Two-mode alternative-NIP calculator: formulation mode (does my label match my recipe?) and label mode (bracket-aware ingredient statements, category priors, declared-% hard constraints).
Compare 60 commodities · explore blend chemistry and NPN · quality tiers (Tier C compositional / Tier B matched-coefficient) with the honest flags.
Data sources (UniProt · USDA FDC · AFCD) → build → engine → surfaces, with the documentation band.
Sequence-derived k is for process control, spec setting and research reporting — the declared factor on a label is usually prescribed by regulation, and neither Tier C nor Tier B of the quality score is a measured DIAAS. Every protein traces to a verified UniProt sequence, every share to a documented source, every coefficient to a citation.