Chinese company DeepSeek is previewing a new AI model after a year of shaking up American rivals

Chinese AI company DeepSeek released a preview of its upcoming next-generation AI model V4 on Friday, Saying That the open source model can compete with leading closed source systems from US competitors including Anthropic, Google and OpenAI.

DeepSeek says the V4 represents a significant improvement over previous models, especially in programming, a capability that has become key for AI customers and has helped drive the success of tools like ChatGPT script and Claude Code. The release is also a milestone for China’s chip industry, with DeepSeek Explicitly highlight Compatibility with local Huawei technology.

The release comes a year after DeepSeek The artificial intelligence industry in the United States has been shaken With the R1, a model it claimed was trained at a fraction of the cost of leading American systems. DeepSeek did not disclose V4 training costs or the hardware it was trained on. US officials accused the company Using banned Nvidia chips And the anthropic Claims DeepSeek misused Claude to improve its products.