Indian companies are showing growing interest in small language models (SLMs) as enterprises seek AI systems that are cheaper to deploy, easier to control, and better suited for multilingual and domain-specific applications, with companies citing lower latency, reduced compute costs, and greater operational predictability compared to large general-purpose models.

Ganesh Gopalan, Co-Founder & CEO, Gnani.ai, explained that large language models (LLMs) are trained on massive datasets with billions or trillions of parameters, enabling broad reasoning capabilities, complex content generation, and handling diverse tasks across domains. They require significant compute power, often depending on large cloud or data-centre infrastructure. SLMs are designed to be smaller, faster, and more efficient, optimized for specific domains or enterprise tasks.

“SLMs can outperform LLMs where low latency, lower inference cost, privacy, edge deployment, and predictable outcomes matter, like customer support, voice assistants, industry-specific workflows, or on-device AI, because they deliver targeted intelligence with lower resource consumption. SLMs often win where efficiency and specialization are more important than scale,” he said.

Multilingual Usage

Gnani.ai’s own SLMs, he said, have assisted a leading bank to collect over $1 billion in overdue EMIs. The company’s voice-focused SLMs are used across customer support, lead qualification, EMI collection, and insurance renewal workflows, particularly in multilingual enterprise environments in India.

Gartner predicts that by 2027, organizations will implement small, task-specific AI models, with usage volume at least three times more than that of general-purpose LLMs.

For instance, Pankaj Gautam, Chief Technology and Security Officer, Healthcare Global Enterprise, noted that the company is deploying SLMs for language-intensive use cases that require consistency, accuracy, and predictability rather than open-ended reasoning.

“We prefer SLMs over LLMs because they provide tighter domain control, lower latency, reduced operating costs, and easier deployment within secure, compliant enterprise environments. For regulated healthcare operations, this balance delivers more practical and scalable value,” Gautam said.

Cost Efficient

Meanwhile, Vaidam Health is deploying SLMs across targeted enterprise use cases like customer support automation, workflow assistance, document summarization, and internal knowledge management. SLMs offer a strong advantage here because they are faster, more cost-efficient, and easier to deploy at scale than LLMs.

“Unlike generative models that rely on broader generative skills, the task-specific nature of such workflows means that SLMs can work more effectively and offer enhanced security and industry customizability,” said Pankaj Chandna, Co-Founder, Vaidam Health.

Several Indian IT companies are also showing greater interest in SLMs as enterprise adoption is increasingly driven by practical business outcomes. According to the Gnani.ai co-founder, most enterprise use cases prioritize lower deployment costs, faster responses, data privacy, and easier customization.

In a market like India, where businesses often operate across multiple languages and cost-sensitive environments, the focus is shifting from building the largest model to the most efficient and purpose-built model for real-world applications.

“SLMs generally depend less on large-scale data centre infrastructure because they require less computational power for training and inference. Their smaller model size allows them to run on localized servers, enterprise infrastructure, edge devices, or compact GPU deployments. This can be attractive for enterprises seeking lower latency, stronger data privacy, reduced operating costs, and greater deployment flexibility,” Gopalan shared.

Unfamiliar Domains

However, Sushovan Mukhopadhyay, Director Analyst at Gartner, explained, SLMs fail when tasks require broad generalization, deep multi-step reasoning, long-context synthesis, or reliable performance across unfamiliar domains. They are more brittle when prompts move outside the data and workflows they were optimized for. Frontier LLMs remain better for complex coding, research synthesis, strategic analysis, and agentic AI use cases that require planning, tool selection and exception handling.

The Vaidam Health co-founder noted that challenges associated with implementing SLMs in company health centers include ensuring data accuracy, privacy, and relevancy in very sensitive patient-doctor encounters.

The company has encountered inaccurate responses, hallucinations, or incomplete medical context when deploying AI models in healthcare environments. This issue is tackled through the validation of responses created by an AI model through reliable medical databases, frameworks, and human supervision, wherever required. “Our team does not see AI models as a replacement for expert medical knowledge but rather as an additional support to provide patients with accurate information,” he said.

Published on May 26, 2026