Delta Lake versus Iceberg: which is winning?

Summary

• There's no official scoreboard for table formats. As of October 2026, Delta Lake and Apache Iceberg both ship regular releases under the Apache License 2.0, and both document support across major open-source engines.
• Delta Lake is a Linux Foundation project, run under a charter that gives technical oversight to a Technical Steering Committee whose voting members are the project's maintainers. Apache Iceberg is an Apache Software Foundation top-level project, run by a Project Management Committee.
• Each format has a public specification with its own versioning model: Delta Lake uses protocol versions plus named table features, and Iceberg uses numbered format versions, with versions 1, 2 and 3 complete.
• The practical test is fit: which engines in your stack read and write each format, which spec features they support, and which catalog API you'll standardize on.
• Databricks' public position supports both formats: Unity Catalog manages Delta Lake and Apache Iceberg tables and serves Iceberg tables to external engines through an Iceberg REST catalog.

Delta Lake versus Iceberg: which is winning?

There's no documented scoreboard, and the record the two projects publish shows active development on both sides. As of October 2026, the latest releases are Delta Lake 4.4.1 and Apache Iceberg 1.12.0, each project sits in an open-source foundation, and each lists Apache Spark, Apache Flink and Apache Hive among the engines it works with. A ranking can't tell you which one fits your stack. Governance, specification, engine coverage and catalog fit can, and the facts for each follow.

Who governs each project

Delta Lake's site states that it is "an independent open-source project and not controlled by any single company," and that it joined the Linux Foundation in 2019. Its technical charter, adopted September 17, 2019, establishes the Delta Lake Project as a Series of LF Projects, LLC. The charter makes a Technical Steering Committee responsible for all technical oversight, names the project's maintainers as its voting members, and says its meetings are intended to be open to the public.
Apache Iceberg became an Apache Software Foundation top-level project in May 2020. A Project Management Committee (PMC) oversees it under ASF processes, and the community holds regular development syncs, including subproject syncs for Python, Rust and Go and a catalog community sync.

Delta Lake vs Apache Iceberg: ecosystem and governance facts

Each cell comes from that project's own site, specification, charter or release notes.

Fact	Delta Lake	Apache Iceberg
Foundation	Linux Foundation (LF Projects, LLC), since 2019	Apache Software Foundation, top-level project since May 2020
Technical decisions	Technical Steering Committee; voting members are the maintainers	Project Management Committee
License	Apache License 2.0	Apache License 2.0
Specification	Delta Transaction Log Protocol	Iceberg Table Spec
How the spec evolves	Reader and writer protocol versions plus named table features	Numbered format versions; versions 1, 2 and 3 are complete and adopted
Latest release, as of October 2026	Delta Lake 4.4.1 (October 5, 2026)	Apache Iceberg 1.12.0 (September 29, 2026)
Engines the project lists	Spark, Flink and Hive among others, with APIs for Scala, Java, Rust and Python	Spark, Flink, Hive and Impala among others
Native libraries	Delta Kernel (Java and Rust); delta-rs (Rust, with Python bindings)	Java libraries, plus PyIceberg, Iceberg Rust, Iceberg Go and Iceberg C++
Catalog model	Self-describing log, so no separate metastore is needed to read a table (catalog-managed tables excepted)	REST Catalog API published as an OpenAPI specification

How to read these facts for your own decision

• Spec features and reader support. Both specs gate new capabilities. Delta Lake readers "must implement and respect all features listed in readerFeatures," and Iceberg increments its format version when older readers "would not read newer table features correctly." Before you enable a feature, confirm every engine that reads the table supports it.
• Engine versions. Release notes show what each project targets. Iceberg 1.12.0 ships runtime JARs for Spark 3.5, 4.0 and 4.1 and for Flink 1.20, 2.1, 2.2 and 2.3. Delta Lake 4.4.0 added support for Apache Spark 4.2. Match these against the versions you run.
• Governance fit. If your team plans to contribute to the spec, read the decision process you'd join: an ASF PMC for Iceberg, or Delta Lake's TSC and the contribution rules its charter points to.
• Catalog standard. Engines find and commit tables through a catalog, so decide which catalog API your engines will share before you weigh formats in the abstract.

Interoperability changes the stakes

The two formats don't exclude each other. Delta Lake's Universal Format (UniForm) generates Iceberg metadata for a Delta table so Iceberg clients can read it, and one copy of the Parquet data files serves both. Apache XTable, incubating at the ASF, translates table metadata among Delta Lake, Iceberg and Apache Hudi without introducing a new format.

Databricks' position: both formats in Unity Catalog

In Unity Catalog, a managed table can be Delta Lake or Apache Iceberg, and the docs recommend managed tables for both. Databricks lists support for Iceberg spec versions 1, 2 and 3. External engines can use Unity Catalog's Iceberg REST catalog to read and write its Iceberg tables, and the docs name Apache Spark and Apache Flink among supported clients. Delta tables can serve Iceberg clients too: with Iceberg reads (UniForm) enabled, a Delta table generates Iceberg metadata automatically and its data files aren't rewritten. Announcing full Iceberg support in June 2025, Databricks wrote that "the Delta and Apache Iceberg communities are aligning concepts and contributions."

FAQs

Is Delta Lake controlled by Databricks?

Delta Lake's site describes it as "an independent open-source project and not controlled by any single company." It has been a Linux Foundation project since 2019, and its charter assigns technical oversight to a Technical Steering Committee of maintainers.

Are Delta Lake and Iceberg licensed differently?

No. Both projects release their code under the Apache License 2.0.

Do engines have to pick one format?

Not necessarily. Apache Spark, Apache Flink and Apache Hive appear in both projects' integration lists, and Delta Lake's Universal Format lets Iceberg clients read Delta tables.