Chinese AI company DeepSeek has released its latest artificial intelligence model, aiming to improve performance while cutting computing costs. According to Beyond Time News, the new launch marks another step forward in the global AI competition.
The company first gained international attention last year with a low-cost AI model that performed at a level similar to leading US systems.
Handles Much Larger Data at Once
The new model, DeepSeek-V4, is designed to process extremely long inputs, supporting up to one million words in a single context. This allows it to analyse large documents and complex information more efficiently.
DeepSeek says the upgrade improves reasoning ability, knowledge handling, and overall accuracy.
Focus on Lower Cost and Efficiency
A key feature of the new model is its reduced computing and memory requirements. The company claims it delivers stronger performance while using fewer resources, making it more cost-efficient.
Two versions have been introduced, including a more powerful option and a lighter version designed for lower-cost use.
China’s high-tech industry sales rise 13.9 percent in 2025
Growing Competition in Global AI
According to Beyond Time News, the release reflects the growing AI race between China and the United States, where companies are competing to build faster and more efficient systems.
Experts say such developments could make advanced AI tools more accessible to businesses and developers by reducing costs.
Wider Use Across Industries
DeepSeek’s open-source approach has already helped its technology spread across different sectors, including business, healthcare, and public services.
With improved efficiency and lower costs, the new model is expected to support wider adoption of AI in practical applications.
