Skip to content
OpenSourceSoftware.ai
Meilisearch logo

Meilisearch

meilisearch · Search & retrieval

GitHub stars

59,684
—Awaiting first sync

Awaiting first sync

No measured star history yet

What is Meilisearch?

Meilisearch is a search engine designed to drop into websites and apps without a dedicated search team. It is fast enough for search-as-you-type, forgiving of typos, and flexible enough for faceted product search or semantic search.

Meilisearch is an open-source search engine for fast full-text, semantic and hybrid search. It returns results as users type, tolerates typos, and supports filters, facets, sorting and synonyms. It can be self-hosted from a single binary or Docker image, or used through Meilisearch Cloud.

Core capabilities

  • Return relevant results as users type, in under 50 milliseconds.
  • Combine full-text and semantic search, with typo tolerance and synonyms.
  • Build filters, facets and sorting for a product catalogue in a few lines of code.

Relevant results while people type

Search-as-you-type returns results in under 50 milliseconds, typo tolerance still finds matches for misspelled queries, and synonyms widen results to the terms people really use. Hybrid search combines semantic and full-text matching.

  • Search-as-you-type
  • Typo tolerance
  • Hybrid search

Filter, facet and sort

Custom filters and faceted search let users narrow a catalogue by attributes such as category or brand, and results can be sorted by price, date or other fields. Quick starts cover JavaScript, Python, PHP, Ruby, Go, Rust and React.

  • Filtering
  • Faceted search
  • Sorting
  • SDKs

Self-host or use the cloud

The same engine and API run in Meilisearch Cloud or self-hosted from a single binary or the Docker image, so a project can start on one and move to the other. Most of the code is MIT-licensed, while enterprise features fall under the Business Source License.

  • Self-hosted
  • Docker
  • MIT

Where it fits

Use cases

  1. 01

    Add instant search to an online store

    Index the product catalogue and let shoppers search as they type, with filters and facets for category or brand and sorting by price or date.

  2. 02

    Search a documentation or content site

    Index articles or docs so readers find the right page even when they misspell a term, with synonyms for the words your audience actually uses.

  3. 03

    Power semantic search

    Combine semantic and full-text matching in hybrid search to return results that match meaning as well as keywords.