Vice President of India C. P. Radhakrishnan launched Gnani Artha, an indigenous sovereign Artificial Intelligence stack, at the Uprashtrapati Bhavan in New Delhi on 28 August 2026. The stack was developed by Bengaluru-based voice and conversational AI startup GNANI AI and is designed for data residency and sovereignty for Indian public institutions and enterprises.
Gnani Artha Architecture
Gnani Artha combines two core components, namely Evon v3.3 and Plexus. Evon v3.3 is a foundational language model, while Plexus is an enterprise-grade agentic AI orchestration platform.
Evon v3.3 is an open-weights model with 30 billion parameters. It uses a Mixture-of-Experts architecture and activates about 3.5 billion parameters per token.
Language Support and Model Access
Evon v3.3 supports 11 Indian languages. GNANI AI has made the model weights available on request through Hugging Face under the Apache 2.0 licence.
The model tokenizer for Indian scripts uses about 20% fewer tokens per Indian-language word than the GPT-5 family. It also uses less than half the tokens required by models such as DeepSeek, Llama, and Qwen.
Plexus and Enterprise Use
Plexus is a commercial platform for deploying autonomous, goal-oriented AI agents across internal workflows. It is designed for secure use by organisations that require controlled AI deployment within enterprise systems.
Important Facts for Exams
- Artificial Intelligence refers to computer systems that perform tasks associated with human intelligence, including language processing and decision-making.
- A Mixture-of-Experts model activates only selected parts of the network for each token or task.
- Open-weights models allow access to model parameters, while open-source licences define conditions for use and distribution.
- Hugging Face is a widely used platform for hosting and sharing machine learning models and datasets.
IndiaAI Mission Context
The IndiaAI Mission is a Government of India initiative related to the development of artificial intelligence capabilities in the country. Sovereign AI stacks are associated with domestic control over data, infrastructure, and model deployment.
