CNBC’s Jim Cramer has expressed skepticism about DeepSeek, a Chinese AI startup that recently revealed an AI model it claims cost $6 million to develop. Cramer believes the company may not have disclosed the full story and suggested that its reported costs may be misleading.
A new report from SemiAnalysis, a semiconductor research firm, found that DeepSeek’s framing of its model development was “misleading” and that the actual cost could be as high as $500 million. Cramer agrees with this assessment, stating that it “may just be one more long knife aimed at Nvidia.”
Cramer’s comments come after investors were surprised by DeepSeek’s announcement, which suggested that Nvidia may not need to spend as much on its chips if DeepSeek’s model can outperform industry favorites like OpenAI. As a result, Nvidia shares plummeted, with the stock losing nearly $600 billion in one session.
However, Cramer remains skeptical about the potential impact of DeepSeek’s AI model, citing concerns that tech giants may not have invested as heavily in Nvidia without thorough due diligence. He also noted that DeepSeek received significant attention this week, which may be more a result of media coverage than a genuine threat to Nvidia’s sales.
Nvidia and DeepSeek declined to comment on the matter.
Source: https://www.cnbc.com/2025/01/31/deepseeks-threat-to-nvidia-may-not-be-so-serious-jim-cramer-says.html