Gnani AI launches Artha in New Delhi for secure LLMs
Technology
Gnani AI just launched Artha in New Delhi, a made-in-India AI platform designed so businesses and government can run powerful language models securely on their own systems.
Alongside Artha, they introduced the Evon 3.3 model (with 30 billion parameters) and the Plexus platform, all tailored to handle sensitive data and meet strict regulations in fields such as banking and insurance.
Evon 3.3 outperforms Sarvam on MILU
Evon 3.3 is trained on more than 2 trillion tokens across 11 Indian languages, outperformed Sarvam's 30-billion-parameter model and 105-billion-parameter model in 10 of 11 languages on MILU while using much less computing power, thanks to smart technology that cuts token usage by 20%.