AI Food Label Scanner

Photo the nutrition label — get the full breakdown plus hidden-sugar, sodium, and trans-fat flags in 5 seconds. No typing, no database.

What the Scanner Extracts

  • Macro panel: calories, protein, total carbs, sugar, added sugar, fiber, total fat, saturated fat, trans fat, sodium.
  • Serving size + servings per container — so you know what the "per serving" numbers actually mean.
  • Ingredient list parsed and categorized.
  • Added sugars in disguise — 60+ name variants.
  • Industrial fats — hydrogenated and partially hydrogenated oils.
  • High-sodium warnings — anything above 20% DV per serving.
  • Allergens — if you\'ve set them in Nouri (gluten, dairy, nuts, shellfish, etc.).
  • Artificial additives — colors, flavors, preservatives.

How to Scan a Label

STEP 1

📸 Photo the label

Include both the nutrition facts panel and the ingredients list. Keep the label flat, well-lit, in focus.

STEP 2

💬 Send to Nouri

Telegram message. Optionally add "for one serving" or "I\'ll eat the whole package."

STEP 3

⚠️ Get breakdown + flags

Macros, total for your intended portion, concerns highlighted, and an optional log into your daily total.

When Label Scanning Matters Most

🛒 Grocery-store decisions

You\'re picking between two yogurts. Scan both labels. See which has less added sugar and more protein in 30 seconds.

🍬 Hidden sugar hunting

"No added sugar" on the front, three sugars in the first five ingredients. Scanner catches the mismatch.

🧂 Sodium-sensitive diets

2,300mg/day target. Manually adding up sodium is tedious. Scanner flags anything >460mg/serving automatically.

🌾 Allergen-avoidant shopping

Set allergens in your profile. Scanner alerts when a scanned product contains them — including derivative names.

🏷 International travel

Labels in a language you don\'t read. Scanner translates relevant nutrition + ingredient info into your language.

📦 Processed food audit

Doing a pantry cleanup. Scan everything, see which products actually contain the ingredients you want to avoid.

Why Reading Labels Manually Is Hard

US food manufacturers are required to disclose nutrition facts, but the design of the disclosure is strategic. Three common tricks:

  1. Tiny serving sizes. A "20 calorie" cooking spray is 20 calories per 0.25-second spritz. A realistic use is 5 seconds = 400 calories of fat.
  2. Sugar under alias names. Up to 60 ingredients qualify as added sugar. Split into 3–5 different names, none ranks in the top 3 of the ingredient list — but summed, they\'re the main thing you\'re eating.
  3. Rounding rules that hide trans fat. Under 0.5g of trans fat per serving is legally "0g." A 40-serving container with 0.4g trans fat per serving = 16g of trans fat total.

The scanner catches all three by pattern-matching across the label + ingredient list. You don\'t need to know the tricks — it already does.

Limitations

  • Fresh produce and unpackaged food have no labels. Use the photo calorie counter for those.
  • Bulk bins and store-packaged items often have minimal labels. Useful data may be missing.
  • "Clean label" marketing claims — terms like "natural," "clean," "minimally processed" — have no regulatory definition. Scanner won\'t validate them.
  • Supplements and medications use a different labeling standard (Supplement Facts panel). Scanner handles these but differently.

Learning to Read Labels (So You Don\'t Always Need the Scanner)

The scanner is a shortcut, not a replacement for nutritional literacy. If you want the underlying education:

Once you can read a label in 10 seconds on your own, the scanner is still useful for volume (doing a full pantry audit), for languages you don\'t read, and for catching the marketing tricks that fool even careful readers.

Frequently Asked Questions

How does the AI food label scanner work?

Send a photo of a nutrition facts panel + ingredients list. Nouri extracts the full macro breakdown (calories, protein, carbs, fat, fiber, sugar, sodium), parses the ingredients, and flags concerns — added sugars hiding under 12 different names, industrial seed oils, sodium above 20% DV per serving, artificial additives. Result in under 5 seconds.

Can it read labels in other languages?

Yes, though accuracy varies. Strong on English, Spanish, French, German, Italian, Portuguese. Good on Japanese, Chinese, Korean. Weaker on Arabic and less-common scripts. For international travel, combine label scan with the photo-based calorie counter which doesn't require label text at all.

Does it flag hidden sugar?

Yes — that's one of the main use cases. The scanner cross-references over 60 names for added sugar (dextrose, maltodextrin, cane juice, high-fructose corn syrup, agave, rice syrup, etc.) and calls them out when they appear in the ingredients list. Bonus: it compares "sugar per serving" to realistic serving sizes you'd actually eat.

What counts as a "concerning" ingredient?

Nouri flags: added sugars (any of the 60+ variants), partially hydrogenated oils (trans fat even when declared "0g"), sodium >20% DV/serving, artificial colors and flavors, high-intensity sweeteners if you've asked to avoid them, allergens if you've configured them. Flags are informational — not medical advice.

Can it tell me if a food is healthy?

"Healthy" depends on your goal, context, and overall diet. Nouri won't say "this is healthy" or "this is unhealthy" as a blanket judgment. It gives you the facts (grams of added sugar, sodium, fiber) plus pattern flags, and you decide.

Does it work with handwritten labels or restaurant menus?

Printed labels only. Handwritten text is often too variable for OCR. For restaurant menus, use the restaurant calorie checker which photographs the food, not the menu.

Can I compare two products?

Yes. Send two label photos in sequence and Nouri produces a side-by-side: calories, protein, sugar, fiber, sodium, ingredient concerns. Useful at the grocery store when you're picking between versions of the same thing.

Related Reading

See What\'s Actually in It

Free label scanning. Hidden sugars + sodium flags included.

Start Free in Telegram