From models to infrastructure: Sarvam's next bet is becoming India's AI platform

Sovereign AI startup Sarvam is expanding beyond AI models to build a sovereign, full-stack AI platform, betting on infrastructure, enterprise software and India-hosted compute to challenge global AI leaders.

Sovereign AI startup Sarvam is no longer positioning itself as just another Large Language Model (LLM) developer, but is instead betting on becoming a platform. At its inaugural EPOCH conference, the company outlined an ambitious roadmap to become a full-stack AI platform spanning infrastructure, foundation models, enterprise agents and developer tools, signalling its intent to compete not only with global model makers such as OpenAI and Anthropic but also with hyperscale cloud providers.

The strategy marks a significant shift for the Bengaluru-based startup, which says the next phase of AI adoption in India will be driven less by standalone models and more by complete AI stacks that enterprises can deploy within the country.

"We have established Sarvam more as a platform player. We want to be a horizontal platform player where people can build," the company's founders Pratyush Kumar and Vivek Raghavan said during the press briefing.

Expanding beyond AI models

The company unveiled a suite of products, including voice agents, coding assistants, enterprise work agents, speech recognition, text-to-speech models, document intelligence and AI-powered dubbing. It also introduced model training as a service, allowing enterprises to fine-tune models on proprietary datasets with minimal engineering effort.

Central to that vision is what Sarvam calls a sovereign AI stack: AI models built from scratch, hosted in India and designed for enterprise use cases where data residency and regulatory compliance are critical. The company claimed its voice and chat models are five to ten times cheaper than comparable frontier offerings while keeping customer data within India.

Betting big on AI infrastructure

But perhaps the bigger announcement was on infrastructure. Sarvam disclosed that it currently has access to around 2,000 NVIDIA Blackwell GPUs and aims to scale to 10,000 GPUs, alongside investments in India-based AI infrastructure and sovereign data centres, including a partnership with HCLTech in Odisha.

The startup also committed to building a trillion-parameter foundation model within the next six months, arguing that India must develop frontier AI capabilities rather than rely solely on overseas providers.

A domestic AI partner for enterprises

Interestingly, Sarvam isn't positioning OpenAI as its primary rival. Instead, it believes enterprises increasingly want a domestic partner capable of providing infrastructure, models, deployment and ongoing support as one integrated offering.

"Our goal is to serve a significant portion of the AI tokens consumed in India," the founders said, describing infrastructure, model training and enterprise applications as parts of a single platform strategy.

Without disclosing details, the company said over 500 enterprises, startups and organisations are already using its products, with the bulk of its revenue coming from the private sector rather than government projects.

Why it matters

For India's AI ecosystem, the announcements reflect a broader shift. As the market matures, the competitive advantage may no longer lie in building a better model alone, but in owning the complete AI stack. From GPUs and inference infrastructure to enterprise applications and sovereign deployment, Sarvam is betting that this is where India's next AI race will be won. However, monetisation challenges still persist.