Sarvam AI secured the capital in a recent funding sequence to construct its sovereign technology in New Delhi. This transaction anchors the enterprise at a $1.5 billion valuation after a single duration of expansion.
HCLTech led the sequence with a $150 million interest to transition away from the legacy outsourcing architectures. This capital infusion triggers a seven-fold escalation in valuation for the startup over twelve months.
And CEO C. Vijayakumar targets a position in the development of sovereign AI stacks for the global market. He masks legacy service methods with new AI assets to avoid a reduction in the traditional revenue.
Bessemer Venture Partners and Nvidia entered the transaction alongside the legacy backers like Khosla and PeakXV. They seek scale in the AI factory architecture for the mining and the port management operations.
Now the enterprise reveals its Sarvam 105B model to address the specific requirements of the Bharat audience. This system uses Mixture of Experts technology to process twenty-two official languages across the Indian region.
But Sarvam still relies on the Nvidia H100 chips and the hardware from Yotta Data Services to execute. So, the enterprise must manage the high-power expenses and the rising expense of elite technical talent.
Yet Indian AI startups captured $3.94 billion in early 2026 as the state pushes its sovereign objective. Success stands at risk.
Discussion