Built by a packaging designer who kept having to rebuild the panel

Our story

Before there was a label generator, there was a packaging brief.

Waqar Anwar has spent his career in packaging design, working through the agency Proton Effect on retail programmes for the North American market. That work is the whole package: the structure, the front panel, the colour that has to survive a press check, the barcode that has to scan on the first pass, and the regulatory panel on the back. Products he designed sit on Dollar Tree shelves across the United States, and his packaging work includes programmes for WD-40 Canada and Grainger Canada.

The Nutrition Facts panel was always the part that went wrong.

Everything else on a package is a design decision. The panel is not. It is a specification. 21 CFR 101.9 fixes the type sizes, the rule weights, the indentation, which nutrients are mandatory, how every value rounds, and the point at which a number becomes "0 g" or drops off the panel altogether. Get the artwork wrong and a client asks for a revision. Get the panel wrong and a print run becomes scrap.

In practice the panel arrived as a low-resolution JPEG from somebody's nutritionist, or a screenshot from software nobody on the project had a licence for, or a Word table that had to be rebuilt by hand in Illustrator at the correct point sizes. Rebuilding it by hand meant re-typesetting a legal document. Every one had to be checked against the regulation, and every revision meant checking it again.

The tools that were supposed to solve this did not. They were priced for companies with a regulatory department, built for the browsers of a decade ago, and several of them put the basic FDA formats behind a paywall while advertising free labels. None produced a file a designer could place into artwork without redrawing it.

The realisation was that the panel is deterministic. Given the nutrition numbers and the format, exactly one panel is correct, and software can produce it more reliably than a person with a ruler and a deadline. That is not a design problem. It is an engineering problem that had been handed to designers because nobody had built the tool.

So he built the tool. First for his own briefs, then for the other designers and small brands who kept asking how the panel had been done. The US formats came first because that is where the packaging work was. The other 114 countries followed, because a client who sells in the United States usually sells somewhere else too, and the same panel has to be rebuilt again under a different rule each time.

What we do

Nutrition Fact Label turns nutrition data into a print-ready regulatory panel. You enter the numbers, or build them from a recipe and its ingredients, and the generator produces a correctly typeset Nutrition Facts panel with the Daily Values calculated and the rounding rules already applied.

It covers 152 label formats across 114 countries: the US FDA panels and Supplement Facts, Canada, the EU and UK, Mexico, Brazil, China, India, Japan, Australia and New Zealand, the Gulf states and the rest. Around the panel it handles the parts of a label that travel with it, including ingredient statements, allergen declarations, serving sizes, nutrient content claims, recipe costing and inventory.

The output is made to be placed into artwork rather than looked at on a screen: vector PDF and SVG with embedded fonts, sized to the package, alongside high-resolution raster export.

Who we are

Waqar Anwar is the founder and sole developer of Nutrition Fact Label. He is a packaging designer by trade, with retail packaging work for the US and Canadian markets behind him, which is where the frustration with existing label tools came from: the panel is one small element of a package, and it was the element that took the longest to get right. He built and maintains the label engine, the 152 regulatory formats, and the compliance research behind them.

NFL is deliberately a one-person product. That is why the roadmap is narrow and the rounding rules are argued from the regulation text rather than copied from a competitor: there is one person to ask, and his name is on it.

You can reach him on LinkedIn or Facebook, or through the contact page. Support goes to support@nutritionfactlabel.com and is answered by the same person who wrote the code.

Why it is built the way it is

The nutrition-label software market is crowded but old: interfaces from 2010, no dark mode, paywalls in front of the basic FDA formats, and manual data entry for every nutrient. The brief here was a premium tool that is fast, legible in the dark, does the nutrient maths itself, and is priced so a one-person operation can afford it.

The design bias comes from the same place as the product. A panel that is technically compliant but visually wrong still fails on press: the wrong rule weight, a substituted typeface, or a panel that cannot hold its size on a small package. The engine treats those as correctness problems, not aesthetics.

How the regulatory content is checked

Every one of the 152 format pages in the regulation library carries a "Reviewed against" line naming the primary document it was written from, with a link, and a "Page updated" date that comes from the page's own change history rather than being typed in. When the underlying text changes, the page changes and the date moves.

The pages for the most-used formats (the US Nutrition Facts and Supplement Facts panels, Canada, the EU and UK, Mexico, China, India, the Gulf states, Brazil, the Philippines, Codex and others) were rewritten in September 2026 from the primary documents themselves: eCFR and the Federal Register, EUR-Lex, legislation.gov.uk, the Canada Gazette, Mexico's Diario Oficial, China's NHC announcements, FSSAI's gazette compendia, GSO and SFDA publications, the FAO/WHO Codex texts, Brazil's Diário Oficial. Each section on those pages ends with the source it rests on, and the page closes with a Sources list. Where a fact could not be confirmed from a primary document it was left out rather than guessed.

The same rule applies to the guides: every FDA article is written from a dated fact sheet with a source URL and a status tag (final, proposed, voluntary or guidance) on each fact, and a proposed rule is always called a proposal. Our editorial guidelines set this out in full, including how to report an error.

What we care about

Regulatory accuracy. We implement 21 CFR 101.9 rounding and insignificance thresholds exactly. No approximations where the regulation is specific.

Design. The printed label has to look right on a real package: Helvetica, crisp rules, the correct vertical border weights. The editor around it should be calm and uncluttered.

Data portability. Your recipes are yours. JSON import/export, deep-linkable URLs, and no proprietary formats locking you in.

Saying what is free. The free tier is three US FDA formats and unlimited clean JPG export, with no account, no card and no watermark. The other formats and the vector exports are paid. We would rather state that plainly than promise "free labels" and gate the download.

What is on the site, in numbers

152 label formats covering 114 countries and regions, each with its own regulation page. 31 free tools, from a Daily Value calculator to a front-of-pack checker for Canada, the UK, Mexico and Australia. 89 real-food label pages rendered from USDA FoodData Central through the same engine that builds your label. 38 guides on FDA labelling rules. All of it is public and linkable, so you can check our work before you trust it with yours.

Who uses it

Small CPG brands, co-packers, ghost kitchens, nutritionists, and schools teaching food-industry courses. Anyone who needs a compliant label and doesn't need a six-figure ERP to get one. See who uses NFL for the segments and what each one uses it for.

Clean-room implementation

Every format is implemented from 21 CFR 101.9 and FDA specimens directly. No text or code is reproduced from competitor products. See our ATTRIBUTION and CLEAN_ROOM_PROCESS files in the repo for the full trail.