Yuka is the app most people mean when they say they scan their food. It puts a number out of 100 on a product and a colour beside it, and it does that for millions of barcodes. This page explains how that number is worked out, what it covers, and the one thing a single number cannot do, whichever app calculates it.
Everything here about Yuka comes from Yuka’s own help pages, linked as it goes. Prices and features change, so the date above is the one that matters.
How the Yuka score is calculated
Yuka publishes the recipe, which is more than most apps do. A food product’s score has three parts:
- Nutritional quality is 60% of the score. This comes from the Nutri-Score, the label that weighs sugar, salt, saturated fat and calories against protein, fibre and the fruit and vegetable content.
- The presence of additives is 30% of the score. Yuka rates additives against published research and regulator opinions. A single additive it classes as high risk caps the whole product at 49 out of 100.
- The organic dimension is 10% of the score. A bonus for products carrying an official organic label.
Those three lines are quoted from Yuka’s own explanation of how food products are rated.
What Yuka does well
Three things, and they are the reason it has the users it has.
It publishes its method. You can read the weighting above and decide whether you agree with it. An app that scores your food and will not say how is asking for more trust than it has earned.
It gives one answer. A number and a colour is a decision you can make in a supermarket aisle in two seconds. That is a real design achievement and it is why the format spread.
It names additives. Most shoppers cannot tell E322 from E621. Yuka puts a name and a risk classification next to each one, which is the part of a label that is genuinely hard to read unaided.
Where Yuka falls short
This is the honest limit, and it is a limit of the format rather than a fault in the execution. A score out of 100 is a property of the product. It is the same number for every person who scans it.
That works when the question is “is this product, in general, a nutritionally balanced one”. It does less when the question is “does this suit me”. Two readers standing in the same aisle can have opposite answers to the second question about the same product, and one number cannot hold both.
The weighting is a judgement call too, and reasonable people land elsewhere. Nutrition at 60% and additives at 30% is Yuka’s view of what matters. If you are avoiding one specific ingredient, neither of those percentages is about you.
Is Yuka accurate?
Two different questions hide inside that one, and they have different answers.
The first is whether the label data is right. Yuka reads a product record, and a record can be wrong or out of date whoever is reading it. That applies to Kaibu exactly as much as it applies to Yuka: we both work from crowd-sourced label data, and neither of us can be right about a product whose record is wrong. This is why the pack in your hand is the last word and an app never is.
The second is whether the scoring is right, and that one is not really about accuracy at all. The arithmetic is applied the same way to every product: 60% nutrition, 30% additives, 10% organic. What people usually mean when they call a score wrong is that they disagree with those proportions, or with Yuka classing a particular additive as high risk. That is a judgement about what matters, not a calculation error, and Yuka publishes it so you can take issue with it.
The place the two get confused most often is the cap. A single additive Yuka rates as high risk holds the whole product at 49 out of 100 however unremarkable the rest of the label is. People meet that, see a product they consider ordinary marked poorly, and read it as a fault. It is deliberate, and whether it is right depends on whether you share Yuka’s view of that additive.
What the score cannot be is accurate about you, because it was never about you. It is one number per product, the same for every person who scans it, which is the limit above rather than a flaw in the execution.
Is Yuka trustworthy?
The scepticism people bring to Yuka is mostly about where it draws its lines, not about whether it is honest. Three things are worth weighing.
It publishes the method. The weighting above is on Yuka’s own help site. You can disagree with it, which is only possible because it is written down.
It says it takes no money from brands. Yuka’s independence page states that “Revenues come from users, never from brands” and that “Brands cannot pay Yuka to advertise their products”. Its income is the Premium subscription. That is Yuka’s own account of itself, not an audit, but it is a public commitment a company can be held to.
It has defended its warnings in court and won.
What the Yuka controversy is about
The disputes are a matter of public record, and they went the other way from what “controversy” usually implies.
By Yuka’s own account, between January and June 2021 it faced three lawsuits from the French charcuterie federation and two associated companies over the warnings it showed on nitrites in cured meat, with more than €1.4 million claimed against it. Yuka lost at first instance and appealed.
On 7 June 2023 the Paris Court of Appeal found for Yuka and ordered the federation to pay €60,000 in damages, as reported by FoodTimes. The court held that informing consumers on the question took precedence over the industry bodies’ commercial interests.
So the short answer to somebody arriving here worried: the controversy is that an industry group sued the app over what it told shoppers, and the app won.
What people ask about Yuka
The same handful of questions come up wherever the app is discussed. Summarised rather than quoted, because a forum thread is one person’s experience and not evidence:
- “Why did this product score so low?” Usually one additive Yuka classes as high risk, which caps the whole product at 49 out of 100 whatever else is in it.
- “Why do two similar products score differently?” Usually the organic bonus, or one additive on one label and not the other.
- “Is the additive rating stricter than the regulator?” In places, yes, and deliberately so. That is the real disagreement, and it is a judgement about risk rather than a factual error.
- “Why does it not know about my allergy?” Because the score is not about you. The Premium alerts cover seven preferences; anything outside that list is not part of the answer.
The last one is the reason this site exists, and it is the honest limit of every app that produces one number per product.
Is Yuka free?
The scanning is free. Yuka’s paid tier adds three things, in its own words: an advanced search bar, so you can look a product up without scanning it; an offline mode; and alerts, which “allow you to personalize the app according to your dietary preferences”. The alert list covers vegetarian and vegan, palm oil-free, gluten-free, lactose-free, pork-free, soy-free and sulfite-free.
The annual Premium subscription starts at $10 in the US, £10 in the UK and €10 in Europe, from Yuka’s page on what membership costs, with the feature list on its page about the paid version.
This is often reported wrongly in both directions: Yuka does personalise, but the personalisation is a set of alerts over seven preferences that Yuka chose. What does not change per reader is the score itself.
Does Yuka work for cosmetics?
Yes, and it is a separate scoring system from the food one: Yuka rates cosmetics and personal care products as well as food. If that is what you are scanning, Yuka covers it and Kaibu does not.
Kaibu is food only. That is a deliberate scope rather than a gap to fill later: the data it reads, the grades it shows and the allergen list it matches against are all about what you eat.
Where Kaibu differs
Kaibu starts from the second question rather than the first. You save the diets you follow and the allergens you avoid, and a product is read against that list. The grade on the screen is the Nutri-Score from the record, shown as the A to E it already is. The part written for you is the verdict underneath it.
Three differences follow from that:
- The allergen answer is a data lookup, not a score. Kaibu shows what the record lists, what it flags as a trace warning, and, separately, what it has nothing on file for. An ingredient with no data is shown as unknown rather than folded into a number.
- Your list is your own. It is not a fixed set of preferences. It is whatever you saved.
- The ingredient list is explained. Tap any ingredient and Kaibu says what it is and what it does in food, in plain English, free for everyone. Premium goes one step further and reads that ingredient against your own diet and allergen list.
Both apps read the same kind of public data. Kaibu’s product records come from Open Food Facts, the open database volunteers build, and where that record has no value for something, Kaibu says so rather than showing the absence as a zero.
Which one suits you
If you want one general read on a product’s nutritional balance and its additives, and you are happy with somebody else’s weighting of the two, Yuka does that well and the free tier covers it.
If you are shopping around a specific list, an allergen, an intolerance, a diet you follow, then what you need is an answer about that list rather than a score, and that is what Kaibu is for. There is a page on scanning against your own list that explains exactly what it can and cannot tell you.
Neither app is a medical device, and neither replaces the packet. Label data is crowd-sourced and can be wrong or out of date, so check the pack before you eat.