Everything you need to know about myzest.app, how it works, and how it differs from other apps.
myzest.app is a personal recipe book with soul — a culinary diary that treats food as experience, not as data. You save recipes, link memories (photos, dates, locations) to them, and share Moment Cards with your family. Available as a free cloud SaaS at myzest.app and as an open-source Docker image on GitHub for self-hosting.
Yes. Zest is free forever. No trial, no credit card. The Free plan lets you save up to 99 recipes, unlimited memories with photos, Moment Cards and share-by-link. Pro and Ultra plans are planned for users who need more capacity, but you don't need them today. The self-hosted version is completely free under MIT license.
Yes. The public repository is at github.com/MartinSantosT/myzest under MIT license. It is a single Docker image: FastAPI + SQLite, no external database, no Node.js, no build steps. Runs on a Raspberry Pi, a Proxmox LXC container, or any Linux box. See /selfhosted for deployment instructions.
Moment Cards are server-generated images (built with Pillow) that summarize a recipe plus its attached memory into a single visual card. You can share it via WhatsApp or email and the recipient sees the full recipe without needing to create an account. The point is to preserve the moment, not just the ingredients.
Mealie and Tandoor are excellent mature self-hosted recipe managers focused on inventory and meal planning. Zest is lighter (single container, SQLite, no external DB), has simpler setup, and is explicitly designed around the "memory" dimension of cooking. Paprika is a closed commercial app; Zest is open source on self-host and free on cloud. They are all valid choices depending on your use case — Zest doesn't aim to replace them, just to offer an alternative with a different focus.
Zest was built by Martin Santos, a systems engineer with experience in restaurant management, based in Mexico. It's a personal project publicly launched in March 2026. The self-hosted code is open on GitHub for community contribution.
Backend: FastAPI (Python) + PostgreSQL (Supabase) on the cloud version; SQLite on the self-hosted version. Frontend: vanilla JavaScript (ES Modules) + Tailwind CSS, no Node.js or build steps. Docker deployment. 4-tier recipe scraper: JSON-LD → Microdata → Schema.org heuristics → LLM fallback. Everything designed to be lightweight and easy to maintain.
On the cloud version (myzest.app), authentication uses HttpOnly cookies with SameSite=Lax and all traffic is HTTPS. On the self-hosted version, 100% of your data lives on your own hardware — neither Martin nor anyone else sees it. The auditable code is on GitHub.
Yes. Zest's scraper works in 4 levels: first it tries JSON-LD (the schema.org/Recipe standard used by most serious food blogs), then Microdata, then heuristics, and as a final fallback an LLM extracts the data. Paste a URL and the recipe is imported with clean ingredients and steps, without the ads or the blogger's personal story.
Still have questions?
We'll get back to you soon.