Sarvam AI Approves Nvidia-Backed $74M Series B Extension
Sarvam AI approved a roughly ₹697.8 crore ($74 million) Series B extension at a July 24 shareholder meeting, with Nvidia slated to invest ₹237.9 crore, according to MCA filings reported by Inc42. Reports differ on whether the issuance is proposed or closed; it follows Sarvam's June first close and a separate plan for a trillion-plus-parameter model on 10,000 Nvidia Blackwell GPUs.
Sarvam AI approved a roughly ₹697.8 crore ($74 million) extension to its Series B at a July 24 extraordinary general meeting, according to Ministry of Corporate Affairs filings reviewed by Inc42. The reported issuance would add Nvidia, Glade Brook, Gaja Capital, Indigo Ventures, and other investors to the Bengaluru company's ongoing round.
Inc42 reported that Nvidia is slated to invest ₹237.9 crore, followed by Glade Brook at ₹190.3 crore, Gaja Capital at ₹95.1 crore, and Indigo Ventures at ₹30 crore. The remaining capital would come from high-net-worth investors and family offices. LDS did not retrieve the corporate filing directly, so those figures remain attributed to the publications that reviewed or described it.
Proposed issuance, not a uniformly reported close
The transaction's stage is described inconsistently. Inc42 says Sarvam is set to raise the capital after the shareholder approval. Free Press Journal calls the extension closed and reports an issuance of 20,244 Series B preference shares plus seven equity shares at ₹344,570 each. India Today says Sarvam secured Nvidia's backing and quotes co-founder Pratyush Kumar confirming the chipmaker as a strategic investor.
The evidence supports Nvidia's participation and the approved issuance, but not a single uncontested closing description. Framing the event as an approved extension preserves that distinction until Sarvam or a later filing confirms completion.
Sarvam's official June 15 announcement provides the funding context. The company said it had raised $234 million in the first close of a planned $300 million Series B at a $1.5 billion post-money valuation, with HCLTech and Bessemer Venture Partners joining existing investors Khosla Ventures and Peak XV Partners. The August extension is part of that broader round; Sarvam and Nvidia have not published a retrieved exact-event announcement for it.
Capital and compute are converging
The financing was reported days after Sarvam outlined a trillion-plus-parameter foundation-model project. India Today reported that Sarvam plans to train it on 10,000 Nvidia Blackwell GPUs for coding, cybersecurity, scientific research, and enterprise applications. Inc42 separately reported a six-month development target. Those are company plans, not a completed model or independently verified performance result.
Nvidia's reported role links capital, accelerators, and training software in one relationship. For infrastructure teams, the important questions go beyond the announced GPU count: cluster availability, interconnect and storage design, training reliability, data composition, evaluation, and the cost of serving the resulting model. None of the retrieved exact-event reports provides those details, so the investment signals an infrastructure commitment without establishing how the proposed system will perform.
Key Points
- 1Sarvam approved a roughly ₹697.8 crore ($74 million) Series B extension, with Nvidia reportedly allocated ₹237.9 crore of the issuance.
- 2Reports agree on Nvidia's participation but differ on whether the extension is still proposed or already closed, so the article treats it as an approved issuance.
- 3The financing follows Sarvam's plan for a trillion-plus-parameter model on 10,000 Blackwell GPUs, but architecture, data, evaluation, and operating details remain undisclosed.
Scoring Rationale
Nvidia's reported participation in Sarvam's Series B extension links financing with a large planned Blackwell training program. The event is material to India's AI infrastructure ecosystem, while the transaction stage and technical implementation still require careful attribution.
Sources
Public references used for this report.
View 1 more source
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
