Meilisearch Documentation

What Meilisearch is

Meilisearch is an open source search engine written in Rust, designed for search-as-you-type experiences. You push JSON documents into an index over a REST API and it gives you relevant results almost instantly, with sensible defaults: typo tolerance, ranking rules, filtering, sorting, faceted search and geosearch. Newer versions add vector and hybrid search, so keyword and semantic results can be combined.

What the docs cover

• Getting started: running Meilisearch locally (a single binary or Docker), adding documents and making your first search.
• Indexing: primary keys, updating documents and how indexing tasks run in the background.
• Relevancy: ranking rules, searchable attributes, synonyms, stop words and typo settings.
• Filtering and facets: filterable attributes and facet counts, the building blocks of a product listing page.
• Security: API keys and tenant tokens, so a front end can search without exposing your admin key.
• SDKs: official clients for JavaScript, Python, PHP, Go, Java and more, plus front-end widget libraries.

Is it worth learning?

For a student project, a small shop or an internal tool, yes. It is the fastest way to see what a dedicated search engine adds over LIKE '%term%' in SQL. Its defaults are good, so you spend time on your app rather than tuning analysers. For very large clusters or log analytics, Elasticsearch or OpenSearch are the more common choice.

How it fits with your database

Meilisearch is not your source of truth. Keep data in PostgreSQL, MongoDB or similar, and copy the fields people search into Meilisearch whenever a record changes. Plan how you will resync the index if it falls behind.

Prerequisites and time

Basic JSON and HTTP. An evening is enough to index a dataset of products or books and build a search page with filters.

Free or paid?

The Community Edition is free and MIT licensed. Meilisearch Cloud is a paid hosted service with a trial; an Enterprise Edition exists under a commercial licence.

Best for: web and full-stack developers who want good search in a side project or small product without running a heavy cluster.