General information, not medical or dietary advice. If you are managing a condition, your care team’s limits outrank any number on any website. Full disclaimer.
A nutrition panel under a recipe looks like a fact. Usually it is an estimate, often it is a guess, and increasingly it is a sentence a language model produced because the format demanded a number. None of that is visible from looking at it — a wrong figure is typeset exactly like a right one.
You do not need a database to catch most bad numbers. Four of the five checks below are arithmetic and common sense, and they take under a minute. The fifth is the one that separates a recipe you can plan a condition around from one you cannot.
Calories are not an independent number — they are what the protein, carbohydrate and fat add up to. Protein and carbohydrate carry about 4 calories per gram, fat about 9. Multiply them out and the total should land within roughly ten per cent of the stated calories.
When it does not, the panel was not calculated from the ingredients — it was invented. This single check catches the majority of fabricated nutrition panels, because a model generating plausible-looking numbers rarely makes them consistent with each other.
Plate Planner runs this check on every dish it plans, and repairs the panel when it fails rather than displaying it. Fibre is checked against carbohydrate too: fibre is a carbohydrate, so a dish claiming more fibre than carbs is describing something impossible.
Per-serving numbers mean nothing until you know how big a serving is, and this is the most common quiet distortion in recipe nutrition. The same lasagne is a triumph at eight servings and a disaster at four. If a recipe states a per-serving figure but never says how many servings it makes, the number is decoration.
Check that the ingredient amounts and the serving count are consistent with each other: two chicken breasts do not feed six, whatever the header says.
In Plate Planner every recipe states its servings, and the ingredient amounts are sized for the number of people you actually set. Rescaling a recipe rewrites the amounts, and the per-serving nutrition stays per-serving.
Sodium is the number most often missing and most often understated, which matters because it is the number most people are told to watch. A recipe with soy sauce, stock, cheese, cured meat or anything tinned and a sodium figure in the low hundreds is not being honest with you.
Rough anchors worth carrying: a teaspoon of table salt is about 2,300 mg of sodium — an entire day’s general allowance. A tablespoon of soy sauce is roughly 900 mg. A cup of shop-bought stock is often 800 mg or more.
Plate Planner shows sodium on every dish and totals it per day, and it will hold a whole week under a ceiling you set — 1,500 mg a day is a preset, because that is the figure most often given to people managing blood pressure or kidney disease.
Calories, protein, carbs and fat are the easy four, and for many people they are not the ones that matter. Kidney disease turns on potassium and phosphorus. Heart guidance turns on saturated fat and sodium. Someone on a GLP-1 medication cares most about protein and fibre holding up while total intake falls.
A site that only ever shows the easy four is not withholding the others — it usually does not have them. Absence is information: it tells you the source is a generic calculator, not a food database.
Plate Planner shows potassium and phosphorus alongside sodium when a renal need is switched on, and highlights whichever numbers the people actually eating that dish need to watch, rather than showing everyone everything.
This is the check that decides whether the rest can be trusted. Pick any single figure and ask where it came from. Can the site show you the ingredient behind it — this much chicken, this much rice, matched to an actual entry in a food dataset? Or is the number simply asserted?
If nothing explains it, treat the entire panel as a sketch. That is fine for a weeknight. It is not fine for planning around a lab result.
Plate Planner matches each ingredient line against a bundled USDA-derived food dataset and will show you that breakdown, including the lines it could not match — which it says plainly rather than quietly estimating over. A dish only carries a verified mark when the match is genuine, so an unmarked dish is telling you something real about itself.
Do not throw it away — cook it, and stop treating its panel as data. A recipe can be a good recipe and a bad nutrition source at the same time, and most of the internet’s best cooking is exactly that.
The distinction that matters is what you are using the number for. Curiosity survives a bad estimate. A potassium limit does not.
Not inherently, and confidently wrong is its normal failure mode. A language model produces a number because the format calls for one; nothing in that process consults a food database. It becomes trustworthy only when each ingredient is matched against real food data afterwards and the arithmetic is checked — and when the app admits the lines it could not match.
Multiply protein and carbohydrate by 4 and fat by 9, add them up, and compare with the stated calories. Within about ten per cent means it was probably calculated; wildly off means it was not. Then check that fibre is less than carbohydrate, and that the serving count matches the ingredient amounts.
Because it is harder to estimate than the macros — it depends on brands, stock, sauces and how much salt is actually added — and because it makes recipes look worse. Missing sodium usually means nobody calculated it, not that the dish is low in it.
That the dish’s ingredient lines were matched to entries in a bundled USDA-derived dataset rather than estimated. It is not a coverage badge — unmatched lines are reported instead of hidden, and a dish that cannot be verified is not shown as verified anywhere on the site.