A new analysis of 439 large language models (LLMs) from 63 different providers reveals widespread price arbitrage opportunities, suggesting that users may be overspending by a factor of 10 or more. The study found instances of identical LLMs offered for free on certain platforms while carrying substantial costs on others. In addition to price discrepancies, the research also identified variations in latency and performance, with cheaper models sometimes outperforming their more expensive counterparts. The findings underscore the critical need for users to carefully evaluate and compare pricing across various LLM providers to optimize both cost efficiency and performance. The original analysis was shared on Reddit and offers a deeper dive into these discrepancies.