Enterprise AI Spending Pivots Toward Low-Cost Open Models

New market data indicates a dramatic shift in enterprise artificial intelligence spending as corporate technology executives seek to control ballooning API invoices. While proprietary frontier models dominated early enterprise adoption, top CIOs are now routing routine tasks to lower-cost, highly optimized open-weights models like LG's EXAONE 2.0 and specialized fine-tunes.

Companies deploying AI across millions of customer support interactions or internal code parsing pipelines report that using top-tier cloud models for basic classification tasks yields diminishing financial returns. By implementing intelligent model routing, systems analyze query complexity before sending low-risk requests to compact, on-premise hardware.

This trend is accelerating the growth of local enterprise model hubs, where customized open models deliver near-parity performance on domain-specific datasets at a fraction of the token cost, reshaping corporate software budgets.