Gartner has recently released a forecast that global end-user spending on generative AI models will skyrocket to $14.2 billion in 2025, reflecting an explosive 148.3% year-over-year growth.

The 2024 spend of $5.7 billion marked a significant milestone in GenAI adoption, but 2025 is expected to bring further acceleration, particularly in specialized models tailored for specific industries and business functions.

While general-purpose foundation models (such as LLMs) still dominate in terms of market share, a fast-emerging category of domain-specific GenAI models, also known as DSLMs (Domain-Specific Language Models), is seeing exponential growth. Gartner estimates that spending on DSLMs will rise from $302 million in 2024 to $1.1 billion in 2025, a 279.2% increase.

These models, fine-tuned using targeted data and workflows, are gaining traction for their performance, relevance, and cost efficiency in enterprise deployments.

Why It Matters: This forecast reflects another huge shift in enterprise AI strategy, from experimentation with general models to the integration of AI into mission-critical, domain-specific operations. As DSLMs gain maturity and deliver tangible ROI, their adoption signifies that enterprises are done exploring AI andoperationalizing GenAI rather than simply exploring it. This transition has deep implications for software vendors, cloud providers, and IT decision-makers focused on vertical solutions.

• 2025 GenAI Spending to Exceed $14 Billion: Gartner projects that total end-user spending on GenAI models worldwide will reach $14.2 billion in 2025, up from $5.7 billion in 2024. This reflects a 148.3% growth rate year-over-year, demonstrating that GenAI is moving from pilot projects to scaled deployment across industries.

• Foundation Models Lead but Growth Slows: Spending on foundation GenAI models is forecast to increase from $5.4 billion in 2024 to $13.05 billion in 2025. While this segment continues to account for the majority of spending, its growth rate (141%) is notably lower than that of domain-specific models, indicating a relative plateauing of broad-model investments.

• Specialized GenAI Models Surge 279%: Spending on domain-specific GenAI models is expected to rise from $302 million in 2024 to $1.146 billion in 2025, a 279.2% jump. These models are designed with industry-specific vocabularies, datasets, and operational logic, which makes them particularly valuable in regulated or niche sectors like healthcare, legal, and finance.

• Enterprise Use of DSLMs to Surpass 50% by 2027: Gartner anticipates that more than half of GenAI models used in enterprise environments will be domain-specific by 2027, a steep climb from just 1% in 2024. This signals a shift toward customized AI solutions optimized for industry workflows and compliance requirements.

• Strategic Value of DSLMs: According to Gartner Analyst, Arunasree Cheparthi, organizations are increasingly turning to vertical models due to their improved performance, lower cost of operation, enhanced reliability, and relevance to enterprise-specific use cases. These advantages make DSLMs more sustainable and scalable in long-term AI strategies.

• Implications for Vendors and CIOs: The rising demand for DSLMs suggests a growth opportunity for software vendors offering vertical AI products. It also pushes CIOs and IT leaders to rethink their GenAI roadmaps, focusing more on integration, model fine-tuning, and governance within the unique contexts of their industries.

Go Deeper -> Gartner Forecasts Worldwide End-User Spending on GenAI Models to Total $14.2 Billion in 2025 – Gartner