According to God of Prompt (@godofprompt), Chinese startup DeepSeek has developed an open source AI language model that matches the performance of OpenAI’s $60/million token GPT model but at a cost of just $0.55 per million tokens, representing a 98% reduction in cost. DeepSeek achieved this with a $6 million investment using restricted GPUs unavailable in China due to US export controls, while OpenAI spent $6 billion for similar outcomes. This breakthrough highlights the rapidly narrowing technological gap between Chinese and US AI firms, undermining OpenAI’s perceived technological moat and suggesting that high AI valuations may be driven more by marketing than by actual technological exclusivity. For the AI industry, DeepSeek’s open source approach and drastic cost reduction signal new business opportunities in global AI accessibility, enterprise adoption, and competitive pricing, while raising questions about the sustainability of current US AI business models (source: @godofprompt, Twitter, Nov 9, 2025).