Apache OpenNLP is a library from the Apache Software Foundation that its site describes as a machine learning based toolkit for the processing of natural language text. It is aimed at developers who need the standard building blocks of a text pipeline inside their own software. Volunteers develop it, and it is free under the Apache License 2.0.
The toolkit covers sentence segmentation, tokenization, lemmatization, part-of-speech tagging, named entity extraction, chunking, parsing, language detection and coreference resolution. Pre-trained models are published for download: 36 language-specific models for sentence detection, tokenization, lemmatization and part-of-speech tagging, plus one language detector that recognizes 103 languages. Developers can also train models on their own data. The project documents Maven and Gradle setup, which points to use from Java and other JVM languages.
The download page lists two current lines, version 2.5.12 and a 3.0.0 milestone build, so the project is still being released. It is a classic statistical NLP toolkit: the site makes no mention of large language models or neural network runtimes, and teams that need generative features will have to pair it with other software. There is no hosted service, pricing or commercial support from the project itself.
