Worldwide end-user spending on generative AI (GenAI) models is forecast to reach $14.2 billion in 2025, according to Gartner. Of this, $1.1 billion will be spent on specialized GenAI models, such as domain-specific language models (DSLMs), marking a 279% year-on-year growth in this subsegment.
Foundation models — including large language models (LLMs) — will continue to dominate GenAI investments, growing from $5.4 billion in 2024 to over $13 billion in 2025, though their growth rate will slow from 306% to 141%. In contrast, specialized models are gaining momentum rapidly due to their improved performance and relevance for specific industries and business functions.
Gartner predicts that by 2027, more than 50% of GenAI models used by enterprises will be domain-specific, a massive leap from just 1% in 2024.
“While foundation models remain the largest spending category, domain-specific models are emerging as preferred options for targeted enterprise use cases due to their enhanced reliability, cost efficiency, and business relevance,” said Arunasree Cheparthi, Senior Principal Research Analyst at Gartner.
This surge in spending reflects a growing enterprise focus on applying GenAI to tailored business outcomes across industries.