Meilisearch
A self-hosted search engine that combines full-text and semantic search, stores vectors for RAG, and integrates with LangChain and MCP.
Meilisearch combines keyword search and AI retrieval in a search engine you can host on your own server. It's for developers building search into websites, applications, product catalogs, or internal data tools. Meilisearch Cloud provides a separate, fully managed hosted service.
Hybrid search combines full-text matches with semantic relevance, so results can account for both the words people type and their meaning. Search-as-you-type returns results as queries develop, while typo tolerance and synonyms help match misspellings and alternative terms. Filters, facets, sorting, and geographic search let users narrow results by attributes or location. Federated search queries multiple data sources together.
For AI applications, Meilisearch stores and retrieves vectors for similarity search and retrieval-augmented generation (RAG). It also supports searching images, video, and audio through embeddings. Conversational search produces AI answers grounded in search results, and integrations with LangChain and the Model Context Protocol (MCP) connect retrieval to AI tools.
The engine uses Rust and exposes a REST API, with SDKs for languages including JavaScript, Python, and Swift. API keys provide granular data access permissions, and tenant-specific search supports applications serving separate customer groups. Enterprise code uses a commercial license or Business Source License 1.1 and requires a commercial agreement for production use. Meilisearch collects anonymized telemetry, which you can disable.
