#2 · Primary category: NLP Tools & Text Processing

compromise

named-entity-recognition nlp part-of-speech

modest natural-language processing

Project last updated:08/23/26

GitHub Stars

12.1K

Forks

667

Contributors

126

License

MIT

Why we included this project

Compromise is a compact JavaScript library that parses sentences into parts of speech and named entities, then lets you reshape the text by conjugating verbs, pluralizing nouns, or expanding contractions. It relies on a hand-built lexicon of about 14,000 words and pattern matching rather than a neural network, so it runs fast enough for keypress-level interaction and works in both the browser and Node. That makes it a good fit for client-side features like autocomplete and text normalization, or for a quick grammar check without the overhead of a deep-learning model. Small teams can get usable results with minimal dependencies.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category