The report said internal testing by DeepSeek employees indicates that the V4 model could outperform rival systems from Anthropic and OpenAI, including Claude and the GPT series, particularly in coding-related tasks.
V4 is also said to have achieved significant improvements in handling and processing extremely long coding prompts, a capability that could give it an edge among developers working on large and complex software projects.
Reuters was unable to independently verify the report, and DeepSeek did not immediately respond to a request for comment.
Based in Hangzhou, DeepSeek has rapidly emerged as a prominent player in China’s efforts to develop a self-reliant artificial intelligence ecosystem and strengthen its domestic semiconductor industry. The startup attracted global attention after several Silicon Valley executives publicly praised its earlier DeepSeek-V3 and DeepSeek-R1 models.
In January, DeepSeek said it had developed a low-cost rival to OpenAI’s ChatGPT, further intensifying interest in the company’s technological progress. However, the firm has also faced scrutiny in some countries over concerns related to data security and privacy practices, according to previous Reuters reporting.
If confirmed, the launch of V4 would mark another major step in DeepSeek’s challenge to leading Western AI developers, particularly in the highly competitive area of coding and developer-focused AI tools.