DeepSeek Permanently Slashes Flagship AI Model Price by 75%Memory Now Accounts for Nearly Two-Thirds of AI Chip CostsFirms Scramble to Rebrand Amid 'AI Washing' TrendOpenClaw Adds iMessage Tapback Approval ControlsDeepSeek Permanently Slashes Flagship AI Model Price by 75%

DeepSeek will permanently reduce the price of its flagship AI model by 75%, according to Bloomberg. The company previously offered temporary discounts to capture market share and undercut competitors. Making this price cut permanent signals DeepSeek’s long-term strategy to maintain a low-cost advantage in the AI inference market. This aggressive pricing move will likely squeeze profit margins across the entire AI API industry. Developers can now run and test high-concurrency AI applications at a fraction of the previous cost.

Memory now makes up nearly two-thirds of the total component costs in AI chips, according to a recent data analysis by Epoch AI. As large models demand rapidly increasing memory capacity and bandwidth, the cost of memory chips has significantly outpaced logic computing units. This shift indicates that the bottleneck for AI hardware cost control has moved from computing to storage. For chipmakers and data centers, driving down the cost of next-gen AI infrastructure will heavily depend on more affordable high-bandwidth memory solutions.

Companies are increasingly using PR campaigns to rebrand themselves as tech-focused businesses, a trend now known as ‘AI washing,’ The Guardian reports. As AI hype grows, marketing and PR firms are scrambling to pitch traditional companies by heavily relying on AI buzzwords and concepts. This marketing-first approach is creating significant noise and confusion in the market. For consumers and investors, distinguishing between genuine AI innovation and superficial marketing stunts is becoming increasingly difficult.