A sovereign-AI bet gets a big cheque
Sarvam has raised $234 million at a $1.5 billion valuation, the Bengaluru company announced, making it India's newest AI unicorn at a moment when governments and companies are scrambling for more control over critical AI technology and the computing behind it.
Who's backing it, and why
HCLTech - the IT subsidiary of India's HCL Group - is putting in $150 million as lead strategic investor, with Bessemer Venture Partners joining alongside existing backers Khosla Ventures and Peak XV Partners; Sarvam hopes to raise $300 million in total for the Series B. The plan is to pair Sarvam's models with HCLTech's enterprise relationships, engineering workforce, and software assets to sell AI products to businesses and governments. It's a meaningful step up from the $41 million Sarvam had raised across its seed and Series A more than two years ago, and follows the startup's release of open-source models in 30-billion and 105-billion-parameter sizes earlier this year.
The sovereignty backdrop
The timing is pointed. The debate over AI sovereignty gained fresh urgency when Anthropic disabled access to its newest models, Fable 5 and Mythos 5, after the US government ordered it to suspend use by any foreign national - a reminder of how concentrated frontier-AI access is among a few overseas providers. India is a heavyweight consumer of AI (both OpenAI and Anthropic call it their second-largest market after the US) but has produced few frontier-model contenders, hampered by high compute costs and limited capital, which is exactly the gap Sarvam is trying to fill with a full-stack approach spanning models, inference infrastructure, and applications.
Scale already in production
Sarvam isn't starting from zero: it says its conversational platform handles more than 2 million interactions a day, its inference platform processes about 10 million API calls daily, its speech models transcribe over 500,000 hours of audio a month, and its document AI is digitizing more than 35 million pages. Those tools are deployed at national scale - multilingual voice agents have gathered data from 17 million farmers for India's agriculture ministry, a voice campaign helped an insurer handle renewals for 45 million policyholders, and a large fintech is using its agentic platform to support a salesforce of over 350,000 people. The fresh capital will fund next-generation models focused on agentic, coding, and cybersecurity uses, plus more compute.