Arguments about whether $20-200 plans provide adequate value, complaints about hitting limits in minutes, comparisons of different plan tiers, and discussion of pricing psychology
The transition from enthusiast to frustrated critic is a recurring theme among users who grapple with opaque token limits that can drain a $20 session's value in mere minutes. This has fueled a contentious debate over whether entry-level tiers are being intentionally "nerfed" to push users toward $200 plans, yet even top-tier subscribers report feeling nickeled-and-dimed by new features being monetized outside their premium packages. While some defend the immense utility of low-cost AI assistance, a growing number of developers are migrating to competitors like Codex, citing a preference for predictable performance over the "anxiety-inducing" usage bars and perceived "enshittification" of current flagship models. Ultimately, the latest model updates have intensified these concerns, as increased reasoning capabilities paradoxically result in higher token consumption, leaving many to conclude that the era of unsustainable, cheap frontier AI is rapidly coming to an end.
56 comments tagged with this topic