Issuer-disclosed cost position — AISC, C1 cash cost and development capex from primary filings — assembled across African assets and grouped by commodity so like is compared with like. Every figure is Sourced to the bankability layer or marked Absent / Pending. Nothing here is re-modelled or forecast.
This is a disclosure view, not a cost curve. Figures are as disclosed by the issuer (study, guidance or filing) — Afrimintel normalises and source-pins them, it does not re-derive them. Cost metrics are not comparable across commodities: gold AISC is $/oz, copper C1/AISC is $/lb or $/t, graphite is $/t — reading one against another is meaningless. Tables are grouped by commodity for that reason and are not ordered by cost. Capex is absolute, not capital intensity \u2014 a larger figure can simply mean a larger mine; read it against the asset\u2019s production scale, not on its own. And cost figures carry each issuer\u2019s own study vintage and price deck (e.g. gold AISC at decks from ~2023 to 2026), so they are indicative, not strictly like-for-like even within a commodity.
Deliberate boundary. Afrimintel does not build a global cost curve, supply-demand forecast or price-normalised ranking — that is a different game (global publishers / forecasters). This is a cost view: NPV / return figures are deliberately left off it, because same-commodity assets here sit on different study price decks (gold alone spans ~$1,500–$4,000/oz) and a single NPV column invites a ranking that is not valid — those figures live, with their caveats, on the bankability layer. Capex shown is gross initial development capex as disclosed, not capital intensity (capex per unit of capacity) — do not read lower capex as lower-cost without scale. The verifiable slice we stand behind is the African cost disclosure itself: located, normalised, three-state honest.