Sarvam Enters Unicorn Club With $234 Mn Funding Round Led By HCLTech

Sarvam has become India’s 130th unicorn after raising $234 Mn (about ₹2,210 Cr) as part of its $300 Mn Series B funding round at a post-money valuation of $1.5 Bn

The funding round was led by HCLTech, and also saw participation from Bessemer Venture Partners, Khosla Ventures and Peak XV Partners. HCLTech will invest $150 Mn as part of the round

The startup will use the fresh capital for its research efforts, including the development of next-generation AI models focused on agentic AI, coding and cybersecurity applications

The funding round was led by HCLTech, and also saw participation from Bessemer Venture Partners, Khosla Ventures and Peak XV Partners. HCLTech will invest $150 Mn as part of the round.

As per HCLTech’s filing with the exchanges, the company acquired 41,421 equity shares or 10.46% stake in Sarvam for ₹1,427.3 Cr.

The startup will use the fresh capital for its research efforts, including the development of next-generation AI models focused on agentic AI, coding and cybersecurity applications. It also plans to expand its compute infrastructure and scale deployments across sectors.

Founded in 2023 by Pratyush Kumar and Vivek Raghavan, Bengaluru-based Sarvam builds AI models, infrastructure and enterprise products. The startup operates across language, speech, vision and document AI, serving customers in sectors such as banking, insurance, government and defence.

Sarvam’s FY26 revenue, as per unaudited financials shared by HCLTech, stood at ₹45.1 Cr. This represents a multifold jump from ₹1.5 Cr revenue in FY25. Since its inception, Sarvam AI has launched multiple AI models over the years. The journey began in December 2023 with the release of OpenHathi-v1, a fine-tuned, open-source Hindi model built on Meta’s Llama 2 architecture that proved foundational models could be effectively adapted for Indian languages.

Building on this proof of concept, Sarvam shifted focus toward efficiency and accessibility in mid-2024 by launching Sarvam-2B, a lightweight model trained from scratch to deliver high-speed Indic language processing on cost-effective hardware.

By October 2024, they expanded their breadth significantly with Sarvam-1, a highly optimised 2 Bn parameter model capable of seamlessly understanding and generating text across 10 major Indic languages alongside English.

This year, Sarvam entered the frontier LLM space with Sarvam-30B and Sarvam-105B, scaled-up, open-weight models designed for complex enterprise reasoning, multi-turn dialogue, and expansive long-document understanding across more than 22 local languages.

As of now, the startup’s AI systems have been deployed across sectors including banking, insurance, government and defence

As per Sarvam, its conversational AI platform now handles more than 2 Mn interactions daily, while its inference platform processes over 10 Mn API calls every day. Sarvam said the usage on both platforms has grown significantly in recent months.

Sarvam’s voice AI business has emerged as one of its key commercial offerings. Inc42 had previously reported that conversational AI agents account for nearly 80% of the startup’s annual recurring revenue (ARR), which was estimated at about $12 Mn. The startup counts Tata Capital among its customers.

Notably, it is also expanding into emerging infrastructure projects in recent times. Earlier this year, Sarvam partnered with spacetech startup Pixxel to develop Pathfinder, an orbital data centre satellite slated for launch in late 2026.