apache/lucene

What it is

Apache Lucene is a high-performance, full-featured text search engine library written in Java (README). It provides indexing and querying primitives for full-text search and information retrieval; applications embed it rather than run it as a standalone server. It suits Java developers building search into their own products, and it underlies several widely used search platforms.

Beyond classic inverted-index search, Lucene's scope includes approximate nearest-neighbour vector search, and the project has explored using the Java Panama Vector API to accelerate those vector operations (HN discussion). Commenters there noted the distinction between vector search (ANN) and the Vector API (SIMD), and flagged that the Panama API was incubating and required specific JVM flags at the time (comment).

Quick start (build from source)

Prerequisites: JDK 25, installed via your package manager or a JDK provider such as OpenJDK or Adoptium.

1. Clone the repository or download the source distribution.
2. From the repository root, run the Gradle wrapper:

./gradlew

The README only covers this basic build setup and points to the latest release documentation and nightly Javadoc for full API guidance. To use Lucene as a dependency, consult the release documentation for the artifact coordinates rather than assuming a version.

Constraints and compatibility

• Building from source requires JDK 25 per the README (build steps); the project does not document older-JDK builds there.
• Lucene is a library, not a service: you must write the indexing and querying code in your application.
• The vector-search acceleration discussed in 2023 depended on an incubating Java API and specific JVM flags, and the surrounding vector index merging has its own operational considerations (comment). Treat those details as version-specific.
• Migration between major versions is not automatic; the project maintains a migration guide.
• Licensed under Apache-2.0 (license).

Community and support

Usage questions go to the Java user mailing list; development discussion happens on the developer lists. New contributors should read the contributing guide, and build-system details live in the help/ directory.

HotRepo's take

Lucene appeared on HotRepo's Java daily list on 2026-11-06 at position 20 with no star change that day, and it has no recorded all-language daily rank. The more informative signal is the 2023 HN thread on Panama Vector API integration (82 points, 16 comments), where practitioners debated whether Lucene's ANN work could make it a serious alternative to dedicated vector databases (thread). That discussion is about a particular integration and is now several years old, so it reflects interest in Lucene's vector direction rather than current performance. The repository's own record shows a mature, steady project rather than a fast-rising one.