Apache Hop
hop.apache.org
- Free plan
- Open source
- Windows
- macOS
- Linux
- Web
- API
- Self-hosted
What Apache Hop is
Apache Hop is a free, open-source platform for data integration and orchestration, built around visual pipelines and workflows. Users create pipelines on a canvas, preview rows at each step, and view live data and metrics during execution. A pipeline can run locally, on Hop Server, on Apache Spark, or on Flink and Dataflow through Apache Beam. Hop includes more than 250 pipeline transforms, over 80 workflow actions, and support for more than 40 database dialects. Listed technologies include PostgreSQL, MySQL, Oracle, SQL Server, Snowflake, BigQuery, Kafka, Salesforce, SAP, S3, REST, and SFTP. Projects can switch environments between development and production, while credentials can be kept out of pipelines. Hop runs on Windows, Linux, and macOS; Hop Web supports modern browsers. Release 2.19.0 requires Java 21. Docker images and a Kubernetes Helm chart are available, though Hop Web is work in progress and provided as-is. The project warns that operations can affect connected databases, file systems, and other infrastructure.
Who it suits
Apache Hop suits teams building data pipelines and orchestration workflows across varied runtimes. It is free under the Apache License 2.0, but users should account for its runtime requirements and the risks of destructive operations.
Good to know, and what to watch
- Free under the Apache License 2.0.
- Pipelines can run across several runtimes.
- Includes more than 250 pipeline transforms.
- Projects can switch environments between development and production.
- Credentials can be kept out of pipelines.
- Release 2.19.0 requires Java 21.
- Hop Web is work in progress and provided as-is.
- Community support has no promised response times.
- Operations can be destructive to connected infrastructure.
Plans
Every plan and priceApache Hop
Free plan
Getting started
- Open the Apache Hop website
- Choose a Linux, macOS, or Windows installation, or use self-hosted deployment
- Install Java 21 for release 2.19.0
- Create a project and design pipelines or workflows on the canvas
- Configure environments so projects can move from development to production
- Run pipelines locally, on Hop Server, Spark, or Beam
The Club's review
Apache Hop is an open-source environment for building data-integration pipelines and orchestrated workflows visually. It suits engineering and data teams that need to move the same designs across local, server, and distributed runtimes. The trade-off is operational responsibility: Java 21, tightly scoped permissions, and attention to the maturity of Hop Web are part of adopting it.
Overview
Hop combines a visual pipeline canvas with workflow orchestration rather than limiting users to a single execution engine. That makes it a credible choice for teams whose data work spans databases, files, APIs, streaming systems, and cloud services. It is less suitable for organizations seeking a fully managed service with predictable vendor response and minimal runtime administration.
Key features
The canvas lets users preview rows at each step and observe live data and metrics while a pipeline runs. In practice, that gives developers immediate visibility into intermediate results instead of forcing every check into separate scripts or jobs. The catalog includes more than 250 pipeline transforms, over 80 workflow actions, and 40-plus database dialects, providing substantial coverage for recurring integration patterns.
A single pipeline can execute locally, on Hop Server, on Apache Spark, or on Flink and Dataflow through Apache Beam. This portability is valuable when development, batch processing, and distributed workloads need different runtimes. It also means teams must decide which runtime to operate and account for its infrastructure and permissions.
Connectors cover PostgreSQL, MySQL, Oracle, SQL Server, Snowflake, BigQuery, Kafka, Salesforce, SAP, S3, REST, and SFTP. Environment switching supports promotion from development to production, while credentials can remain outside pipeline definitions. Incremental loading, change-data capture, and custom-code transforms extend Hop beyond simple full-refresh jobs, although custom code increases the need for engineering discipline.
Hop is released under the Apache License 2.0. Docker images are provided for Hop and Hop Web, alongside a Kubernetes Helm chart. Hop Web is described as work in progress and supplied as-is, so browser-based operation should not be treated as the most mature deployment path.
