Yuka vs Open Food Facts: private score vs open database
Yuka is a private company that returns a 0–100 score computed from a database it holds. Open Food Facts is a non-profit whose entire product database is downloadable under the Open Database License 1.0. That single difference drives coverage, freshness and repair. Open Food Facts lists 4.6 million products on its own counter.
This page compares the two on ownership, licence, coverage, freshness and who can fix a wrong record, then says which shopper each one fits. It is not a ranking of two competitors doing the same job. The rest of the field sits in label scanner apps compared.
What Yuka and Open Food Facts actually are
Yuka is a private company publishing a score computed from a database it holds. Open Food Facts is a non-profit whose entire product database is downloadable by anyone under the Open Database License 1.0, with a sibling project, Open Beauty Facts, for cosmetics. One publishes a verdict; the other publishes the raw material.
That is a structural difference rather than a feature difference, and it propagates into everything downstream. Ownership decides who is accountable for a wrong record, whether an outsider can audit the data, and whether the answer you get today can be reproduced by someone else in six months.
4.6 million
That counter is published by the project itself and moves as contributors add products. Yuka reports 28 million users in the United States and 85 million worldwide on its press page (checked 10 August 2026). The two figures measure different things — one counts records in a database, the other counts people holding a phone — and neither is a coverage guarantee in your own store.
Yuka vs Open Food Facts, point by point
Yuka scans a barcode and returns a score with a colour band. Open Food Facts scans a barcode and returns Nutri-Score, NOVA group and the underlying fields. The table below records what each project publishes about itself, checked on 10 August 2026, with licence and repair as rows rather than footnotes.
| Criterion | Yuka | Open Food Facts |
|---|---|---|
| Who runs it | Private company | Non-profit project |
| Data licence | Not published for reuse | Whole database under ODbL 1.0 |
| Stated database size | No public count in the pages checked | 4.6 million products on its homepage counter |
| Where records come from | Collected and maintained by the company | Contributed by users, text and label photographs |
| If a record is wrong | Not described on Yuka’s public pages | Edit the record yourself and save |
| Main output | 0–100 score with a colour band | Nutri-Score, NOVA group and the raw fields |
| Cosmetics | In the same app as food | In the sibling project Open Beauty Facts |
| Price | Free to scan; Premium from $10 a year in the US | Free, no paid tier |
| Advertising | None in either tier | None |
| Reuse by other tools | Not offered | Full export, used by third-party apps |
Two rows carry most of the argument: the licence and the repair path. Everything else in the table follows from them. A wider field, with a barcode scanner set against a curated one and against a text checker, is laid out in four Yuka alternatives side by side, and the dietary-restriction case — where the question is Low FODMAP or alpha-gal rather than a general score — is handled in Yuka compared with Fig, whose web search covers 363,250+ products without a login.
Who can fix a wrong product record
Open Food Facts is the only one of the two where a shopper can correct a wrong product record directly. Editing is the normal way data enters the project: open the product page, change the field, add a photograph of the package, save. Yuka accepts reports through its app and decides internally what happens next.
Step 1
Open the product page
Search the barcode on the website or in the app. If no page exists, the product is simply missing, and creating it is the same operation as fixing it.
Step 2
Change the field that is wrong
Ingredient text, quantity, categories and labels are editable fields rather than a support ticket. The edit is attributed and visible to everyone who looks at the product afterwards.
Step 3
Photograph the package as evidence
A photo of the ingredient panel lets the next contributor check the text against the package instead of trusting the typing. This is what makes the correction reviewable rather than just asserted.
The repair path matters most exactly where product databases are weakest: store brands, regional lines and packaging reformulated last quarter. On Open Food Facts one shopper spends five minutes and the correction lands in the next public export under ODbL 1.0, where every other tool built on that data inherits it.
