Best food scanner apps 2026: EatMatrix vs Yuka, on the evidence

For a shopper standing in a grocery aisle in 2026, two food scanner apps are worth comparing: EatMatrix and Yuka. Every other name in this category is either a regional variant of one of these two, or a stripped-down barcode lookup with no scoring layer at all.

Neither app wins on every dimension. EatMatrix wins when the reason behind a flag matters to you. Yuka wins when you want one fast number and don’t plan to look further.

Both apps score products by building on the NOVA food processing classification system, which groups foods by the extent and purpose of industrial processing they have undergone rather than by nutrient content alone.^[1] That system splits every food into four categories based on how far it has been altered from its original form.^[1] Where the two apps diverge is what they do with that starting point, and that difference is the whole story of this comparison.

What each app scores

EatMatrix scores a flagged additive against its specific regulatory record. If an ingredient shows up in a scan, the app links that flag to the actual EFSA, FDA, or JECFA assessment behind it, rather than assigning the same penalty to every additive regardless of what regulators concluded about it.

Yuka combines a Nutri-Score-derived nutritional grade with a fixed penalty for any additive it classifies as a risk, applied the same way whether the amount present in a product is trivial or substantial.

Both approaches start from the same NOVA logic: more industrial processing, more scrutiny. The split happens after that, in how each app treats what the scrutiny finds.

The core difference: EatMatrix’s flag tells you what the regulator said. Yuka’s flag tells you that an additive is on a list. Those are not the same piece of information, and the gap between them is where a shopper can get misled.

Where the scores agree with the regulators: titanium dioxide

Titanium dioxide (E171) is the clearest case where both apps’ skepticism lines up with what regulators found. EFSA concluded in 2021 that E171 could no longer be considered safe as a food additive, after a persistent concern for genotoxicity could not be ruled out.^[2]

The detail that matters is which part of the safety picture triggered that conclusion. EFSA’s panel reviewed an extended one-generation reproductive toxicity study in animals and found no adverse effects on general or organ toxicity at doses up to 1,000 mg per kg of body weight per day, the highest dose tested.^[2] The concern that ended E171’s approval was narrower and more specific than a blanket “this is dangerous” verdict: the panel could not rule out that titanium dioxide particles cause DNA strand breaks and chromosomal damage, and no safe threshold for that specific effect could be identified from the data available.^[2]

This is a case where flagging the additive hard is the right call, and both apps do it. Neither app oversells the finding into “titanium dioxide causes cancer”, which the EFSA opinion does not establish. The genotoxicity concern is real, specific, and sourced, which is exactly the kind of flag a scoring app should carry loudly.

Where the scores get ahead of the regulators: carrageenan

Carrageenan is a different story, and it’s where a uniform scoring system starts to mislead rather than inform.

EFSA’s 2018 re-evaluation of carrageenan found no adverse effects and no genotoxicity concern in the animal toxicity studies it reviewed. No adverse effects were detected in rats even at 7,500 mg per kg of body weight per day, the highest dose tested in the chronic toxicity studies.^[3]

But EFSA did not clear carrageenan outright. The panel kept the additive’s group acceptable daily intake temporary, at 75 mg per kg of body weight per day, and asked for better exposure and toxicological data within five years because of remaining uncertainties in the underlying database, not because of a safety signal in the toxicology.^[3]

7,500 mg/kg bw/day highest carrageenan dose tested in EFSA's chronic toxicity review, with no adverse effects observed PMID 32625873

Compare the two additives directly:

Two additives, two different EFSA verdicts

Both show up as flags in a food scanner app. The regulatory record behind them is not the same.

One additive has a specific, unresolved genotoxicity concern that ended its approval. The other has no adverse-effect finding at all, and a provisional status driven by a documentation gap. A scoring system that penalizes both the same way because both sit on a “watch” list can’t tell a reader which situation they’re looking at.

It’s a bit like a credit score that dings you the same amount whether you missed one payment by a day or defaulted for six months. The number moves in the same direction either way. It just stops telling you how bad the miss was, and a shopper who only sees the grade has no way to ask the follow-up question.

Where both apps can mislead you

Every scoring system has an honest downside, and this one has more than one.

• The uniform-penalty problem. A processing-based flag that treats every additive the same can’t distinguish an additive with a genuinely withdrawn safety margin, like titanium dioxide, from one regulators have repeatedly reviewed and left in provisional use pending more data, like carrageenan.
• The dose blind spot. A score built around population-wide regulatory thresholds can undersell a real individual concern, such as a sensitivity that shows up in some people below the dose regulators consider safe for the general population. A regulatory clearance describes the average person at typical exposure, and an individual sensitivity can sit outside that average entirely.
• The database lag. Neither app’s barcode scan substitutes for reading the actual ingredient list on the package in your hand. Manufacturers reformulate faster than any product database updates, and a scan result reflects the last time someone logged that product, not necessarily what’s in the box on the shelf today.

We looked for peer-reviewed evidence that would let us rank the two apps’ underlying classification methods directly against each other, for example on how consistently trained reviewers agree which processing category a given product belongs in, or how well either app’s output correlates with actual health outcomes. We could not find a study that answers that specific question well enough to cite here. That is a real gap in what’s publicly known about how these classification systems perform in practice, and we would rather say so plainly than paper over it with a confident-sounding guess dressed up as evidence.

That gap cuts against both apps equally. Neither EatMatrix nor Yuka has published independent, peer-reviewed validation of its scoring output against measured health outcomes. What EatMatrix can offer instead is transparency about which regulatory opinion sits behind each individual flag, so a reader can go check the primary source themselves rather than trusting a black-box grade. That is a difference in how each app handles uncertainty, not a claim that either app’s overall scoring has been clinically validated.

Practical considerations

EatMatrix vs Yuka, side by side

What differs for a shopper deciding between them.

Track record. Yuka has a longer history in the European market, where NOVA-style and Nutri-Score-based labeling is already culturally familiar to shoppers, which shows up in how quickly people trust its single-letter grade without digging further.

Depth of explanation. EatMatrix’s additive pages let you look up the specific regulatory opinion behind a flag directly from the scan result, instead of stopping at a single letter grade and leaving the reasoning implicit.

If you want the raw data instead. Open-database alternatives built on Open Food Facts exist for shoppers who want to see the underlying ingredient and additive data without either app’s proprietary scoring layer sitting on top of it.

Choose EatMatrix if, choose Yuka if

Choose EatMatrix if the reason a product got flagged matters as much as the flag itself, and you want the regulatory source sitting right behind the score rather than a grade you have to take on faith.

Choose Yuka if what you want is a fast single-number grade across a wide, well-established product database, and you’re comfortable not seeing the regulatory reasoning underneath it every time you scan something.

Bottom line: Both apps are a reasonable starting point for reading a label faster than you could on your own. Neither is a substitute for knowing what a specific flagged additive’s actual dose and regulatory status are, and right now, EatMatrix is the one that shows you that status instead of stopping at a grade.

For a full breakdown of any specific additive mentioned here, including the source regulatory opinions behind it, EatMatrix’s additive pages carry the underlying record behind every flag.

This article summarises publicly available regulatory assessments and peer-reviewed research about food ingredients. It is general information and is not intended to diagnose, treat, cure, or prevent any disease, and it is not personalised dietary or medical advice. Regulatory positions change; the EatMatrix app reflects the current assessments in our database. If you have an allergy, a medical condition, or are pregnant, talk to a qualified clinician or dietitian before changing what you eat.