Meilisearch Engine

Meilisearch is an open-source, lightning-fast search engine created by Quentin de Quelen and Clementine Hurlin in 2018, headquartered in Paris, providing instant search capabilities with sub-fifty-millisecond response times on datasets of millions of records, written in Rust for native performance, featuring typo-tolerance, synonyms, stop words, filtering, sorting, faceted search, geo-search, and custom ranking, designed as a developer-friendly alternative to Algolia with an open-source core and managed cloud offering, adopted by over forty-five thousand GitHub stars and companies including Linux Foundation and Hunter for their search needs. The search performance is achieved through a Rust implementation with optimized memory management, parallel processing, and a custom prefix search algorithm that indexes all possible word prefixes at indexing time, enabling instant results as the user types each character, with the search returning results in under fifty milliseconds even for complex queries on large datasets, providing a truly instant search experience without debounce delays. The typo-tolerance automatically handles spelling mistakes through a Damerau-Levenshtein distance algorithm that finds results even when the query contains typos, with configurable typo strictness per attribute, and support for typo-tolerant prefix search meaning that partial words with typos still return relevant results. The indexing pipeline processes data uploads through a streaming architecture that enables searching during indexing, with incremental updates that do not require reindexing the entire dataset, and automatic language detection for applying the correct tokenizer for over ten languages. The filtering and faceting enable drill-down navigation. The custom ranking rules. The synonyms and stop words configuration. The geo-search for location-based queries. The self-hosted open-source edition. The cloud managed hosting. Designed for developers needing fast search.