Vice-President Launches Gnani Artha Sovereign AI Stack

Key facts

• Vice-President C P Radhakrishnan launched the sovereign AI stack 'Artha' developed by Bengaluru-based startup Gnani at his official residence in New Delhi on August 28, 2026.
• Artha is an end-to-end sovereign AI stack for Indian enterprises and public institutions, comprising the open-weight foundational model Evon v3.3 and the agentic AI platform Plexus.
• Evon v3.3 is a 30-billion-parameter model with roughly 3.5 billion active parameters on any given token, indicating a mixture-of-experts (MoE) architecture.
• Evon v3.3 supports over 11 Indian languages and its model weights are released under the Apache 2.0 open-source licence.
• Gnani rebuilt Evon v3.3's tokeniser for Indian scripts, claiming it needs about 20 per cent fewer tokens per Indian-language word than the GPT-5 family tokeniser and less than half of byte-level tokenisers such as DeepSeek, Llama and Qwen.
• Evon v3.3 is designed for self-hosting on a single node to help banks, insurers and government bodies meet data residency requirements.
• Plexus is an agentic AI platform that allows users to build and deploy AI agents through natural language prompts, with tool calling and autonomous operation across documents, systems and conversations.
• According to Gnani, Evon v3.3 outperformed similarly sized and 105-billion-parameter Indic models on the Indian-language benchmark MILU.
• Gnani AI is one of the 12 local entities selected under the India AI Mission to build sovereign LLMs.
• The India AI Mission was approved by the Union Cabinet in 2024 with an initial outlay of Rs 10,372 crore.
• At the India AI Impact Summit in February 2026, Gnani launched Vachana TTS, a text-to-speech model that can clone human voices across 12 Indian languages using under 10 seconds of reference audio.