Yes. The full SDK is open source under Apache 2.0. Self-host it forever, no usage limits. Enterprise adds distributed execution at scale, dashboard / API / scheduled / webhook triggers, SSO, RBAC, and audit logs.
What's the difference between Open Source and Enterprise?
Same SDK, same primitives. Open source runs your flows on a single-node server you operate yourself. Enterprise adds distributed execution across deployments — so you can replay, retry, and resume at scale — plus dashboard, API, scheduled, and webhook triggers, SSO, custom RBAC roles, org-wide audit logs, multi-region and air-gapped deployment, and SOC 2 Type II / ISO 27001.
Can I run Kitaru on my own cloud?
Yes. Self-host the server in your own VPC, point it at S3, GCS, or Azure Blob, and run flows on Kubernetes, Vertex, SageMaker, or AzureML. Enterprise customers can run a dedicated control plane in their own region or fully air-gapped.
Do I have to share my agent data with Kitaru?
No. In the open-source path, executions, artifacts, and prompts stay inside your infrastructure end-to-end. Enterprise stores execution metadata and pointers — your artifact store stays under your control.
Can I migrate from Open Source to Enterprise later?
Yes. The schema is the same on both sides, so flows, checkpoints, memory scopes, and stack configs port over. We'll help script the cutover during onboarding.
How do I get support?
Open source: GitHub Discussions, GitHub Issues, and the community Slack. Enterprise: a dedicated solutions engineer, a private Slack Connect channel, and an SLA on response and resolution times.
Get started in seconds
Ship reliable agents. Start free.
Open source today. Talk to us when your team needs more.