Bengaluru-based Sarvam AI secures a $75 million Series B extension, with Nvidia making its debut as a strategic investor, boosting India’s sovereign AI infrastructure ambitions.
Bengaluru-based Sarvam AI has deepened its latest fundraising with a $75 million Series B extension that brings Nvidia on board as a strategic investor, a notable first for an Indian company. The startup, which is positioning itself as a builder of sovereign AI infrastructure for India, now sits at an estimated valuation of about $1.5 billion after the new capital was added to its earlier round, according to the company and reports on the transaction.
The fresh money builds on Sarvam’s earlier $234 million Series B first close, which was led by HCLTech and helped lift the company to unicorn status. That round was announced in June at a $1.5 billion post-money valuation, with HCLTech contributing $150 million and Bessemer Venture Partners joining alongside existing backers Khosla Ventures and Peak XV Partners. Sarvam said the combined financing was intended to support research on its next frontier model, particularly for agentic, coding and cybersecurity use cases, alongside greater access to compute at scale.
According to the latest transaction details reported by the company and filings cited in local media, Nvidia is the largest investor in the extension, contributing roughly $25 million for a stake of about 1.66 per cent. Glade Brook Capital also participated, while other investors included Gaja Capital, Indigo Ventures, and a number of angel investors and family offices. The extension takes Sarvam’s total Series B haul to more than $309 million, and the company says the proceeds will help it expand foundation model training, compute capacity and enterprise and government-facing products.
Sarvam was founded in August 2023 by Vivek Raghavan and Pratyush Kumar, both veterans of India’s digital public infrastructure efforts. The startup has built large language models and AI infrastructure focused on Indian languages and local deployment needs, and it was selected earlier this year by the central government to develop India’s sovereign foundation AI model under the IndiaAI Mission. Its broader pitch is that strategic capital from chipmakers and infrastructure players can help secure the compute supply needed to train a model exceeding one trillion parameters from scratch in India.
Disclaimer: This article is intended to inform and educate, not to recommend or endorse any financial product, investment or strategy. Please consider your own financial circumstances and seek professional advice where appropriate before making financial decisions.