Pricing
Apache Hop — free
The free plan is the open-source platform under the Apache License 2.0, with no seat, usage, or term limits stated. Release 2.19.0 requires Java 21, which can add a meaningful platform-management task even though the software itself costs nothing. Community support is free and volunteer-based, with no promised response time; organizations needing service-level agreements, training, or migration help must use commercial providers.
There is no separate paid Hop tier described. Teams should also budget for the infrastructure and administration of Hop Server, Spark, Beam runners, containers, or Kubernetes where those execution options are used.
Platforms
Hop runs on Windows 7 or later, Linux on x86_64 or ARM, macOS, and self-hosted environments. Hop Web requires a modern browser. The broad operating-system coverage helps mixed estates, while the Java requirement and self-managed deployment model favor teams with established platform operations.
Who it's for
Choose Hop for data engineers and technical integration teams that need visual design plus control over execution placement. It is particularly well matched to organizations integrating heterogeneous databases and services, promoting projects between environments, or combining incremental loads and change-data capture with custom transformations.
Look elsewhere if your priority is a hosted, turnkey experience, guaranteed support response, or a browser product that is already fully mature. Administrators must also restrict permissions carefully: Hop warns that transforms and actions can perform destructive operations against databases, file systems, and other infrastructure. Security maintenance requires upgrading to versions containing fixes because individual binary patches are not produced.
Pros and cons
Pros
- Broad visual toolkit: 250-plus transforms, 80-plus actions, and 40-plus database dialects reduce the need to build common steps from scratch.
- Runtime flexibility: The same design can target local execution, Hop Server, Spark, or Beam-backed Flink and Dataflow.
- Strong integration coverage: Major relational, cloud, SaaS, messaging, object-storage, REST, and SFTP technologies are included out of the box.
- Promotion and loading controls: Environment switching, externalized credentials, incremental loading, and change-data capture support production-oriented workflows.
Cons
- Operational overhead: Java 21 and self-hosted runtime infrastructure require platform expertise.
- Web uncertainty: Hop Web remains work in progress and is provided as-is.
- Support limitations: Volunteer community support has no response-time promise; paid assistance is a separate procurement decision.
- Safety responsibility: Destructive operations are possible, so permissions and restrictions must be designed carefully.
Alternatives
For a broader shortlist, see ETL Software, DataOps Platforms, and Data Transformation Tools.
- Skyvia is a better fit for a hosted, freemium workflow with a free tier capped at 10,000 records monthly, two scheduled integrations, once-daily scheduling, and 30-day schedule expiration.
- EasyMorph suits users wanting a desktop free edition, accepting limits of 20 actions and 20 loop iterations per project.
- Jitterbit Harmony is for teams prepared to buy a commercial platform with defined connection and support tiers.
- Apache Beam is preferable when a programming model is the priority and execution costs are handled by the selected runner and infrastructure.
- Apache NiFi offers another free Apache-licensed integration option without stated commercial plans or usage limits.
- CloverDX targets organizations seeking commercial support and seat-oriented licensing.
- Keboola fits managed, freemium projects with metered compute, including 60 free minutes per month after its first-month allowance.
- Apache Spark is the stronger choice when distributed analytics is the central requirement rather than visual pipeline authoring.
Verdict
Apache Hop is a strong recommendation for technical teams that want an open, visual integration platform with many connectors and the freedom to move pipelines among execution engines. Choose it for control and breadth; look elsewhere when managed operations, guaranteed support, or a more finished web experience matters more than flexibility.
The verdict
Apache Hop offers visual pipeline design and several execution options without a listed software price. Teams need to manage Java requirements, operational permissions, and the work-in-progress status of Hop Web.
Questions people ask
Is Apache Hop free?
Yes. The Apache Hop plan is listed as free, and the platform is open source under Apache License 2.0.
Which operating systems does Hop support?
Hop supports Linux, macOS, and Windows. Hop Web supports modern browsers, and self-hosted deployment is available.
