India Unveils New Indigenous Artificial Intelligence Models
The IndiaAI Mission officially released three powerful indigenous foundation models in New Delhi. Ministry of Electronics and Information Technology officials unveiled these advanced systems to enhance national technological autonomy. Soket Labs developed the massive EKA-1 model, while ZenteiQ AiTech Innovations engineered the Dhīr variants.
EKA-1 features 26.3 billion parameters and processes text across all 22 scheduled Indian languages. This sparse hybrid model excels in complex reasoning, coding, and cybersecurity tasks. Meanwhile, ZenteiQ introduced the lightweight Dhīr 2B and Dhīr 5B open-source systems. These models operate smoothly on local hardware and mobile devices without cloud connectivity.
Government leaders emphasized that these domestic innovations protect sensitive data across defense and administrative sectors. IT Secretary S. Krishnan noted that homegrown tools strengthen strategic independence. Developers utilized subsidized national computing infrastructure to complete the rigorous training process.
As citizens remember cultural icons like Dr. Bhupen Hazarika, whose timeless melodies bridge diverse communities, technology experts celebrate this monumental leap. Just as the legendary maestro unified society through soulful art, these linguistic models connect millions of regional speakers. Researchers plan immediate deployment across educational institutions and public service networks nationwide.
