Pentaho Data Integration (PDI), also known as Kettle, is a long-standing Java ETL tool where you design transformations and jobs in a desktop designer. Airbyte is a data movement platform with 700+ prebuilt connectors that load SaaS, API and database data into any destination, with a free self-hosted edition and managed cloud. If you're maintaining hand-built PDI jobs to pull data from SaaS apps and databases into a warehouse, Airbyte replaces that work with maintained connectors. Pentaho remains a fit for teams invested in complex, on-prem ETL logic built in Spoon.
Airbyte vs Pentaho at a glance
| Airbyte | Pentaho Data Integration |
|---|
| Category | Data movement (ELT and CDC) platform | Java-based ETL tool (Kettle) |
| Ownership | Independent | LEO Software (Constellation Software) since June 2026, previously Hitachi Vantara |
| Prebuilt connectors | 700+, including SaaS APIs, databases, files and unstructured sources | Steps for databases, files and big data; SaaS APIs usually need custom steps or REST calls |
| Building a pipeline | Configure source and destination in UI, API or Terraform | Design transformations and jobs in the Spoon desktop designer |
| Transformation | In the destination with dbt or SQL (ELT) | In-flight, step by step, before loading (ETL) |
| Schema changes | Detected and propagated automatically | Handled in transformation logic you maintain |
| Change data capture | Log-based CDC on every plan | Not built in as log-based CDC; typically timestamp or custom logic |
| Pricing | Free (Core); Cloud from $20/month; capacity-based Data Workers on Pro and Enterprise Flex | Community Edition free; Enterprise Edition quote-based |
| Deployment | Self-hosted (Core), managed cloud, hybrid or on-prem (Enterprise Flex) | Self-managed servers (Carte, Pentaho Server) on-prem or in your cloud |
| Managed cloud option | Airbyte Cloud | None; you run the servers |
| Python and API access | PyAirbyte, Airbyte API, Terraform provider | Java plugins, Carte REST endpoints, command-line Pan and Kitchen |
| Open source | Yes | Community Edition; the community fork Apache Hop is now developed separately |
| Community | 25,000+ members, 1,000+ contributors | Long-standing forums; much of the contributor community has moved to Apache Hop |
| Who can use it | Analysts (UI) and engineers (API, Terraform, PyAirbyte) | ETL developers comfortable with Spoon and Java |
About Airbyte
Airbyte is the open standard for data movement, used by 18% of the Fortune 500. It replicates data from 700+ sources into warehouses, lakehouses, databases and AI context stores, and runs self-hosted, fully managed, or hybrid inside your own boundary.
About Pentaho
Pentaho Data Integration (PDI), originally the open-source Kettle project, is a Java ETL tool. Developers design transformations (row-level data flows) and jobs (control flows) in the Spoon desktop designer, then run them on Carte or Pentaho Server, or from the command line with Pan and Kitchen. Pentaho also sells business analytics, a data catalog and a data optimizer as the Pentaho+ platform. Version 11, released in February 2026, modernized the user experience.
Ownership: another change in 2026
Pentaho was founded in 2004 and grew around open-source business intelligence and the Kettle ETL engine. Hitachi acquired it in 2015 and ran it within Hitachi Vantara. In June 2026, Pentaho was reportedly sold to LEO Software, part of Constellation Software, followed by restructuring.
Many of the original Kettle contributors had already moved on to Apache Hop, a community fork of PDI now developed under the Apache Software Foundation. For teams with years of PDI jobs, two ownership changes and a split community are a good reason to review roadmap, support terms and long-term skills availability before the next renewal.
Pentaho Community Edition is free to download and self-host, with community support. Enterprise Edition adds commercial support, enterprise connectors, security features and additional tooling, and is quote-based.
Airbyte Core is free to self-host and actively maintained, with 1,000+ contributors and 25,000+ community members. Paid cloud and enterprise plans add managed infrastructure, support and governance on the same codebase, so moving from free to supported doesn't mean switching products.
In PDI, each pipeline is a transformation you design step by step: input steps, lookups, joins, calculations and outputs. That suits complex in-flight transformation, but pulling data from a modern SaaS API usually means building REST calls, pagination, rate-limit handling and incremental logic yourself, then maintaining them every time the API changes. Schema changes upstream often break jobs until someone updates them.
Airbyte connectors ship with authentication, pagination, incremental syncs, CDC and schema propagation built in, and are maintained as source APIs change. Data lands in your warehouse, and transformation happens downstream in dbt or SQL, where it's version-controlled and tested. When a source is missing, the Connector Builder creates one from API docs with AI assist.
Change data capture
PDI doesn't include log-based CDC out of the box. Teams typically use timestamp columns, full reloads or custom logic, which adds load on source databases and can miss deletes.
Airbyte includes log-based CDC on every plan, including free self-hosted Core, for databases such as Postgres, MySQL, SQL Server, Oracle and MongoDB, capturing inserts, updates and deletes with minimal load on production systems.
Pricing
Pentaho Community Edition is free; Enterprise Edition pricing is quote-based, and you also pay for the servers that run it and the engineers who maintain jobs.
Airbyte Core is free to self-host. Cloud starts at $20/month (Standard) and $189/month (Plus). Pro and Enterprise Flex price on Data Workers, dedicated compute capacity, so cost doesn't depend on row counts.
Deployment
Pentaho runs on servers you manage, on-prem or in your own cloud, with no vendor-managed cloud option.
Airbyte runs four ways on one codebase: free self-hosted Core, managed cloud, hybrid with an Airbyte-managed control plane, or fully on-prem with Enterprise Flex. It writes to Snowflake, BigQuery, Databricks, Redshift, Postgres, S3, Iceberg lakehouses and vector stores.
Migrating from Pentaho to Airbyte
Most PDI estates mix two kinds of work: moving data from sources into a warehouse, and transforming it along the way. A practical migration separates them:
1. Inventory PDI jobs and transformations, and tag each as data movement, transformation, or both.
2. Replace data movement with Airbyte connectors, using CDC for database sources.
3. Rebuild transformation logic as dbt models or SQL in the warehouse.
4. Run old and new pipelines in parallel, compare row counts and key metrics, then cut over source by source.
Where Pentaho leads
Pentaho is a strong fit for teams with large, working PDI estates, complex in-flight transformations built visually, and on-prem environments where reporting and ETL already run together on the Pentaho platform.
When to choose Airbyte vs Pentaho
| Your situation | Better fit |
|---|
| You need SaaS and API data without building REST steps and pagination yourself | Airbyte |
| You load into cloud warehouses or lakehouses and transform with dbt or SQL | Airbyte |
| You need log-based CDC from production databases | Airbyte |
| You want an actively maintained open-source option with a managed cloud path | Airbyte |
| You want published pricing | Airbyte |
| You're concerned about Pentaho's ownership changes and roadmap | Airbyte |
| You have a large, working PDI estate with complex in-flight transformations | Pentaho |
| You run on-prem reporting and ETL together on the Pentaho platform | Pentaho |
Build with Airbyte
Ship pipelines in minutes, not sprints. Try Airbyte free or talk to sales.
