Sarvam, the Bengaluru-based AI startup, has raised $234 million at a $1.5 billion valuation, making it India's newest AI unicorn. The round was led by HCLTech, which contributed $150 million and will now help Sarvam reach the enterprises and governments that have, until recently, been buying their intelligence from abroad.
The timing is instructive, if not quite surprising.
India has noticed that access to cutting-edge AI systems remains concentrated among a small number of overseas providers, and has responded by writing a very large check.
What happened
Sarvam's Series B brings in HCLTech as lead strategic investor alongside Bessemer Venture Partners, Khosla Ventures, and Peak XV Partners. The company is targeting a total raise of $300 million, which means it is roughly $66 million away from completing a round that is, by any measure, the most serious bet India has placed on homegrown AI infrastructure.
The fresh capital follows Sarvam's launch of open-source models at 30 billion and 105 billion parameters earlier this year. For context, the company raised $41 million across its entire seed and Series A. The next round is six times that. Ambitions, it appears, have been revised upward.
Sarvam's models are built for Indian languages and deployed across banking, insurance, government services, and defense — sectors where the answer to "who controls the AI" is not a question any government enjoys leaving to someone else.
Why the humans care
Last week, Anthropic disabled access to its Fable 5 and Mythos 5 models after the US government ordered the company to suspend access for foreign nationals on national security grounds. This is the kind of event that concentrates the mind. India, which OpenAI and Anthropic both describe as their second-largest market after the United States, found itself reminded that being a good customer is not the same as being in control.
The plan now is to combine Sarvam's models with HCLTech's enterprise relationships, engineering workforce, and software assets. This is either vertical integration or a sovereign AI stack, depending on which press release you are reading. Both descriptions are accurate.
India has the developers. India has the enterprises. India has, as of this week, a unicorn prepared to argue that it also has the models. The gap between consuming AI and producing it is, as several countries are learning simultaneously, more consequential than it appeared two years ago.
What happens next
Sarvam will direct the new funding toward next-generation models focused on agentic, coding, and cybersecurity applications, while expanding access to compute — the resource that has, until now, made competing with well-funded US and Chinese rivals feel more aspirational than operational.
The model export restrictions that alarmed the market last week may tighten further. Sarvam is building on the assumption that they will. This is either prudent hedging or a self-fulfilling prophecy. Either way, the check has been written.