sense2vec is an open-source Python library from Explosion for word vectors. A word vector is a list of numbers that represents a word, learned from large amounts of text so that words used in similar contexts get similar vectors; word2vec is the best-known method. sense2vec (Trask et al., 2015) extends word2vec: each key combines a word or multi-word phrase with its sense, a part-of-speech tag or an entity label, so "natural_language_processing|NOUN" has its own vector.
A program loads a vector package from disk and asks for the vector of a key, its frequency, or its most similar keys; for natural language processing, the nearest keys include machine learning and computer vision. The library also works as a component inside spaCy pipelines, and it can train new vectors from raw text with GloVe or Word2Vec via fastText. Pretrained vectors trained on Reddit comments are offered as downloads: s2v_reddit_2015_md (573 MB, comments from 2015) and s2v_reddit_2019_lg (4 GB, 2019).
sense2vec is free and open source under the MIT License and installs with pip.