quickwit-oss/tantivy

Health v2 · maintenance only

Adoption not part of the score

15992 stars · 980 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Tantivy is a fast full-text search engine library written in Rust, strongly inspired by Apache Lucene's design. It is a crate for building search functionality into applications rather than an off-the-shelf search server.

Use cases

• add full-text search to a Rust application
• build a custom search engine server
• index and query documents with BM25 scoring
• implement phrase and boolean query search
• embed fast search in a command-line tool
• build an alternative to Elasticsearch or Solr

When to choose

• you need a fast, embeddable full-text search engine in Rust
• you want Lucene-like features (BM25, natural query language, phrase queries) without running a separate server
• you need tiny startup time (<10ms) for CLI tools
• you want configurable tokenizers with stemming for many languages

When to avoid

• you need a ready-made distributed search server out of the box (use Quickwit or Elasticsearch instead)
• your project is not in Rust and you don't want FFI bindings
• you only need simple substring matching without indexing

Facets

library · maturity stable

search-engine parser developer-tools rust cross-platform full-text-search lucene-inspired bm25 inverted-index information-retrieval search algorithms

Member repositories

For agents

markdown · JSON · MCP: product_card(name="quickwit-oss/tantivy")