Skip to content

Summary

Official MCP server for Meilisearch — create and configure indexes, add documents, run hybrid and semantic searches, and manage API keys and tasks through conversation.

Features

  • Create, list, delete and measure Meilisearch indexes
  • Add and retrieve documents without writing client code
  • One search tool covering filters, sorting, facets, multi-index, semantic and hybrid modes
  • Read and update index settings — searchable and filterable attributes, ranking rules, synonyms
  • Mint, list and revoke API keys
  • Track, cancel and delete asynchronous tasks
  • Health, version, stats and system-info tools for monitoring a running instance

Installation

Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.

Any MCP-compatible agent

Most agents (Claude, Cursor, Windsurf, VS Code, and more) read a standard mcpServers configuration.

  1. Open your agent's MCP configuration file.
  2. Merge the snippet below into it, filling in the environment variables with your own values.
  3. Restart the agent — the "Meilisearch MCP Server" tools will be available.
{
  "mcpServers": {
    "meilisearch-mcp": {
      "command": "uvx",
      "args": [
        "-n",
        "meilisearch-mcp"
      ],
      "env": {
        "MEILI_HTTP_ADDR": "http://localhost:7700",
        "MEILI_MASTER_KEY": "your_master_key"
      }
    }
  }
}

Description

The Meilisearch MCP Server is the search engine's own MCP integration. It puts the administrative half of Meilisearch — the part you would normally drive with curl against the HTTP API — behind a set of tools an assistant can call, which makes it useful both for operating a running instance and for prototyping a search setup without writing client code first.

Tool groups
  • Connection — get-connection-settings, update-connection-settings, so you can repoint the server at a different instance mid-session.
  • Indexes — create-index, list-indexes, delete-index, get-index-metrics.
  • Documents — get-documents, add-documents.
  • Search — a single search tool covering filtering, sorting, facets, multi-index queries, and Meilisearch's semantic and hybrid modes.
  • Settings — get-settings, update-settings for searchable/filterable attributes, ranking rules, synonyms and the rest.
  • Keys — get-keys, create-key, delete-key.
  • Tasks — get-task, get-tasks, cancel-tasks, delete-tasks, matching Meilisearch's asynchronous task model.
  • Monitoring — health-check, get-health-status, get-version, get-stats, get-system-info.
Setup

Python 3.9 or newer, plus a reachable Meilisearch instance. uvx -n meilisearch-mcp is the recommended launch (no install step); pip install meilisearch-mcp works too. Point it with MEILI_HTTP_ADDR, defaulting to http://localhost:7700, and give it MEILI_MASTER_KEY if the instance is secured.

Worth knowing

create-key and delete-key mean the server can mint and revoke credentials for the instance it is pointed at, so scope the master key you hand it accordingly — or run it against a development instance. MIT licensed, maintained in the Meilisearch GitHub organisation.

Typesense's official remote MCP server — search collections, index documents, and create, resize or clone the Typesense Cloud clusters behind them, over OAuth with per-agent permission presets.

Featured

Weaviate's MCP server, built into the database itself — hybrid search, schema inspection, tenant listing and batch upserts at /v1/mcp, scoped by RBAC.

MCP: XERJ

by XERJ

New

Local-first search engine and MCP server in one Rust binary: indexes code, docs, logs and PDFs for full-text, vector and hybrid search plus agent memory.

Databricks-hosted MCP endpoints that expose Unity Catalog functions, Genie spaces, Vector Search indexes and Databricks SQL to agents, with UC permissions enforced.

Browse all MCP servers →