arunsupe/semantic-grep

Health v2 · maintenance only

Adoption not part of the score

1246 stars · 29 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

w2vgrep is a grep-like command-line tool that finds words semantically similar to a query using word2vec embeddings, with familiar grep options like context lines, line numbers, and similarity thresholds. It supports multiple languages and reads from files or stdin.

Use cases

• search text for words with similar meaning instead of exact matches
• grep a document for synonyms of a concept like 'death'
• pipe text from curl or stdin and find semantically related terms
• search multilingual text with word embeddings from the terminal
• find paraphrased terminology in large text files

When to choose

• you want grep-style UX but need semantic rather than literal matching
• you work with plain text corpora and want lightweight local semantic search without an LLM
• you need a fast, offline tool driven by word2vec embeddings

When to avoid

• you need full semantic search over documents or passages rather than individual words
• you want transformer-based embeddings or contextual understanding
• you need to index and query large document collections with ranking

Facets

cli-tool · maturity active

search-engine nlp cli developer-tools cli go grep word2vec semantic-search text-processing word-embeddings command-line natural-language-processing linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="arunsupe/semantic-grep")