Investment in the generative AI industry is expected to explode in 2025.

Gartner, a global IT market analysis firm, predicted on the 11th in its report, 'Global Generative AI Model Forecast 2023-2029, 2025 Q2 Update', that the global end-user spending on generative AI models this year will reach $14.2 billion.

This is about 2.5 times the $5.7 billion from last year, and an increase rate of a whopping 148%.


Gartner predicts that spending on general-purpose large-scale language models (LLMs) will grow 141% to $13 billion this year.

At the same time, specialized generative AI models, including domain-specific language models (DSLMs), are expected to grow by a whopping 280%, reaching $1.1 billion.

DSLM is trained based on industry or business process-specific data, enabling higher efficiency and performance in specific fields.

Interest in DSLM is growing.

Gartner predicts that by 2027, more than 50% of enterprise generative AI models will be centered around DSLM, up from 1% in 2024.This is a figure that has grown significantly.

The transition from existing general-purpose models to customized AI models tailored to specific industries and functions is accelerating.

“General purpose LLMs have laid the foundation for generative AI and will continue to see the most investment in the coming years,” said Arunashree Sheppard, senior research analyst at Gartner. “Demand for task-specific DSLMs and vertical AI models is also rapidly increasing.”