Extract a clean recipe from any URL
Recipe pages bury the good part under ads, pop-ups, and a long backstory. CookSift pulls out the recipe itself: ingredients with quantities and units, numbered steps, and timing.
It reads the structured recipe data the page publishes — schema.org JSON-LD first, then microdata and conventional Ingredients and Instructions lists — so it works across sites rather than being hard-coded to one.
What gets extracted
- Title, image, and source link — with attribution kept.
- Ingredients with quantity and unit.
- Numbered instructions.
- Servings and prep/cook time when the page includes them.
- Nutrition with a label for every value.
How it reads the page
CookSift fetches the public page server-side — no cookies, no login, no paywall bypass — and parses the recipe markup. If a site publishes schema.org recipe data, CookSift reads that directly.
Nutrition you can verify
Every nutrition value is labelled: from the source, estimated from USDA FoodData Central ingredient data, or unknown. You always know where a number came from.
Turn recipe links into one grocery list
Paste a public recipe URL and CookSift extracts clean ingredients, steps, and nutrition — then combines saved recipes into one grocery list.
Paste a recipe →Frequently asked questions
Does it work with any recipe site?
It works with any public page that uses standard recipe markup. If a page hides the recipe behind a login or paywall, CookSift can’t read it.
Do I need to type the recipe by hand?
No. Paste the URL and CookSift extracts the recipe for you.
What if the page has no nutrition info?
CookSift may estimate it from USDA FoodData Central ingredient data and labels it estimated, or marks it unknown.