Zenframe

AI shopping list from any recipe

You find a recipe, want it for Thursday dinner, and then have to extract the ingredients manually. The AI drops them straight into the shopping list — portion-scaled and sorted by aisle.

The problem families face

You find a recipe on a food blog, decide it's perfect for Thursday, and then spend the next five minutes manually copying each ingredient into a shopping list app — pausing to calculate that two cups of stock is roughly half a litre, checking whether you have paprika already, and wondering if the recipe means fresh or dried thyme. By the time you're done you've spent more effort on the list than on choosing the meal. Multiply this by five dinners a week and it adds up to a genuinely tedious recurring chore with no culinary payoff.

The manual transcription problem compounds when you're planning for a family rather than yourself. You need to scale the recipe from four servings to six, which involves fractional amounts for some ingredients. You need to cross-reference against whatever else is already on the list from Tuesday's dinner so you're not buying two packs of the same herb. And you need to account for what's already in the cupboard — which requires opening a second mental inventory alongside the recipe. That's four simultaneous tasks for what should be a simple process.

  • Manually copying ingredients from recipe to shopping list is time-consuming and error-prone
  • Scaling recipes from 4 to 6 portions requires mental arithmetic that frequently produces wrong quantities
  • Duplicate ingredients across multiple recipes in the same week list aren't consolidated automatically

Common ways families try to solve this today

The most common workaround is using a meal kit service like Gousto or HelloFresh, where the ingredients come pre-portioned alongside the recipe card. This fully solves the transcription problem for kit nights — but it only covers two or three of the week's dinners, limits you to the service's recipe library, and costs more per meal than buying ingredients independently. The three or four nights where you're cooking your own recipes still require manual list-building, and that's where the problem lives for most households.

Some families use grocery store apps that have built-in recipe search and ingredient addition — the Tesco Recipes feature, for instance, or BBC Good Food's integration with certain apps. These work neatly within their own ecosystems, but the recipes you actually want to cook rarely live in those ecosystems. Your grandmother's curry, the pasta dish from a food Instagram account, the Nigella recipe bookmarked three months ago — none of those are in a supermarket app's recipe database.

  • Meal kits (Gousto, HelloFresh, Mindful Chef): solve the problem on kit nights but not for self-chosen recipes
  • Supermarket recipe apps: convenient within the app's library but miss recipes from outside their ecosystem
  • Manual transcription with a shared notes app: works but is the exact friction point the approach is trying to remove

A better system for family planning

The principle that makes AI-generated shopping lists genuinely useful is source-agnostic import: any recipe from any website, any photo of a cookbook page, any hand-typed dish name becomes a structured, scaled ingredient list without manual transcription. The planning workflow shifts from 'find recipe, copy to list, scale manually' to 'find recipe, add to plan' — one step instead of three. The ingredient extraction and scaling happen in the background.

The second level of value is consolidation across the whole week's plan. If Monday calls for chicken thighs and Wednesday uses chicken breast, that's fine — they're different items. But if both Monday and Thursday use the same tin of chopped tomatoes, the AI combines them into one line. And if you already have olive oil in the cupboard, it doesn't appear on the list. The output is a single, accurate, deduplicated list for the entire week's cooking.

  • Import any recipe from any URL or photo without typing a single ingredient manually
  • Automatic portion scaling to your household size handles the maths for every recipe
  • Cross-recipe deduplication across the full week's plan produces one consolidated list

Example of a weekly system

Sunday meal planning: add the week's five to seven recipes — paste the URL, take a photo of the cookbook page, or pick from your saved library. The AI generates a single consolidated shopping list for all planned dinners, with quantities combined across recipes and pantry items deducted. You review the list, adjust anything the AI got wrong (an unusual measurement, a pantry item you know you're out of), and it's ready to use. The planning session is about choosing what to cook, not about transcribing ingredients.

Mid-week, when you find a recipe you want to add for Thursday: paste it into the plan. The new ingredients add to the existing list without duplicating anything already there. When Thursday changes to something simpler because the week has been exhausting: remove the recipe, and its ingredients drop off the list automatically. The shopping list always reflects exactly what's planned — not a frozen snapshot from last Sunday.

  • Sunday: add all recipes via URL, photo, or library — AI builds the consolidated week's list
  • Review the list for accuracy before the main shop — adjust quantities or pantry misses
  • Mid-week additions go straight in without manual deduplication
  • Recipe removals update the list immediately — no stale ingredients to cross off at the shop

How Zenframe helps

Zenframe Assistant handles recipe import: paste a URL from a food blog, BBC Good Food, or any structured recipe site, and the assistant extracts the ingredients, scales them to your registered household size, and drops them into the week's shopping list. No manual entry. The recipe saves to your family library so next time you want to cook it, it's one tap to add rather than hunting for the original source. Photo import works for cookbook pages or handwritten recipes that aren't online.

The shopping list in Zenframe Meals is connected to the rest of the plan — it updates when you swap a dinner in Planner, deducts pantry items you've registered, and syncs in real time across all household members. The AI assistant can also handle proportional scaling: tell it you're cooking for six instead of four and it recalculates every ingredient in the recipe. That removes the fractional arithmetic that makes manual scaling unreliable.

  • Zenframe Assistant: paste any recipe URL and the AI extracts, scales, and adds ingredients to the shopping list instantly
  • Photo import handles cookbook pages and handwritten recipes that aren't online
  • Pantry integration deducts ingredients you already have, so the list only shows what you need to buy

Practical tips families can start with today

  • Paste the recipe URL rather than typing ingredients — it takes two seconds and eliminates transcription errors.
  • Set your household size once in Zenframe; every recipe import scales to the right number of portions automatically.
  • Save frequently cooked recipes to the family library — Friday's pasta doesn't need a new import every week.
  • Add recipes during the week as you find them; the list consolidates automatically without creating duplicates.
  • Do a quick pantry check once a week to keep the deduction logic accurate and avoid being told you have ingredients you've already used.

FAQ

Does AI recipe import work with any website?

It works with the majority of recipe websites that use standard recipe markup — BBC Good Food, Delicious Magazine, The Guardian Food, Nigella.com, and most food blogs. Some sites block automated reading, and a small number use non-standard layouts that produce imperfect extractions. For those cases, photographing the recipe page or pasting the ingredient list manually are reliable fallbacks. The import quality is directly proportional to how clearly the recipe is structured on the source page.

What happens if the AI gets a quantity wrong?

You can edit any ingredient quantity in the list before shopping. The AI's extraction is accurate for clearly written recipes but can misread ambiguous measurements (e.g., '1 large onion' vs '1 cup of onion') or unusual units. The review step — a one-minute check of the generated list before the shopping run — is where you catch and fix those edge cases. Over time, corrections teach the system your household's typical interpretation of ambiguous measurements.

We cook from cookbooks, not websites. Does photo import work reliably?

Photo import works well for clearly printed cookbook pages with good lighting and a straight-on angle. It handles standard recipe formats from most cookbooks published in the last thirty years. Handwritten recipes are more variable — printed handwriting reads better than cursive. For your most frequently cooked family recipes, spending a few minutes creating them in the recipe library once (with the photo import as a starting point to edit from) is more reliable than re-scanning each time.

How does AI recipe import connect to the rest of the Zenframe family plan?

Each recipe you import links to a specific dinner slot in the weekly plan in Zenframe Meals. When you move Thursday's dinner to Friday in the Planner, the ingredients stay attached to the recipe and move with it — the list doesn't need manual updating. If you're cooking for guests on Saturday and double the recipe, the shopping list doubles those quantities while keeping all other dinners at standard household size. The list is a live output of the plan, not a separate document.