Part 5/8:
One of the more glaring issues surrounding GPT-4.5 is its pricing model. When comparing it to earlier iterations and its competitors, the sheer cost per million tokens for input and output is staggering—about 30 times more expensive than other working models like GPT-3.5 and significantly more than Claude 3.7 and others. This pricing brings forth a critical question of usability: who can afford to leverage this model at such a premium, especially when cheaper models yield similar results?
While some have praised its output for creative writing and structured tasks, if the costs associated with GPT-4.5 are not aligned with its performance, users may not find it feasible for general everyday tasks.