Summary: On June 15, Bengaluru-based Sarvam AI confirmed a $234 million Series B first instalment at a $1.5 billion valuation, making it India’s newest AI unicorn. HCLTech led the round with $150 million for roughly a 10.5% stake, while Bessemer, Khosla Ventures, and Peak XV also invested; the company plans to raise the round to $300 million. The arrangement pairs Sarvam’s Indian-language model stack with HCLTech’s enterprise distribution to accelerate commercialization for corporate and government customers.
India’s newest AI unicorn has a name. On June 15, Sarvam AI — headquartered in Bengaluru — confirmed a $234 million raise that puts the company’s valuation at $1.5 billion. HCLTech, the IT services arm of Indian conglomerate HCL Group, headed the round as lead strategic investor with a $150 million check. Bessemer Venture Partners came in as well, and both existing backers, Khosla Ventures and Peak XV Partners, chose to reinvest.
Sovereign AI hardened into official doctrine over the course of 2026, and against that backdrop the deal’s real story is not its size but its shape. An IT services giant has taken the strategic-investor role, plugging a model startup straight into enterprise customers around the world. The anxiety — shared by governments and corporations alike — over who commands critical AI technology and the compute beneath it is turning “local models plus local channels” into an asset the market knows how to price.
Inside the Round: Three Investor Types, One Series B
Sarvam is raising a Series B. The plan is to bring the total to $300 million; the $234 million now announced is only the first instalment, and Reuters reports that HCLTech’s investment translates into a stake of roughly 10.5%. Barely more than two years ago, the company’s seed and Series A rounds combined came to $41 million. A valuation leap of this magnitude reflects two forces at once: the gathering heat of the “India AI national champion” narrative, and the hard reality that sovereign compute and model development consume capital on an industrial scale.
Who writes the checks says as much as the total. HCLTech arrives as the strategic investor carrying customer reach; Bessemer and Khosla, venture firms from the United States, are in effect underwriting the technical roadmap; Peak XV, a fund with deep roots in the Indian market, is wagering that domestic demand takes off. Three pools of capital chase three different prizes, and between them they carry Sarvam into the short list of homegrown model companies valued north of a billion dollars. The provenance of the money says something too: capital from a listed IT services firm is patient and revenue-adjacent — it looks forward to co-selling rather than waiting on a multiple.
Sovereign AI Meets Indian-Language Models
The company runs the whole stack itself: it builds models, operates inference infrastructure, and delivers enterprise applications. Earlier this year it open-sourced two models — 30 billion and 105 billion parameters — tuned for Indian languages and India-specific use cases, and its products already run inside banking, insurance, government services, and defense.
Context explains the excitement: OpenAI and Anthropic each rank India as their second-largest market after the United States. Even so, the frontier-model race has left India without a heavyweight contender of its own — compute is costly, and research talent keeps migrating to labs abroad. Sarvam is one of a very small number of candidates taken seriously, and the worldwide contest over who controls critical AI technology and computing infrastructure is the mightiest tailwind a raise like this could ask for.
Why an IT Services Giant Took the Lead Spot
There is more to HCLTech’s involvement than capital. The blueprint is to pair Sarvam’s models with HCLTech’s enterprise relationships, engineering workforce, and software holdings, then assemble AI products for companies and governments. For HCLTech, the deal secures access to model capability in a market sprinting toward AI-delivered services; for Sarvam, it is a shortcut to commercialization that avoids the grindingly long direct-sales cycle.
Pair a model startup with a systems integrator, and you may be looking at the standard sovereign-AI script taking shape: the startup furnishes the models and the research velocity, while the big services firm owns deployment, day-to-day operations, and the billing relationship.
Startups and Enterprise Buyers, Take Note
Three lessons stand out. Sovereign AI has left the slogan stage and entered the capital-allocation stage: strategically aligned money, typically close to government priorities, now serves as the main fuel for model startups based outside the United States. Distribution builds the moat: in markets where OpenAI and Anthropic sell face-to-face, a homegrown model startup stands out not on parameter counts but on language coverage, onshore deployment, and industry integration — precisely what an IT services major is good at. And enterprise AI procurement teams should adjust their maps: India’s data-sovereignty requirements, joined with its linguistic diversity, are producing a supply chain that runs on logic distinct from both the American and the Chinese versions. The round doubles as a benchmark — how Sarvam performs through HCLTech’s channel will shape what sovereign AI actually means commercially, past the data-center headlines.