Self-Hosted / Projects / MeiliSearch
MeiliSearch
StableSpecs verified October 7, 2026 · against vv1.37
Overview
Self-hosted search engine API for fast, typo-tolerant, and developer-friendly application search.
Meilisearch is a self-hosted search engine and API for applications that need fast on-site search without sending their index and query workload to a hosted search vendor by default.
Why developers choose Meilisearch
The project targets a common implementation problem: building a responsive search experience with typo tolerance, filters, sorting, and relevant results without assembling an indexing and ranking system from scratch. It is a search backend for a product or website, not a general-purpose Google replacement for browsing the web.
What is included
Official sources cover document indexing, keyword search, typo tolerance, prefix search, filtering, faceting, sorting, highlighting, pagination, REST APIs, SDKs, and hybrid or AI retrieval capabilities. Search behavior is configured per index, so relevance tuning and data modeling still require application work.
Requirements and tradeoffs
Meilisearch can run as a binary or Docker container and persists data in its data directory. Current installation docs officially support Linux amd64 or arm64 with glibc 2.35+, macOS 14+, and Windows Server 2022+, with Docker and direct-download paths. Protect the master key, persist and back up the data directory, and size resources for index and query volume; the service does not replace an application database.
Best fit
Choose Meilisearch when you need embedded site or app search with a straightforward API and a self-hosted option. Elasticsearch or OpenSearch may fit larger distributed search and analytics workloads, while a hosted search service is simpler if you do not want to operate indexing infrastructure.
Interface previews
Screenshots of the MeiliSearch interface.
Feature support
| Feature | Support |
|---|---|
| Full-text search APIThe official product and repository describe Meilisearch as a search engine API for websites and applications. | Supported |
| Typo tolerance and prefix searchThese are core search behaviors highlighted in official product and documentation material. | Supported |
| Filtering, faceting, and sortingOfficial docs cover filters, facets, and sortable attributes for structured search experiences. | Supported |
| Search result highlighting and paginationThe API supports common UI-oriented result presentation features including highlighting and pagination. | Supported |
| REST API and SDKsMeilisearch provides a REST API and official or community SDK integrations for application languages. | Supported |
| Hybrid and AI retrievalCurrent product pages promote AI-powered and hybrid retrieval, but implementation depends on version, configuration, and the selected feature set. | Partial support |
| Docker deploymentThe official installation guide documents running Meilisearch with the official Docker image and a persistent data volume. | Supported |
| General-purpose web searchMeilisearch indexes application content; it is not a crawler-based public web search engine like SearXNG. | Not supported |
Community signals
- Stars
- 59.5k
- Forks
- 2.7k
- Open issues
- 321
- Last commit
- 3d ago
- Latest release
- v1.54.3
- Repo created
- Apr 2018
Includes open pull requests
1w ago
Refreshed nightly from the GitHub API.
Questions
What is Meilisearch used for?
Meilisearch is used to add fast search, filtering, faceting, typo tolerance, and relevant results to websites and applications through an API.
Is Meilisearch a self-hosted search engine?
Yes. You can run the Meilisearch binary or official Docker image on infrastructure you control and connect to it through its API.
Can Meilisearch replace Algolia?
It can cover many hosted application-search use cases with a self-hosted deployment, but compare ranking controls, operations, analytics, support, and scale requirements before switching.
Does Meilisearch support Docker?
Yes. The official installation documentation provides a Docker command using the getmeili/meilisearch image and a persistent data directory.
What operating systems does Meilisearch support?
Current official installation docs list Linux amd64 or arm64 with glibc 2.35+, macOS 14 or later, and Windows Server 2022 or later as supported environments.
Is Meilisearch the same as SearXNG?
No. Meilisearch is an application-search backend that indexes data supplied by your software. SearXNG is a metasearch engine that queries external search providers.
Does Meilisearch replace a database?
No. Meilisearch is a search index and API. Your application still needs a source of truth and a process to synchronize records into the index.
support // the lab
Found this write-up useful?
If it saved you time or a rebuild, you can support more practical homelab guides.