Sarvam, the Indian AI startup, is broadening its business by introducing a new division named Chanakya. This initiative will concentrate on applied AI services, providing real-world AI solutions to governments, enterprises, and critical sectors where security and reliability are paramount.
Sarvam made the announcement of Chanakya through a post on X (formerly Twitter). The company disclosed that it has been diligently developing full-stack AI capabilities over the past year aimed at addressing significant challenges that impact national interests and large complex organisations.
The company communicated that it is applying its full-stack AI to issues of national importance and intricate enterprises, now scaling up the vertical associated with it – Chanakya. It was also emphasised that the AI system is intended for deployment in air-gapped environments, securely isolated from the internet.
Chanakya is built to analyse diverse data types, such as text and images, and autonomously execute AI workflows reliably in high-pressure, real-world scenarios.
With robust security features, the company is focusing on organisations that do not rely on public cloud services and operate in highly secure settings. The system supports “dual use,” meaning it can be implemented across both enterprise applications and critical sectors such as defence and national security.
As it broadens its enterprise reach, the AI startup is swiftly enhancing its AI services with sophisticated offerings. At the Indian AI Impact Summit 2026, Sarvam introduced its inaugural AI models, Sarvam-30B and Sarvam-105B, which support 22 Indian languages. The company gained significant attention for their performance levels when compared to global AI models.
Additionally, Sarvam showcased AI-driven smart glasses branded Sarvam Kaze and launched its AI-first application named Indus, among various other innovations. The startup has also made strides on a global scale by joining Nvidia’s Nemotron Coalition, collaborating with leading AI firms to develop next-generation open foundation models, thereby entering a wider international initiative to influence the future of open and sovereign AI.