Sarvam raises $234 million to expand AI stack

Sarvam raised $234 million at a $1.5 billion valuation, according to TechCrunch. TechCrunch reports HCLTech provided $150 million as the lead strategic investor - acquiring over 10% of the company per Business Standard - and that Sarvam is seeking to expand the Series B to $300 million. Business Standard confirms participation from Bessemer Venture Partners, Khosla Ventures, and Peak XV Partners alongside HCLTech. TechCrunch reports Sarvam launched 30-billion and 105-billion-parameter models earlier this year, designed for Indian languages and deployed across banking, insurance, government services, and defense. Moneycontrol reports Sarvam trained its sovereign models on roughly 3,400 Nvidia H100 GPUs and brought an on-premises Blackwell cluster online.

The fundraising and reported compute footprint place Sarvam in the small set of startups attempting a vertically integrated AI stack, combining model development, inference infrastructure, and enterprise applications. Industry-pattern observations: Building and operating multi-billion-parameter models at scale requires sustained access to GPU fleets, custom clusters, and heavy engineering investment, and practitioners note that on-premises Blackwell-class clusters and thousands of H100s materially change cost and deployment trade-offs compared with pure cloud-first startups.

Public reporting links Sarvam's raise to a renewed Indian focus on AI sovereignty after export-control actions that affected access to Anthropic models Fable 5 and Mythos 5, as reported by Economic Times and NDTV. Observers quoted by Economic Times and NDTV framed the export controls as a catalyst for calls to develop indigenous capabilities. TechCrunch frames HCLTech's participation as giving Sarvam a strategic enterprise partner to commercialize models inside Indian enterprises and government agencies.

Practitioners should track whether the $300 million close completes, how Sarvam's agentic AI and coding/cybersecurity model roadmap progresses, and whether government procurement - announced across banking and defense - translates into measurable deployment scale.