The origin of large I-type batholiths remains a disputed topic. One model states that I-type granites form by partial melting of older crustal lithologies (amphibolites or intermediate igneous rocks). In another view, granites are trapped rhyolitic liquids occurring at the end of fractionation trends defining a basalt–andesite–dacite–rhyolite series. This paper explores the thermal implications of both scenarios, using a heat balance model that abstracts the heat production and consumption during crustal melting. Heat is consumed by melting and by losses through the surface (conductive or advective, as a result of eruption). It is supplied as a basal conductive heat flux, as internal heat production or as advective heat carried by an influx of hot basalt into the crust. Using this abstract approach, it is possible to explore the role different parameters play in the balance of granites formed by differentiation of basalts or by crustal melting. Two end-member situations appear equally favourable to generating large volumes of granites: (1) short-lived environments dominated by high basaltic flux, where granites result mostly from basalt differentiation; and (2) long-lived systems with no or minimal basalt flux, with granites resulting chiefly from crustal melting.