HiveMQ MQTT Broker

Overview

HiveMQ MQTT Broker moves messages between connected devices and enterprise systems using MQTT. It is best suited to teams running large IoT or mission-critical workloads that need runtime clustering and enterprise security controls. Its operational depth is a strong fit for those environments, but the connection caps and capacity-based plans make it harder to justify for modest deployments.

Key features

Support for MQTT 3.1, 3.1.1, and 5.0, plus TCP, TLS, and WebSocket, gives teams several client connection options. Persistent messages stored to disk, offline queuing, QoS, configurable delivery guarantees, and dead-letter queues address reliability and failure handling. The 256 MB message-size ceiling allows large payloads, though teams should match it to their actual message patterns rather than treating it as a reason to send oversized messages.

Dynamic clustering allows operators to add or remove broker nodes without downtime and adjust cluster size while running. That is valuable when capacity changes cannot wait for a maintenance window. HiveMQ says the broker scales to millions of connections and billions of messages; that broad claim should be weighed against the specific connection and throughput limits of each plan.

Extensions connect with streaming services, databases, data warehouses, and security services, including Kafka, Amazon Kinesis, and Google Pub/Sub. Security options 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 it is ISO/IEC 27001:2022 certified and SOC 2 Type II compliant, which may matter to teams with formal security requirements.

Self-managed deployments can run on private, hybrid, or public clouds, with pre-built images for Kubernetes, OpenShift, and DC/OS. One practical constraint is the evaluation mode: without a valid license, a self-managed broker supports no more than 25 client connections. That is enough for a narrow technical evaluation, not a meaningful load test.

Pricing

HiveMQ has a free plan, paid options, and a 15-day trial. The named plans vary considerably in deployment model, capacity, and support, so teams should compare both connection and message-rate limits against expected load.

• Lab: Custom pricing; 100 connections and 100 messages per second, based on a normalized 5 KB message size. Its small stated capacity makes it a fit for limited workloads, not a large deployment.
• Launch: 299.00 USD per month, billed $299/month. It includes one broker deployment, 10,000 connections, 500 messages per second at a normalized 5 KB message size, and Community support. It suits a team with a defined single-deployment workload that can operate without paid support.
• Run: 499.00 USD per month, billed $499/month. It keeps the one-deployment and 10,000-connection limits, raises throughput to 1,000 messages per second at a normalized 5 KB size, and adds Silver support. The higher rate is relevant when throughput or support needs exceed Launch, but it does not increase the stated connection count.
• Self-Managed Enterprise: Custom pricing for self-managed multi-broker and multi-site deployments with custom capacity. This is the clearest fit for teams that need control over deployment and capacity beyond the fixed single-broker plans.
• Serverless: Custom pricing, with 100 connections and 10 GB/month of traffic. Those caps make it a constrained option for small workloads.
• Starter: Custom pricing, billed From $0.34/hour+ $0.80/million messages. It includes 10,000 connections, 1 MB/sec, a 99.95% Uptime SLA, and Silver support. The usage-based charge and throughput ceiling matter for workloads with variable traffic.
• Fully-Managed Enterprise: Custom pricing, with uptime of up to 99.99%. Teams considering it should weigh that uptime target against the needs of their deployment.

The self-managed evaluation cap is separate from plan capacity: without a valid license, evaluation mode stops at 25 client connections. Support options range from Community through Silver, Gold, and Platinum; Gold and Platinum are available up to 24×7.

Platforms

HiveMQ supports API, Linux, self-hosted, and Windows environments. Its hybrid deployment model and support for private, hybrid, and public clouds give infrastructure teams flexibility, while the pre-built Kubernetes, OpenShift, and DC/OS images suit teams already operating those platforms.

Who it's for

HiveMQ is aimed at enterprises and IoT teams that need MQTT messaging at scale, delivery controls, runtime cluster changes, and integration with existing data and security systems. The multiple authentication options and stated compliance credentials may also suit organizations with formal access-control and audit requirements. Smaller projects with low connection needs or no need for enterprise controls may find the capacity-oriented plans excessive, particularly if the self-managed evaluation cap blocks a useful test.

Pros and cons

• Pro: Runtime cluster resizing without downtime gives operators a way to adjust capacity without a planned broker outage.
• Pro: Persistent messaging, offline queues, QoS, configurable delivery guarantees, and dead-letter queues provide several tools for managing delivery and failure.
• Pro: Broad security controls and extensions to named streaming services make it easier to align the broker with enterprise systems.
• Con: Self-managed evaluation is limited to 25 client connections without a valid license, which restricts meaningful evaluation at scale.
• Con: Launch and Run both state a 10,000-connection ceiling and one broker deployment, so Run's higher price buys throughput and Silver support rather than more connections.
• Con: Several plans have custom pricing, making it harder to compare total cost before matching capacity and deployment needs.

Alternatives

For another MQTT-focused choice with a free plan, consider EMQX, which also offers a serverless free tier and supports Linux, macOS, self-hosted, web, and API environments. Choose it when those platform options or its free serverless entry point better suit the stack.

Aiven for Apache Kafka is the better direction for teams choosing Kafka, with a free tier capped at 250 KiB/s, three days of retention, five topics, and two partitions per topic, plus a $35.00 USD per month Developer plan.

Amazon SQS offers a free tier of one million requests and standard and FIFO queues; consider it when a managed queue service is a better fit than an MQTT broker.

Confluent Cloud has a free Basic option with serverless managed clusters, autoscaling Kafka resources, and more than 80 managed connectors. It is worth considering for teams centered on managed Kafka and connector breadth.

Google Cloud Pub/Sub includes a 10 GiB free usage allowance and a pay-as-you-go throughput option after that allowance. It is an alternative for teams whose messaging needs align with that model.

Apache ActiveMQ is a free, open-source option under the Apache License 2.0. Consider it if that licensing model is the priority.

Apache Pulsar is another free option; its suitability depends on the requirements of the deployment.

Apache RocketMQ is free software under the Apache License 2.0 and can be considered when that licensing model is preferred.

For broader category comparisons, browse Message Queue (MQ) Software and Message Broker Software.

Verdict

Choose HiveMQ when your team needs an MQTT broker with runtime clustering, delivery controls, and enterprise-grade security options for a substantial IoT deployment. Those capabilities are its strongest case. Look elsewhere if a small evaluation, a low connection cap, or upfront price transparency is more important than deployment flexibility and enterprise controls.