Where crowd-sourced data thins out
Crowd-sourced coverage is uneven by construction. Open Food Facts grows where its contributors shop, so coverage tracks volunteer activity rather than the shelves in front of any particular person, and a record is only as current as the last package someone photographed. These are real costs of the open model.
Coverage follows contributors
A database filled by volunteers grows fastest where volunteers are active. A regional US store brand can have no record at all until one shopper decides to add it.
A record can be older than the package
Reformulation happens on the shelf before it happens in any database. An entry reflects the day it was last photographed, which is why the photo matters as much as the typed text.
Fields can be partly filled
A product may carry a name and a barcode but no ingredient text yet. The lookup then succeeds and still leaves the question unanswered, which is a different failure from a missing product.
This is also why the alternative question splits in two. Search results for people leaving Open Food Facts divide between shoppers who want a finished app and developers who want the API and the dump, which are almost opposite requirements; Open Food Facts alternatives for shoppers separates the two before recommending anything.
What a private database does not let you check
A private database cannot be audited from outside. Yuka publishes a description of how its score is built, and a description is not the same object as the data: nobody outside the company can recompute a score, count how many records carry a given additive, or see when a particular entry was last touched.
There is a fair trade on the other side of that. Yuka runs no advertising in either tier, so the money comes from the people using it rather than from the brands being rated, and a single owner produces consistent records without waiting for a volunteer. Consistency and independence from advertisers are genuine properties of the closed model, not concessions.
Both models produce wrong records. Only one of them lets the person who spotted the error open the record, change it, and have everyone else inherit the fix.
Which one fits which shopper
Use Yuka if the job is ranking two packaged products in an aisle in seconds and the Premium features — offline use, search without scanning, dietary filters, from $10 a year in the US — match how you actually shop. Use Open Food Facts if you want the underlying fields, the licence, or the ability to repair what is wrong.
The conditions matter more than the preference. If your basket is mostly national brands and you never intend to look under the score, the open licence buys you nothing you will use. If you are building anything on top of the data, or you care that a stranger can check the same record you did, only one of the two is even a candidate. Running both on one shopping trip is a cheap way to find out which describes your stores.
When neither has a record for what you are holding
Both apps answer from a product record, so both go quiet when no record exists — bulk bins, deli counters, imported packaging, a store brand nobody has added yet. A text checker sidesteps the record entirely, because the input is the ingredient list printed on the package rather than a barcode.
That is what LabelPeek does, and it is a narrower job than either app on this page. Two free checkers run in the browser with no account and no upload: the ultra-processed food checker matches pasted label text against 74 markers of industrial processing, and the pore-clogging ingredients checker matches pasted INCI text against a published list of 86 entries. Both print the full list on the same page as the tool.
The output has three states, not two: matched, recognised but not on the list, and not recognised at all. That third state is the point. A checker that prints a clean verdict while silently counting names it could not parse is reporting a gap in its data as a statement about the product.
Where LabelPeek is the wrong choice
LabelPeek is the wrong choice for in-store scanning. There is no barcode database behind the web checkers, nothing can be pointed at a shelf, and the LabelPeek app was not listed in any store as of 10 August 2026. For scanning today, a released scanner is the right tool and this page names two of them.
It is also the wrong choice for nutrition. LabelPeek reads ingredient lists, not the nutrition panel, so sugar per serving, sodium and protein sit outside what it answers — which is precisely what Nutri-Score and NOVA group on Open Food Facts exist for. And it holds no product records, so it cannot build a shopping list, compare two brands of yoghurt, or tell you anything about a package you are not currently reading.
On skincare it will disagree with other tools, deliberately. No regulator anywhere certifies a comedogenic list, published lists contradict each other on the same ingredient, and the honest response is to print the list being used rather than to imply an authority that does not exist.
None of this is medical advice, and none of it is a verdict on any named brand’s product. The page compares how two projects are built and where each one stops; whether a particular product suits you depends on your own circumstances, and on a professional when the stakes are health-related.