Key facts
- Sarvam AI has raised $234 million in a funding round led by HCLTech.
- The company is valued at $1.5 billion.
- The capital will be used for compute infrastructure, model training, and software development.
- Sarvam focuses on enterprise and government clients, aiming for sovereign AI solutions.
- The startup handles over 2 million daily calls on its conversational AI platform.
Indian artificial intelligence startup Sarvam has successfully raised $234 million in a funding round, achieving a valuation of $1.5 billion. The round, which is expected to reach $300 million, was led by HCLTech, with participation from Bessemer Venture Partners, Khosla Ventures, and Peak XV Partners. HCLTech's significant investment of $150 million marks a strategic move by an Indian IT services company into the foundation model ecosystem, aligning with growing discussions around sovereign AI in India.
Sarvam co-founder Vivek Raghavan stated that the newly acquired capital will be allocated towards building compute infrastructure, training and fine-tuning AI models, scaling inferencing capabilities, and developing software layers for enterprise and government clients. He highlighted the company's current scale, noting that its conversational AI platform handles over 2 million calls daily and that it processed more than 300 million API calls in the previous month.
Raghavan emphasized the importance of sovereign AI, drawing parallels to restricted access to advanced global models. He noted that while the raised capital is substantial for India, it remains relatively small in a global context, necessitating continued capital efficiency alongside efforts to develop frontier capabilities. The company's strategy involves playing across all layers of AI development, from compute infrastructure to large-scale inferencing and agentic orchestration, to achieve self-reliance.
Sarvam's focus is primarily on enterprise and government sectors, differentiating itself from consumer-focused AI products. Raghavan believes that approximately 95% of AI needs can be met through sovereign models, which can also extend to enterprise-level sovereignty. The company aims for a significant portion of token consumption in India to originate from Sarvam, positioning it as a reliable alternative to global AI models, akin to the success of UPI in India.
Pankaj Mitra of Bessemer Venture Partners expressed confidence in India's AI decade, viewing Sarvam as a major bet within the country's technological landscape. He highlighted that mission-critical, regulated enterprise use cases will necessitate sovereignty, which can be achieved through local data and fine-tuned models. Mitra sees Sarvam as a key player in diffusing AI across citizen, enterprise, and government workflows.