What runtimes can execute a pipeline?
The same pipeline can run locally, on Hop Server, on Apache Spark, or on Flink and Dataflow through Apache Beam.
Does Hop support major data services?
Listed integrations include PostgreSQL, MySQL, Oracle, SQL Server, Snowflake, BigQuery, Kafka, Salesforce, SAP, S3, REST, and SFTP.
What does Hop require at runtime?
Apache Hop 2.19.0 requires Java 21.
What the maker publishes
- Built-in scale
- The homepage states that the distribution includes more than 250 pipeline transforms, 80 workflow actions and 40 database dialects. source
- Cloud runtimes
- Run configurations include Databricks, native Spark, Beam Spark, Beam Flink and Google Dataflow options. source
- Commercial listing
- The Apache Hop project says its commercial-support listing is informational and that it does not endorse, rank, or vet the listed companies. source
- Commercial support
- The commercial-support directory lists Avizing, know.bi and Yupiik, while stating that Apache does not endorse or vet them. source
- Configuration
- Projects can be moved between development and production by switching environments, and credentials can be kept out of pipelines. source
- Deployment
- The download page provides Docker images for Hop and Hop Web and a Helm chart for Kubernetes; Hop Web is described as work in progress and provided as-is. source
- Download
- The Apache archive lists source code and a 333 MB apache-hop-client-2.19.0.zip released on 17 August 2026. source
- Execution
- The same pipeline can run locally, on Hop Server, on Apache Spark, or on Flink and Dataflow through Apache Beam. source
- Included components
- The platform includes 250+ pipeline transforms, 80+ workflow actions, and 40+ database dialects. source
- Independent documentation criticism
- A 2025/2026 academic project report said limited official documentation can make developing plugins from scratch challenging. source
- Independent usage description
- An OpenSchoolMaps training document describes Hop GUI, Hop Run CLI and Hop Server REST execution as the three primary clients. source
- Integrations
- Listed out-of-the-box technologies include PostgreSQL, MySQL, Oracle, SQL Server, Snowflake, BigQuery, Kafka, Salesforce, SAP, S3, REST, and SFTP. source
- Integrations
- Listed integrations include PostgreSQL, MySQL, Oracle, SQL Server, Snowflake, BigQuery, Redshift, Databricks, MongoDB, Neo4j, Kafka, Cassandra, Salesforce, SAP, Parquet, Avro, S3, Azure Blob, Google Cloud Storage, REST and SFTP. source
- Java requirement
- Current Hop 2.19.0 documentation requires a 64-bit Java 21 runtime. source
- Java SDK
- A documented Hop SDK provides Java initialization APIs and supports loading plugins from configured folders. source
- Licence
- Apache Hop is released under the Apache License 2.0. source
- License
- The project says Hop is developed in the open under the Apache License 2.0. source
- Limit: CDC
- The CDC guide says Hop has no single CDC transform and no built-in Debezium or GoldenGate connector; users must choose among log, watermark or snapshot approaches. source
- Limit: connector parsing
- A Reddit commenter reported BigQuery connector problems parsing queries containing a LIMIT clause and difficulty scaling component-based designs. source
- Limit: maturity issues
- An independent migration article listed GUI path-reset problems, repeated GitHub credential prompts, Hop Server ID and metrics issues, and inability to copy transforms in Hop Web. source
- Limit: native Spark
- Native Spark does not support sorted Group By, Beam-only I/O, barriers, result-row hops, Kafka consumer or streaming windows. source
- Limit: scheduler
- A Reddit user reported that Hop has no built-in scheduler, requiring Cronicle, cron or another external scheduler for scheduling jobs. source
- Maker legal status
- The Apache Software Foundation is a U.S. 501(c)(3) nonprofit incorporated in 1999. source
- Marketplace
- Optional plugins can be installed from Apache or Maven repositories through a GUI, CLI or declarative hop-env.yaml file. source
Checked against the maker's pages on 15 Oct 2026.