KissanAI转发了
Is India finally on the brink of its own large language model (LLM) revolution? With the recent launch of AI Kosha and the IndiaAI Compute Portal, the ?10,370 crore IndiaAI mission is taking a bold step toward AI sovereignty. IT Minister Ashwini Vaishnaw emphasized, “We provide the cheapest compute in the world at ?67 per GPU hour.” This massive investment is set to power India’s AI ambitions and challenge Western and Chinese dominance. The government has received 67 proposals, with 22 focused on LLM development, indicating a strong push to build homegrown AI models. Key players such as Sarvam AI, Krutrim, Zoho, CoRover.ai, and International Institute of Information Technology Hyderabad (IIITH) are leading the charge, while startups like Genloop and GFuture Tech Pvt. Ltd. are set to submit their proposals by March 15. With 18,693 GPUs allocated under public-private partnerships, companies like Jio Platforms Limited (JPL), Tata Communications, and Yotta Data Services Private Limited are providing the backbone for India's AI revolution. However, the big question remains—will this momentum translate into a game-changing Indian LLM, or will it be another six-month reset? The coming months will be crucial in determining whether India can build a model that truly rivals the likes of OpenAI and DeepSeek AI, securing its place on the global AI map.?