
Tell us who's eating and what you can spend - we pick the dinners, merge the shopping list, and price it at your Kroger with today's prices.
You set the constraints: how many people you are feeding, how many dinners you want, a budget if you have one, any diets to respect, and what is already in your cupboard. We search a library of 748 dinners that have each been priced ingredient by ingredient at a real Kroger, and pick the combination that feeds you for the least money - favoring menus that share ingredients, because a package bought for Monday should not be bought again for Thursday. The winning week is then re-priced against today's shelf prices before you see it, so the headline number is current rather than cached. You get one combined shopping list, a cost per serving for every meal, and a one-tap Kroger cart.
A meal planner that does not know prices is a calendar. This one starts from what things actually cost at your store, so the plan and the budget are the same object.
The honesty rules matter more than the optimization. If your budget cannot be met, we say so instead of shaving a meal or pretending a smaller portion counts. Uncertain ingredient matches are flagged rather than substituted behind your back. Pantry staples come out of the total instead of inflating it. Package remainder is treated as an asset, not waste: the half-package left after Tuesday is carried into next week's plan so you are not charged for it twice. And in-store pickup pricing is never mixed with delivery pricing, which carries its own markup.
Across the 748 dinners we have priced, the median costs $4.62 per serving, and the cheap tail goes a long way below that - Simmered Beans comes to $0.41 a serving. In one real two-week test the first week's shopping came to $34.60; the second week, built to spend down what the first week left in the cupboard, came to $13.91 for the same number of dinners. The remainder is where the saving lives.
Kroger stores. You give a ZIP code, we price the week at the Kroger nearest it, using that location's own shelf prices rather than a national average. The same menu costs different amounts at two stores in one metro, so the store is always named on the result.
Every dinner in the pool was already priced item by item against real products, and the winning week is re-priced live before you see the number. About 93% of ingredient matches are ones a shopper accepts unchanged; the rest are flagged for you to check rather than silently swapped.
You are told so plainly. The optimizer never quietly drops a meal or shrinks a portion to hit a number - if five dinners cannot be built under your budget at your store, it says the budget is infeasible and shows what the cheapest real week actually costs.
Yes. Stores sell packages and recipes call for amounts, so most of a basket's dollars are food the recipe does not finish. Whatever is left in usable quantities is carried forward, and the next week's build starts with those items already in your pantry.