Chinese AI Closes Gap: DeepSeek Challenges US Dominance

Chinese AI models are now lagging behind US rivals by months, not years. Artificial Analysis, an AI benchmarking platform, reports this gap has dramatically narrowed to just three to nine months. What's interesting is DeepSeek is leading this charge for China. Their R1 0528 model tied for the world's second-ranked lab overall. OpenAI's o3 model still holds the top spot for American models. Chinese large language models surged from 3% to 13% of global market share within a two-month window. However, American models still dominate raw usage, capturing about 93% of LLM site visits. Here's the thing: DeepSeek trained its V3 model for approximately $5.6 million. This challenges the idea that frontier AI requires billions in spending. The bottom line for listeners is that AI-related digital asset prices dipped after DeepSeek's competitive models emerged, suggesting continued volatility for tokens tied to AI developments.