India launches Gnani Artha sovereign AI stack for enterprises

India’s Vice President CP Radhakrishnan has launched Gnani Artha, a sovereign AI stack designed for Indian enterprises and public institutions. Unveiled at Uprashtrapati Bhavan in New Delhi, the platform aims to bolster technological self-reliance and support India’s goal of becoming a developed nation by 2047.

The Artha stack integrates two primary components: Evon v3.3, a 30-billion-parameter open-weights large language model, and Plexus, an agentic AI platform for workflow automation. Evon v3.3 was trained on over 2 trillion tokens and natively supports 11 Indian languages. The model utilizes a Mixture-of-Experts architecture and is designed for high efficiency, capable of running on a single graphics card.

Gnani AI CEO Ganesh Gopalan stated that the company plans to expand the model family to include 70-billion and 100-billion parameter versions and increase language coverage from 11 to 22 in the near future. The initiative is supported by the government’s India AI Mission, which seeks to develop indigenous AI solutions to reduce dependence on global technologies.