OpenClaw Enterprise
BetaOpen source, vendor-neutral platform for managing persistent agents in sensitive environments
OpenClaw Enterprise is a free, MIT-licensed control plane for deploying and governing persistent AI agents, self-hosted on Docker Compose or Kubernetes and developed with Red Hat and NVIDIA. Conditional because it is pre-1.0: its own announcement offers it for internal pilot workloads, with 1.0 expected later in 2026, and we have not run it. Suits platform and security teams piloting agents in sensitive environments; wait for 1.0 before production-critical rollouts.
Control plane for deploying and governing AI agents; LLM-based reviews are one of its security controls, per the announcement
Is it right for you?
Good for
- Piloting persistent agents inside an organization that needs security and governance controls: the announcement lists multi-tenancy, hard boundaries between trusted and untrusted workloads, sandboxing, LLM-based reviews and fine-grained permissions
- Self-hosting with no licence cost: MIT licensed, and the announcement says it will always be free for any organization to use
- Running on infrastructure you already operate: Docker Compose for local setup, Kubernetes for deployment
- Swapping components: the announcement says core primitives like the harness, model and sandbox can be swapped out with third party solutions
Not good for
- Production-critical rollouts today: the announcement offers it for internal pilot workloads and expects 1.0 later in 2026
- Teams wanting a hosted service: the pages we read describe self-hosting only and list no managed offering
- Quick evaluation on a laptop without container tooling: the README quickstart lists Docker or Podman, k3d, kubectl, Helm, Go, Node.js 24+ and pnpm
- Teams with no model API access: the README quickstart needs an OpenAI API key for the default model or a selected override to send requests to the first deployed agent
Our experience
Pricing
Open Source- Announcement: will always be free for any organization to use
- Self-hosted via Docker Compose or Kubernetes
- Quickstart needs an OpenAI API key for the default model or a selected override (README)
Workflow recipes
Stand up a local pilot control plane
Run OpenClaw Enterprise locally before any shared deployment
- Install the README prerequisites: Docker Engine or Podman, k3d, kubectl, Helm, Bash, Python 3, the Go version in go.mod, Node.js 24+ and the pnpm version in package.json
- Clone github.com/openclaw/openclaw-enterprise and follow the local-setup quickstart
- Provide an OpenAI API key for the default model or a selected override, which the README says is needed to send model requests to the first deployed agent
- Deploy a first agent through the OpenClaw Control Plane and review how deploying and managing agents works
- Move to a Kubernetes cluster only after reviewing the docs at docs-enterprise.openclaw.org
Use these instead for specific tasks
No verdict changes yet
The clock starts day one. Changes land here as our verdict evolves.
Sources
Verification log
- Profile· No changes
Automated agent
Want this running in your business?
We tested this tool and we build with tools like it every day. Tell us the workflow and we will set it up, integrate it and hand it over working.
The conversation runs on Gnosari, one of the tools in this directory. A real conversation, not a sales script.