No API keys. No cloud accounts. No surprises. Just a working AI Agent.
Getting started with AI Agents in Camunda has never been easier, and Camunda Provided LLM is the reason why.
If you've been exploring Camunda's agentic orchestration capabilities but keep hitting the same wall — getting budget approval and IT security approval, setting up LLM credentials — Camunda Provided LLM removes that wall entirely. Camunda Provided LLM is Camunda's managed LLM provider built directly into Camunda 8 SaaS. You can deploy and run an AI Agent in minutes without ever touching an API key.
In this post, we'll walk through:
- What Camunda Provided LLM is and why it's a game-changer for teams exploring agentic orchestration
- How budgets work to keep your experiments on track
- A step-by-step guide to running your first blueprint
- How to transition to your own LLM provider when you're ready for production
What is Camunda Provided LLM?
Camunda Provided LLM is a Camunda-managed LLM provider option available to Camunda 8 SaaS users. Instead of requiring you to bring your own provider credentials from day one, Camunda provides and manages the LLM access behind the scenes.
Here's what that means in practice:
- No LLM account setup. You don't need an additional sign up with a model provider or paste credentials into secrets just to explore Camunda agents.
- Compare different LLM providers. You can run your agent with leading providers' models and easily switch between them, so you freely experiment which one is better for your use case.
- No surprise bills. Your organization gets a free, preconfigured budget for testing and experimentation.
- Instant blueprints. AI Agent blueprints that use Camunda Provided LLM work out of the box; deploy and run, no configuration needed.
- Seamless transition. When you're ready for production, switch to a customer-managed provider like AWS Bedrock, without changing your process architecture.
Camunda Provided LLM is available today for:
- SaaS trial organizations: A meaningful free budget to let individuals and small teams explore AI Agents quickly.
- SaaS enterprise organizations: A more generous budget so multiple teams can run proofs of concept without wrestling with credentials on day one.
Why we built it
The feedback from early adopters was consistent: people wanted to see Camunda's agentic orchestration in action now, not after battling internal LLM approvals, signing up to providers and creating API keys just to be able to run a Camunda agent.
Previously, most AI Agent examples assumed the users had access to providers like AWS Bedrock. They're powerful, but nontrivial to configure in enterprise environments. There was no AI Agent that "just ran." Evaluation teams got stuck on credentials before they could demo anything; many never hit that "aha" moment.
Camunda Provided LLM flips that experience. You get immediate access to ready-to-run AI agent blueprints backed by an LLM managed by Camunda, so you spend your first hour testing agentic patterns instead of wading through LLM provider docs.
If you haven't decided on the model provider and the model, it gives you the ability to compare them without having to sign up to multiple vendors. You can quickly compare how your agent works for example with OpenAI's GPT-5 and Google's Gemini Pro. Let's see how this works!
How it works
Camunda Provided LLM uses a managed LLM gateway that Camunda operates and maintains. This gateway routes requests to model providers — for example Anthropic's Claude via an OpenAI-compatible API — so you don't have to deal with the LLM configuration.
In your SaaS cluster, if Camunda Provided LLM is enabled, you can see two automatically provisioned secrets:
CAMUNDA_PROVIDED_LLM_API_ENDPOINT
CAMUNDA_PROVIDED_LLM_API_KEYMost AI Agent blueprints are already wired to use these secrets by default. You don't configure anything — you just run.
Get started: Run your first AI Agent
Step 1: Enable AI features
If you're on a Camunda SaaS Trial plan, AI features are available by default. Just sign up for a SaaS Trial and you're ready to go.
If you're on a Camunda SaaS Enterprise plan, navigate to Camunda Console and enable AI-powered features for your organization. This toggle brings Camunda Provided LLM online and provisions the managed LLM credentials for your clusters.
Step 2: Deploy and run a blueprint
The fastest path to a working AI Agent is a Camunda Provided LLM-compatible blueprint:
- Open Web Modeler and create a new project.
- Click Browse blueprints.
- Select any blueprint marked as compatible with Camunda Provided LLM. We recommend the AI Agent Chat Quick Start blueprint to get started.
- Click Deploy and Run.
That's it. No LLM configuration. No secrets to paste. A working, end-to-end AI Agent in minutes.
Optional next step: Build an agent from scratch
Once you've seen a blueprint run, you might want to build something of your own. Camunda Provided LLM makes that easy too. When manually configuring an AI Agent connector, just point it to the pre-provisioned secrets:
- Provider: OpenAI Compatible
- API Endpoint:
{{secrets.CAMUNDA_PROVIDED_LLM_API_ENDPOINT}} - API Key:
{{secrets.CAMUNDA_PROVIDED_LLM_API_KEY}} - Model: Select one of the available Camunda Provided LLM models
From there, the focus shifts entirely to what matters:
- Designing your BPMN agent flow
- Tuning your prompts
- Modeling error handling
- Adding human-in-the-loop steps
- Connecting to your existing systems via connectors
Step 3: Monitor usage
The budget is shared across your organization, so you will want to keep an eye on consumption. Camunda Console shows Camunda Provided LLM usage statistics. You can see how much budget has been used, by which processes, and plan your transition to a customer-managed setup accordingly.
Moving to your own LLM Provider
Camunda Provided LLM is built intentionally as a launchpad, not a lock-in mechanism. When you're ready to move a process to production, or you need tighter controls, a different provider, or higher throughput, we have made advancing to production straightforward.
Before you switch:
- Confirm your organization has access to your preferred provider (e.g., AWS Bedrock).
- Collect credentials: keys, regions, model IDs, authorization settings.
- Identify which connectors currently rely on Camunda Provided LLM defaults.
Then:
- Add your provider credentials as secrets in Camunda Console.
- Update your AI Agent connector configuration to point to the new provider.
- Redeploy your process.
- Test your AI Agent using task testing.
- Run an end-to-end test in Operate and Tasklist.
The key benefit: your orchestration model doesn't have to change. The BPMN, event choreography, and human touchpoints you designed with Camunda Provided LLM carry forward unchanged — only the LLM backend configuration shifts.
Who Camunda Provided LLM is for
Camunda Provided LLM is especially useful for:
- New SaaS trial users who want to see an AI Agent run end-to-end before spending any time on LLM setup.
- Enterprise teams exploring agentic orchestration who need a low-friction way to build and socialize PoCs internally.
- Partners and consultants who spin up demos frequently and don't want to redo provider configuration every time.
The pattern is the same in every case: start with Camunda Provided LLM to hit the "aha" moment fast, learn what works for your use case, then bring your own LLM provider when you're ready to go further.
Try it now
If you're already on Camunda 8 SaaS, you have everything you need to get started today:
- Enable AI features in Console (Enterprise) or start your trial.
- Open Web Modeler and browse the Camunda Provided LLM-compatible blueprints.
- Deploy one and run it.
Your first AI Agent should be up before your coffee gets cold. From there, adapt it into something that matches your real process, and when you're ready, bring your own LLM and take it to production.
A quick note about budgets
Camunda Provided LLM is designed for experimentation and onboarding, not production use. Budgets are enforced at the organization level, so everyone shares the same pool.
- Trial organizations get an optimized budget tuned for quick first runs.
- Enterprise organizations get a larger budget suitable for multi-team PoCs.
Usage data is available in Camunda Console, so admins and project leads can monitor consumption and build the business case for moving to a fully provisioned, customer-managed setup.
