Choose how you
run Magistrala.

Same platform, three ways to run it:

• Community: free to run yourself. Open-source backend services (Apache 2.0) plus a free-to-use UI you deploy as a Docker container.
• Private Cloud: a fully managed, dedicated instance of Magistrala Enterprise Edition that we host and operate for you.
• Self-managed: Magistrala Enterprise Edition in your own environment, licensed to you to run yourself, or deployed and operated by us on your infrastructure.

Every Private Cloud tier covers license, hosting, deployment, monitoring, dashboards, audit logs, and the updates listed for that tier. Higher tiers add the rules engine, alarms, reports, and scheduled backups. You focus on your devices and data; we run the platform.

Business takes a snapshot of your instance every week and keeps the last four, so you can restore to any of the previous four weekly points, covering 28 days. Enterprise takes one every day and keeps the last seven: at most 24 hours of data is ever at risk, across a 7-day window. Daily is the stronger guarantee even though its window is shorter, because what matters in a recovery is how recent your restore point is, not how far back the archive goes. Restores are performed by our team on request. Long-term archiving, point-in-time recovery, and custom retention are part of the Custom plan.

A device is any client that connects to the platform and exchanges messages, regardless of protocol (MQTT, CoAP, HTTP, or WebSocket). The device and message figures on each tier are guidance for typical workloads, not hard quotas.

Yes. Every Private Cloud plan runs on a dedicated instance with no shared infrastructure. We provision, run, and manage the instance for you.

Because each tier is a dedicated instance, that figure is a rough estimate of how many concurrent devices and messages the instance can comfortably handle. It is not a hard cap. We do not limit you artificially; the only limit is the hardware and how it is set up. We deliberately avoid a messages-per-second metric, because it is a weak and misleading number for a few reasons:

1. Message size varies enormously. A ping is a message, and so is a 100 MB binary upload. They do not cost the same, and they are not the same priority.
2. Networking hiccups and load balancing have an outsized impact as you approach the upper limit of throughput.
3. Messages carry different delivery modes and quality-of-service guarantees. Not all messages are equally important.
4. Throughput also depends on delivery semantics: whether it includes consumer delivery and message persistence. Slow consumers can effectively serialize your infrastructure, and sometimes that is the intended behaviour, especially in low-bandwidth, high-QoS environments.

Limits are soft. We monitor usage and reach out to move you to the tier that fits, rather than cutting you off mid-cycle. If a burst is temporary, nothing breaks.

Yes. Upgrades are available at any time. Downgrades require a migration on our side, so please contact us to arrange one. Changes are prorated, so you only pay for what you use in a billing period.

Annual Private Cloud plans are billed once per year and work out roughly 10% cheaper than paying monthly. You can switch between monthly and annual billing when your term renews.

Both cover a single-cluster deployment of the full Magistrala Enterprise Edition feature set. The difference is how updates work. A monthly license (€300 per cluster) gives you access to the latest release for as long as your subscription is active. A perpetual license (€3,999 per cluster) is a one-time payment that lets you run the version you bought indefinitely, including patches and minor updates for that version, but not upgrades to future major releases.

Yes. With a self-managed license you run Magistrala Enterprise yourself, anywhere. With the “Managed by us” option we deploy and operate it on your infrastructure: AWS, GCP, Azure, or on-premises, including air-gapped environments, aligned to your SLA, security, and data-residency policies.

A formal uptime SLA and high-availability clustering are part of the Custom plan and the self-managed “Managed by us” option only. The standard Private Cloud tiers (Prototype, Business, Enterprise) each run as a single dedicated instance with proactive monitoring and support, but without a contractual SLA or clustering. If you need a guaranteed uptime target or active-active high availability, that is what the Custom plan is for. Response times and architecture are tailored to your requirements.

Partly. The Magistrala backend is open source under the Apache 2.0 license and free to self-host forever, and the Community Edition UI is free to use and deploy as a Docker container. The Enterprise Edition, which unlocks the full feature set including the rules engine, alarms, reports, audit logs, and dashboards, requires a self-managed license or a managed plan.