English Sentence Analyzer

What does the analyzer show?

The analyzer colors each token by predicted part of speech, then shows estimated morphology (case, gender, and number when available), its dependency role and head, and nominal phrase summaries. A compact TensorFlow.js model and rule-based fallback run in the browser; results are probabilistic and can be wrong for ambiguous, unusual, or complex sentences.

Where is the sentence processed?

Your text stays in this browser. The quick rules and TensorFlow.js language model run locally without sending the sentence to an analysis service.

Model data and license

The English model was prepared from the Universal Dependencies English EWT, CHILDES, Atis, and ESLSpok treebanks.

• Universal Dependencies English EWT
• Universal Dependencies English CHILDES
• Universal Dependencies English Atis
• Universal Dependencies English ESLSpok

Dataset release 2.18. The treebanks and derived model data are provided under CC BY-SA 4.0. Attribution belongs to the Universal Dependencies contributors and the cited treebank projects.