Scoring
Last updated: July 22, 2026
Every product you scan gets a score from 0 to 100, and higher is better. But that number isn’t a universal verdict on the product — it’s a score for you. Two people can scan the same jar and get different numbers, and each of them can see exactly why.
This page explains what goes into the score, how your own profile changes it, and the rules we hold ourselves to.
The short version
- A score from 0 to 100. Higher is better.
- Built from four things: nutrition, additives, processing and contaminants.
- Your health profile and preferences change how much each of those counts.
- Every score opens into a breakdown showing which grams and which of your settings produced it.
- The same product, the same profile and the same version of Nime always give the same score.
What goes into the score
| Component | Weight | What it covers |
|---|---|---|
| Nutrition | 45% | Sugar, salt, saturated fat, fibre, protein |
| Additives | 25% | Every E-number, graded by risk rather than counted |
| Processing | 15% | How heavily processed, adjusted by category |
| Contaminants | 15% | PFAS, microplastics, heavy metals, pesticide residue |
Nutrition — 45%
The biggest share, because it’s the part with the strongest evidence behind it. We use the 2024 Nutri-Score algorithm, which is noticeably stricter than the 2017 version it replaced: sugar can now cost up to 15 penalty points instead of 10, salt up to 20 instead of 10, fats are judged on their saturated-fat ratio, and drinks — including milk and plant drinks — are held to a separate, tougher standard.
We keep the underlying grams, not just the A–E letter. That matters: if you tell Nime you’re managing your blood sugar, we need the actual grams of sugar in order to weight them for you. A single letter grade throws that information away.
Additives — 25%
Most apps count additives. We grade them. Every E-number sits in one of four tiers, based on EFSA re-evaluations and ANSES advisories:
| Tier | What it means | Examples |
|---|---|---|
| Benign | Nature-identical, or no safety concern | E300 vitamin C, E330 citric acid, E440 pectin |
| Low | Minor or no documented concern | Most lecithins, most natural colours |
| Moderate | Limited or mixed evidence of concern | E320/E321 (BHA/BHT), E407 carrageenan |
| At-risk | Flagged in EFSA or ANSES safety reviews | E249–E252 nitrites, E102/E110/E129 azo dyes, E171 titanium dioxide (EU-banned since 2022) |
One at-risk additive costs more than three benign ones. A long list of harmless ingredients won’t drag a product down, and there’s a ceiling on this component so nothing is punished endlessly just for having a lot of entries on the label.
Processing — 15%
We look at how heavily processed a product is, using the NOVA classification — but we deliberately keep this weight modest, and we adjust it by category.
Here’s why. The evidence that ultra-processed food is harmful is strongest for processed meat and sugary drinks. It is much weaker for foods like wholegrain bread and yoghurt, and for some outcomes it points the other way entirely. A flat “ultra-processed equals bad” penalty would hit a fortified wholegrain cereal as hard as a hot dog. We don’t think that’s honest, so we don’t do it.
Contaminants — 15%
Where we have data on PFAS, microplastics, heavy metals or pesticide residue, it counts here — measured against regulatory limits from bodies such as EFSA, the EPA and the FDA. We name the standard we used. We don’t take a campaign group’s threshold and present it as if it were law.
When we have no contaminant data for a product, we don’t invent a penalty. That 15% is redistributed across the other three components, and the product’s confidence level drops so you can see the score rests on less. Unknown isn’t the same as safe — but it isn’t the same as bad, either.
What the score bands mean
| Score | Band | What it means |
|---|---|---|
| 90–100 | Excellent | Among the best in its category |
| 70–89 | Good | A solid choice, with minor trade-offs |
| 50–69 | Fair | Some real drawbacks worth knowing about |
| 40–49 | Poor | Significant issues, given your profile |
| 0–39 | Avoid | Best left on the shelf |
The bands themselves are fixed — it’s your score that moves between them. That’s why the same product can land in Fair for one person and Poor for another.
Your score is yours
This is what makes Nime’s number different from a shelf-wide rating. The conditions, allergens, diet and goals you set during onboarding change the maths in two ways.
1. They change how much things count
| If you’ve told us about | What changes |
|---|---|
| Diabetes or prediabetes | Sugar counts double |
| High blood pressure | Salt counts double |
| A heart-health goal | Saturated fat and salt count 1.5× |
| Pregnancy | At-risk additives count double |
| A kidney condition | Salt and phosphate additives weigh heavier |
You also set the overall caution level yourself. The sensitivity slider — Relaxed, Balanced or Strict— scales the precautionary parts of the score, meaning additives and processing. If you find most food apps alarmist, turn it down. If you’d rather err on the side of caution, turn it up. We don’t make that choice for you.
2. Some things are absolute
Not everything should be a gentle nudge. Two kinds of things override the score outright:
- Allergens.If a product contains something on your allergen list, it’s flagged and the score is capped — no matter how good the nutrition looks.
- Diet.A non-vegan product on a vegan profile, or a non-halal one on a halal profile, gets an explicit, named flag — never a quiet deduction you can’t trace.
The same cereal, two people
Take a sweetened wholegrain breakfast cereal. Per 100g: 22g sugar, 6g fibre, 8g protein, 0.9g salt. It contains BHT (E321, moderate tier) and it’s ultra-processed.
Someone with no specific health flags, on Balanced: around 64 — “fair”. High in sugar, but the wholegrain fibre and protein pull it back up.
Someone managing diabetes, on Strict: around 41 — “poor”. The sugar counts double, which makes it the dominant factor, and the caution setting nudges the additive and processing terms up.
Same product. Same engine. Two honest answers — each with its own explanation attached.
Every score shows its work
Tap any score and it opens into its parts. For each factor you see what it was, how much it moved the number, and why — including your own settings, written in plain language: “Sugar weighted 2× because you flagged diabetes.” No black box, and no single letter you have to take on faith.
Alongside the score you’ll see a confidence level, so you know how complete the data behind it is.
The rules we hold ourselves to
- Same in, same out.The same product with the same profile on the same version of Nime always returns the same score. Scores aren’t improvised by an AI that might answer differently tomorrow.
- Every weight is written down.Nothing on this page is a magic number buried in code. Each weight carries a rationale and an evidence level — strong, moderate or limited — so you can see how firm the ground under it is.
- We benchmark against regulators. When a contaminant or an additive drives a penalty, we cite the specific standard behind it.
- We version the algorithm publicly.If you want to know why Nime treats something the way it does, the answer is a named, dated setting — and some of them you can adjust yourself.
- Missing data lowers confidence, not honesty. We’d rather tell you we don’t know than dress a guess up as a number.
The weights on this page are our current settings. They are deliberately adjustable, and we update this page whenever they change.
What Nime isn’t
Nime is a shopping companion, not medical advice and not a diagnostic tool. It doesn’t replace guidance from a doctor or a registered dietitian. The score is there to make what’s actually in a product readable in the time it takes to scan a barcode — what you do with that is your call.
Questions about a score?
If something here doesn’t add up, or you think we’ve got a weight or a classification wrong, we want to hear it — corrections with sources get read and actioned. Get in touch through our contact form.