incident status
  /view-process       — visualize BPMN, DMN, or Form files
  /setup-environment  — install and start Camunda 8 Run locally

Connect to Camunda via MCPOrchestration Cluster MCP|Docs MCP

Terminal
// mcp.json
{
  "servers": {
    // Locally running C8Run instance
    "camunda": {
      "type": "http",
      "url": "http://localhost:8080/mcp/cluster"
    },
    // Knowledge from the Camunda docs website
    "camunda docs": {
      "type": "http",
      "url": "https://camunda-docs.mcp.kapa.ai"
    }
  }
}

Then ask your agent:

💬"Deploy the order-process.bpmn and start an instance with orderId=42"
💬"Show me all incidents on the payment-flow process"
💬"Investigate incidents on the payment-flow process and resolve any that are due to missing payment details"

Agentic orchestration

Orchestrate AI agents in BPMN-based workflows. Design, deploy, and monitor intelligent agents combining human tasks with LLM reasoning.