DeepSeek V4 Model Coming February With Coding Focus

DeepSeek will launch its V4 model in mid-February with enhanced coding capabilities that outperform Anthropic’s Claude and OpenAI’s GPT series. The Chinese startup’s new model excels at handling extremely long coding prompts.

Read more: How Pakistan’s AI Ambitions Can Drive Economic Growth

Chinese AI startup DeepSeek plans to launch its next-generation AI model V4 in mid-February, according to The Information. The Hangzhou-based company’s new model will feature enhanced coding capabilities, with internal tests suggesting it could outperform rivals like Anthropic’s Claude and OpenAI’s GPT series in coding tasks. DeepSeek employees testing the V4 model found it demonstrates particular strength in handling extremely long coding prompts, offering a significant advantage for developers working on complex software projects.

This development positions DeepSeek as a key player in China’s push to build its own AI ecosystem and strengthen the domestic chip sector. The company has already gained global attention with its cost-efficient DeepSeek-V3 series, which rivals GPT-5 in performance while maintaining lower operational costs. Reuters could not immediately verify the report, and DeepSeek has not yet responded to requests for comment regarding the February launch timeline.