DeepSeek to launch coding-focused AI model, the Information reports

Chinese AI startup DeepSeek is expected to launch its next-generation AI model V4, featuring strong coding capabilities, in mid-February, The Information reported on Friday citing people familiar with the matter.

Internal tests by DeepSeek employees suggested V4 could outperform rivals such as Anthropic's Claude and OpenAI's GPT series in coding tasks, the report said.

The latest V4 model has also made breakthroughs in handling and processing extremely long coding prompts, a potential advantage for developers working on complex software projects, the Information added.

Reuters could not immediately verify the report. DeepSeek did not immediately respond to a Reuters request for comment.

Hangzhou-based startup DeepSeek has emerged as a key player in China's push to build its own AI ecosystem and bolster the domestic chip sector. Its DeepSeek-V3 and DeepSeek-R1 models drew praise from Silicon Valley executives for their demonstrated capabilities.

Reuters previously reported that the Chinese artificial intelligence startup, which said in January it had built a low-cost rival to ChatGPT, has faced scrutiny in some countries over its security and privacy practices.