Live Market Updates
Latest Financial News
News Feed
1 articles
Personalized
Live Market Updates
Latest Financial News
positive
40 days agoChina’s DeepSeek Trains AI Model at Record-Low $294K
China’s DeepSeek has announced its latest AI model R1 was trained at a cost of just $294,000, using 512 Nvidia H800 chips. This is a fraction of the cost of U.S. large language models, which often run into millions. The development highlights a new phase in the AI race—efficiency over brute force scale. For India, this trend could lower entry barriers for research institutions and startups aspiring to build indigenous AI systems, democratizing access to advanced model development.
positive
40 days agoChina’s DeepSeek Trains AI Model at Record-Low $294K
China’s DeepSeek has announced its latest AI model R1 was trained at a cost of just $294,000, using 512 Nvidia H800 chips. This is a fraction of the cost of U.S. large language models, which often run into millions. The development highlights a new phase in the AI race—efficiency over brute force scale. For India, this trend could lower entry barriers for research institutions and startups aspiring to build indigenous AI systems, democratizing access to advanced model development.
positive
China’s DeepSeek Trains AI Model at Record-Low $294K
about 1 month ago
1 min read
78 words
DeepSeek’s cost-efficient training sets new benchmark; India may find entry routes into AI R&D.
China’s DeepSeek has announced its latest AI model R1 was trained at a cost of just $294,000, using 512 Nvidia H800 chips. This is a fraction of the cost of U.S. large language models, which often run into millions. The development highlights a new phase in the AI race—efficiency over brute force scale. For India, this trend could lower entry barriers for research institutions and startups aspiring to build indigenous AI systems, democratizing access to advanced model development.
China’s DeepSeek has announced its latest AI model R1 was trained at a cost of just $294,000, using 512 Nvidia H800 chips. This is a fraction of the cost of U.S. large language models, which often run into millions. The development highlights a new phase in the AI race—efficiency over brute force scale. For India, this trend could lower entry barriers for research institutions and startups aspiring to build indigenous AI systems, democratizing access to advanced model development.
Companies:
DeepSeek
Nvidia
Tags:
DeepSeek
AI training
DeepSeek
AI training
China AI
low-cost AI
Sep 21, 2025 • 03:56 IST