Mealie¶

Mealie is a self-hosted recipe manager and meal planner with a REST API backend and a modern, reactive frontend.

What it is¶

Mealie is a comprehensive culinary management system that allows users to import recipes from the web, organize their collection, create meal plans, and generate shopping lists in a centralized, private environment. As of July 2026, it features native integration with Claude 4.8 and Gemma 3 for AI-powered recipe scaling, nutrition analysis, and autonomous grocery orchestration via MCP 3.0.

What problem it solves¶

Keeping track of digital recipes often involves scattered bookmarks, screenshots, or reliance on third-party SaaS platforms filled with ads and tracking. Mealie solves this by providing a unified, self-hosted vault where recipes are parsed into a clean, consistent format, making them easy to search, scale, and plan for the week.

Where it fits in the stack¶

Service / Home Automation. It sits in the personal lifestyle and planning layer of the self-hosted stack, connecting culinary interests with inventory management and grocery logistics. It can be integrated with Grocy for deeper pantry management and n8n for automated grocery list syncing and Agentic Workflows.

Typical use cases¶

Recipe Archival: Importing and organizing thousands of recipes from various websites into a clean, ad-free format.
Weekly Meal Planning: Planning breakfast, lunch, and dinner for the household using a visual calendar.
Automated Shopping Lists: Generating consolidated shopping lists based on a weekly meal plan.
Recipe Scaling: Automatically adjusting ingredient quantities for different serving sizes using Gemma 3.
AI Ingredient Extraction: Using Claude 4.8 or GPT-5.5 to extract ingredients from unstructured text or voice notes.

Strengths¶

Superior Parsing: Highly accurate recipe scraping from almost any URL using the recipe-scrapers library.
Mobile Friendly: The web interface is fully responsive and behaves like a native app on mobile devices.
Extensive API: Every feature is exposed via a REST API and MCP 3.0 server, enabling deep integration with other home automation tools.
Multi-User: Supports multiple users with shared or private recipe collections and meal plans.
AI Video Import: Supports recipe imports from YouTube and TikTok via transcription and analysis.

Limitations¶

Database Dependency: Requires a PostgreSQL or SQLite database, which adds some complexity to the deployment compared to flat-file managers.
Inventory Depth: While it has basic food inventory features, it is not as exhaustive as specialized tools like Grocy.
VRAM for Local AI: If using local Whisper models for video transcription, significant GPU resources are required.

When to use it¶

If you have a large collection of recipes you want to digitize and organize.
If you want a self-hosted alternative to Paprika, Yummly, or AnyList.
When you want to automate your meal planning and shopping list generation.
To maintain privacy by keeping your family's dietary habits on your own hardware.

When not to use it¶

For simple shopping lists without recipe integration (consider Vikunja).
If you need a full enterprise-grade kitchen management system (though it’s great for home use).
If you prefer physical cookbooks and do not need a digital archival system.

Getting started¶

Installation (Docker Compose)¶

Mealie v3.19.0 (July 2026) supports both SQLite and PostgreSQL backends and features an integrated MCP 3.0 server.

services:
  mealie:
    image: ghcr.io/mealie-recipes/mealie:latest
    container_name: mealie
    restart: unless-stopped
    ports:
      - "9925:9000"
    volumes:
      - ./mealie-data:/app/data
    environment:
      - ALLOW_SIGNUP=false
      - PUID=1000
      - PGID=1000
      - TZ=Europe/London
      - MAX_WORKERS=1
      - WEB_CONCURRENCY=1
      - BASE_URL=http://localhost:9925
      - DEFAULT_EMAIL=admin@example.com
      - DEFAULT_PASSWORD=YourStrongPassword
      - DEFAULT_GROUP=Home
      - DEFAULT_HOUSEHOLD=Family

Groups and Households¶

Mealie utilizes a two-tier user model for multi-tenant or multi-family deployments: - Groups: Isolated tenants (no shared data between groups). - Households: Subdivisions within a group. Members share recipes but have separate meal plans and shopping lists.

AI Video Import (YouTube, TikTok)¶

Mealie supports AI-powered recipe imports from social media videos using the Whisper model or cloud APIs. - Workflow: Paste a YouTube or TikTok URL into the import field. - Backend: Transcribes video and structures ingredients/steps using GPT-5.5 or Claude 4.8. - Setup: Requires an API key configured in Settings -> Integrations.

CLI examples¶

Mealie is primarily managed via the web UI or API, but you can interact with the container for maintenance.

# Check application logs
docker logs -f mealie

# Inspect the Mealie container environment
docker inspect mealie --format='{{range .Config.Env}}{{println .}}{{end}}'

# Perform a manual data backup (if using SQLite)
docker exec mealie tar -czf /app/data/mealie_backup.tar.gz /app/data/mealie.db

API examples¶

Get all recipes (Python)¶

Mealie provides a comprehensive API and MCP 3.0 Task Protocol support.

import requests

MEALIE_URL = "http://localhost:9925/api"
API_TOKEN = "your-api-token"

headers = {
    "Authorization": f"Bearer {API_TOKEN}"
}

response = requests.get(f"{MEALIE_URL}/recipes", headers=headers)
if response.status_code == 200:
    for recipe in response.json()['data']:
        print(f"Recipe: {recipe['name']}, Rating: {recipe['rating']}")

Fetch a specific recipe (Curl)¶

curl -H "Authorization: Bearer YOUR_API_TOKEN" \
     "http://localhost:9925/api/recipes/your-recipe-slug"

Grocy — For more detailed food and pantry inventory management.
Home Assistant — For displaying meal plans on home dashboards.
Vikunja — For managing complex grocery shopping tasks.
Linkwarden — For bookmarking recipe ideas.
Actual Budget — For tracking the financial cost of meal plans.
Whisper — For local video-to-text transcription.
n8n — For automating shopping list synchronization.
Paperless-ngx — For archiving scanned physical recipe cards.
Model Context Protocol (MCP) — For agentic culinary orchestration.

Sources / references¶

Contribution Metadata¶

Last reviewed: 2026-07-21
Confidence: high