RT team
Bengaluru-based Sarvam AI has been selected by the Government of India, under the IndiaAI Mission, to develop the country’s first sovereign large language model (LLM). This indigenous model will be built entirely in India using local infrastructure and talent, aiming to enhance strategic autonomy and domestic innovation.
The model will be developed from scratch with strong reasoning capabilities, optimized for voice interaction, and fluent in multiple Indian languages. It is intended for secure, large-scale deployment, ensuring that AI feels familiar to citizens and helps enterprises maintain data sovereignty.
Dr. Vivek Raghavan, Sarvam’s co-founder, stated that building a national AI foundation is crucial, emphasizing that their goal is to create multi-modal, multi-scale models capable of powering a range of applications across India. The initiative supports Sarvam’s broader mission to empower Indian builders with cutting-edge AI solutions.
As part of the project, Sarvam is developing three variants: Sarvam-Large for complex tasks, Sarvam-Small for real-time applications, and Sarvam-Edge for on-device AI. The company is collaborating with AI4Bharat at IIT Madras, a leader in Indian language AI research, to bring this vision to life.
Sarvam has already proven its expertise in building foundational models tailored for Indian use cases, offering scalable and cost-effective solutions. Union Minister Ashwini Vaishnaw expressed confidence that Sarvam’s models would compete effectively with global LLMs, marking a significant milestone in India’s AI journey.
