#DeepSeek’s radical architecture, particularly its cache-optimised models, is #disrupting the enterprise #AImarket. The company’s 75% price cut on its flagship #V4Pro model, which is 7x cheaper on inputs and 17x cheaper on outputs than comparable Western models, is forcing a reevaluation of the #tokencostcrisis. This #deflationarypressure, coupled with DeepSeek’s open-weight architecture, is leading to a bifurcation of the market. https://venturebeat.com/infrastructure/how-deepseeks-radical-architecture-is-shattering-silicon-valleys-token-moat?AIagents.at #AIagent #AI #ML #NLP #LLM #GenAI