HiveMQ MQTT Broker

Overview

HiveMQ MQTT Broker is an MQTT messaging broker for moving events between connected devices and enterprise systems. It is best suited to teams building large IoT deployments or mission-critical services that need scale, security controls, and runtime cluster changes. Its strengths are substantial, but smaller projects should weigh the capacity tiers and the self-managed evaluation limit.

Key features

The broker supports MQTT 3.1, 3.1.1, and 5.0, as well as TCP, TLS, and WebSocket. Persistent messages stored to disk, offline message queues, configurable delivery guarantees, QoS support, and dead-letter queues give operators several ways to handle message delivery and failure. A 256 MB maximum message size allows large payloads, though the right fit depends on the application’s message patterns.

Dynamic clustering lets teams add or remove nodes without downtime and change cluster size while the system is running. That is useful when capacity needs shift, particularly in large deployments. HiveMQ says the broker can scale to millions of connections and billions of messages; those are maker claims, not a substitute for matching a plan’s stated capacity to expected load.

Extensions connect the broker with streaming services, databases, data warehouses, and security services, including Kafka, Amazon Kinesis, and Google Pub/Sub. Security controls include TLS 1.2/1.3, mutual TLS, fine-grained access control, pluggable authentication such as OAuth 2.0 and LDAP, and audit logging. HiveMQ states that it is ISO/IEC 27001:2022 certified and SOC 2 Type II compliant, relevant credentials for organizations with formal security requirements.

The self-managed broker can run in private, hybrid, or public clouds, with pre-built images for Kubernetes, OpenShift, and DC/OS. Without a valid license, it starts in evaluation mode with a 25-client connection ceiling. That is enough to assess basic integration, but not to validate a larger deployment under representative connection load.

Pricing

HiveMQ’s pricing model is freemium, with a free plan, a 15-day trial, and paid options. The stated starting price is $0.34; the Starter plan is billed From $0.34/hour+ $0.80/million messages, so usage can add to the hourly charge.

Plan	Price and capacity	What it means
Lab	Custom pricing; 100 connections and 100 msgs/sec, based on a normalized 5 KB message size	A small capacity ceiling makes it a fit for limited workloads, not a growing production fleet.
Launch	$299 a month; 1 broker deployment, 10,000 connections, 500 msgs/sec at normalized 5 KB, Community support	A single deployment with a defined throughput cap suits teams that can work with community support.
Run	$499 a month; 1 broker deployment, 10,000 connections, 1,000 msgs/sec at normalized 5 KB, Silver support	It doubles Launch’s stated throughput and adds Silver support, while keeping the same connection and deployment counts.
Self-Managed Enterprise	Custom pricing; multi-broker and multi-site deployments, custom capacity	For organizations that need to manage their own deployment across sites and set capacity to their requirements.
Serverless	Custom pricing; 100 connections and 10 GB/month traffic	The traffic allowance is a key constraint for teams considering this option.
Starter	Custom pricing; billed From $0.34/hour+ $0.80/million messages; 10,000 connections, 1 MB/sec, 99.95% Uptime SLA, Silver support	Usage-based billing and a stated uptime SLA may suit a team preferring this capacity and support combination.
Fully-Managed Enterprise	Custom pricing; up to 99.99% uptime	The stated uptime ceiling distinguishes it for teams prioritizing managed service availability.

Launch costs less than Run but provides half its stated message rate and Community rather than Silver support. Both cap the deployment count at one and allow 10,000 connections. Teams needing multiple brokers or sites should consider Self-Managed Enterprise; teams considering Starter should account for both hourly and per-message charges. The free-plan and trial details do not establish additional capacity or terms beyond those stated above.

HiveMQ offers Community, Silver, Gold, and Platinum support options; Gold and Platinum support are available up to 24×7. Support availability is an important distinction for teams that need round-the-clock escalation.

Platforms

HiveMQ supports API, Linux, Windows, and self-hosted use. Its hybrid deployment model and cloud options give operators flexibility over where to run the broker, while Kubernetes, OpenShift, and DC/OS images align with teams already using those environments.

Who it's for

HiveMQ is a strong candidate for enterprise IoT teams that need MQTT messaging, runtime clustering, integration extensions, and a broad set of security controls. Its plan capacities and enterprise support options make it more appropriate for organizations planning around defined connection and throughput needs than for casual or very small deployments. A 25-connection evaluation limit also makes the self-managed version a poor way to assess full-scale load before licensing.

Pros and cons

• Pros: Dynamic clustering changes capacity without downtime, useful for services that cannot schedule broker outages.
• Pros: Persistent messaging, offline queuing, configurable delivery guarantees, and dead-letter queues provide several delivery and failure-handling options.
• Pros: MQTT protocol coverage, extensions, and security controls make it adaptable to enterprise IoT architectures.
• Cons: The self-managed evaluation mode is limited to 25 client connections, restricting realistic scale evaluation.
• Cons: Launch and Run are single-deployment plans, and Run’s higher throughput costs more while retaining the same connection ceiling.
• Cons: Starter combines hourly and per-million-message charges, so usage contributes to cost beyond the base hourly rate.

Alternatives

For a self-managed broker without a paid plan, RabbitMQ offers a free, community-supported server under the Mozilla Public License 2.0. Apache Artemis is another free self-hosted option under Apache License 2.0, with Java 17+ required. If core MQTT features and on-premises deployment are the priority, Eclipse Mosquitto is a free alternative.

For a managed notification service with a monthly free allowance, Amazon SNS offers 1 million requests, 100,000 HTTP notifications, and 1,000 email notifications in its Free Tier. Teams comparing queue-oriented services can also consider Amazon SQS, CloudAMQP, Google Cloud Pub/Sub, or browse Message Broker Software and Message Queue (MQ) Software.

Verdict

Choose HiveMQ when a large IoT or enterprise deployment needs MQTT, runtime clustering, integration options, and mature security controls. Its defined Launch and Run capacities, broader enterprise paths, and support tiers give teams choices as requirements grow. Look elsewhere if a free self-managed broker is the priority, if a 25-connection evaluation cap is too restrictive, or if a single-deployment paid plan cannot meet the needed topology.