The Great AI Price Collapse: How DeepSeek & Open Weights Broke the Proprietary LLM Monopoly
Model inference costs have plummeted over 90% in under a year. Here is a granular breakdown of how DeepSeek, open-weight architectures, and prompt caching shattered proprietary pricing power.