DeepSeek releases preview of V4 large language model
Chinese startup DeepSeek has launched a preview of its long-awaited Version 4 model, claiming significant cost reductions as it continues to compete with major U.S. rivals
Chinese artificial intelligence startup DeepSeek has officially released a preview version of its highly anticipated Version 4 (V4) large language model. The launch, which took place on Friday, represents the latest development from the firm as the international competition for AI dominance continues to accelerate between Chinese and American developers. This preview allows users to experience the capabilities of the new system before a full-scale rollout.[1][2]
According to the company, the new V4 model features drastically reduced costs for users and developers compared to earlier versions. This focus on affordability follows DeepSeek's previous success more than a year ago, when it surprised the global tech industry by releasing a low-cost reasoning model that demonstrated capabilities comparable to those of leading U.S.-based artificial intelligence systems. The startup has consistently emphasized its ability to deliver high-performance AI at a fraction of the price of its competitors.[1]
The release of the V4 preview is seen as a strategic move to maintain the startup's momentum in the rapidly evolving AI landscape. By prioritizing cost-efficiency alongside performance, DeepSeek aims to provide a competitive alternative to the resource-intensive models currently dominating the market. This latest release underscores the intensifying race among global tech firms to produce more efficient and powerful large language models for a wide range of applications.[1][2]
