Emerging from stealth mode, Sarvam AI Reveals $41 Million Funding for Rapid Development of Full-Stack Generative AI Solutions in Just Five Months.

Sarvam AI has successfully raised $41 million through both Seed and Series A financing rounds. The Series A round was spearheaded by Lightspeed, while they partnered with Peak XV Partners in leading the Seed round. Notably, Peak XV and Khosla Ventures were also significant contributors to the Series A funding.

Based in Bengaluru, Sarvam AI, led by Vivek Raghavan, is developing extensive language models tailored for Indian languages, as shared with TechCrunch. Additionally, the startup is working on a platform enabling businesses to harness these Large Language Models (LLMs) for their applications.

— “everything from writing an app, deploying it to popular channels, observing logs, and custom evaluation,” he said.

Sarvam AI prioritizes the development of Large Language Models (LLMs) with voice as the default interface for the Indian market. This strategic approach, coupled with a commitment to supporting local languages, addresses the unique requirements of the Indian market.

Raghavan explained, “We need to modify the architecture of current open models and customize their training for the new language. The benefit is that the resulting models are more efficient, in terms of tokens consumed, for comprehending and generating content in Indian languages compared to any existing Large Language Models (LLMs).”

Founded approximately five months ago, Sarvam was established by Raghavan and Pratyush Kumar, who had prior roles at the Nandan Nilekani-backed AI4Bharat of IIT Madras. Notably, Raghavan brings over a decade of experience from UIDAI, the organization overseeing India’s ubiquitous identity system, Aadhaar.