55,000+ Recipes and Counting: A RecipeScrape Milestone

RecipeScrape just crossed 55,000 recipes in the database. That's more than 55,000 structured, normalized, searchable recipes — every one with ingredients, instructions, nutrition data, and a link back to the original source.

To put that in perspective: when we first wrote about the architecture back in June, we had 7,166 recipes. The scraper ran nightly against 4 sites. Today we cover 12 source sites and the database grows by roughly 400–600 new recipes every day.

What Changed

The jump from 7,000 to 55,000 came from two things:

More sources. We added 5 new sites including BBC Good Food, NYT Cooking (public recipes only), Budget Bytes, Love and Lemons, and Once Upon a Chef. Each new source adds thousands of recipes and fills cuisine and dietary gaps.

Historical backfills. For the original 7 sites, the initial scrape only captured the most recent recipes. Adding sitemap pagination let us go back years — Allrecipes alone contributed over 18,000 recipes going back to 2015.

Database Growth

Allrecipes     → 18,247
RecipeTin Eats →  4,830
Food Network   →  3,214
Simply Recipes →  2,901
BBC Good Food  →  9,422
NYT Cooking    →  5,108
Budget Bytes   →  1,876
...and 5 more  →  9,402+
─────────────────
Total          → 55,000+

The health endpoint now reads:

{
  "status": "ok",
  "recipeCount": 55000,
  "lastScrape": "2026-06-30T03:00:00Z"
}

Performance at Scale

Full-text search across 55,000 recipes still returns in under 80ms — PostgreSQL's GIN index scales logarithmically, not linearly. The Neon free tier handles the load without breaking a sweat. Filtering by cuisine, source, or category adds a B-tree lookup and stays under 100ms for most queries.

What's Next

The goal is 100,000 recipes by the end of the year. We're targeting regional sources — Australian, Indian, and Middle Eastern recipe sites — to diversify the cuisine coverage. We're also exploring ingredient-level search so you can ask "what can I make with chicken, garlic, and lemon?" and get ranked results.

Thanks to everyone who's used the API, reported issues, and contributed source sites. 55,000 recipes is a big number, but the database is just getting started.