sumy

Why we included this project

Sumy is a Python library and command line tool that produces extractive summaries by picking the sentences that matter most, which makes it a light option when you need to shrink a long document before it reaches the rest of your pipeline. It implements several classical algorithms, LSA, LexRank, Luhn, Edmundson, and TextRank, so you can try a couple and keep the one that fits your document style instead of being locked into a single method. It handles both plain text and HTML pages, which helps when you want to condense web articles or feeds before passing them along. The tokenizers, stemmers, and parsers are all replaceable, so supporting another language or changing how documents are read is a matter of subclassing rather than reworking the project. That modularity, plus a simple sentence-count API, makes it attractive for teams that want proven, well-understood summarization without dragging in a large model.