EY refers to the global organization, and may refer to one or more, of the member firms of Ernst & Young Global Limited, each of which is a separate legal entity. Ernst & Young Global Limited, a UK company limited by guarantee, does not provide services to clients.
Recent Searches
Building inclusive and scalable AI through the rise of Small Language Models.
In brief
- For focused enterprise use cases, training small language models (SLMs) can be faster and more cost-efficient than large GPT-scale models.
- SLMs can run on private servers or edge devices, enabling data privacy, compliance and predictable costs, which are key for regulated sectors.
- In India, purpose-built SLMs enable startups to quickly develop chatbots for Indian languages and domain-specific AI solutions for real business needs.
In a country defined by linguistic diversity and limited digital infrastructure, small language models (SLMs) are emerging as the practical foundation for AI growth. Unlike massive, large language models (LLMs), SLMs are smaller, more adaptable and more efficient to train. With millions to a few billion parameters, they perform remarkably well when fine-tuned for Indian languages or specific domains, making them a cornerstone of the next wave of AI in India.
SLMs as India’s competitive advantage
The increasing adoption of SLMs in India is driven by three major factors:
LLM challenges in India
India faces structural hurdles in building frontier LLMs. The digital content in Indian languages remains fragmented, especially for technical subjects. High compute costs, limited GPU availability and the still-developing R&D ecosystem add further constraints. Additionally, the rapid evolution of global AI technology makes it difficult for India to maintain pace in large-scale model development.
The hybrid path forward
India’s strategic edge lies in combining both models. SLMs serve as the operational backbone for high-volume, regulated and localized applications, while LLMs act as reasoning engines for broader problem-solving. Synthetic data generation is helping bridge data gaps, creating sovereign and auditable AI systems. With the rise of SLMs in India, the nation is setting a model for scalable, inclusive and resource-conscious AI innovation.
The article is also contributed by Yash Dandavate, Manager, Technology Consulting, EY India.
Summary
Focusing on SLMs positions India to lead the next wave of AI through inclusive, efficient and scalable innovation. By combining local relevance with cost-effective deployment, SLMs can bridge linguistic and infrastructural gaps, enabling enterprises and startups to build practical, compliant and impactful AI solutions for global adoption.
How EY can help
-
Our Consulting approach to the adoption of AI and intelligent automation is human-centered, pragmatic, outcomes-focused and ethical.
Read more -
Drive business growth with technology transformation solutions at EY - integrating tech, operations & architecture to build agile, data-driven organizations
Read more -
EY India delivers technology solutions to drive digital transformation, improve efficiency, manage risk, and support industry-specific business goals.
Read more
The AIdea of India: Outlook 2026