800 Million non paying accounts for free inference, but theyre nickle and diming paid users with lessened Codex rate limits.

OpenAI is navigating a significant backlash from its core developer base after implementing stricter rate limits on its paid API services, particularly affecting Codex models used for code generation. The controversy stems from the perceived imbalance of offering robust, free inference to approximately 800 million ChatGPT users while simultaneously restricting the developers and businesses who pay for API access. Many professional users report hitting new, lower usage caps that disrupt their development workflows and application scaling, labeling the move a counterproductive 'nickel and diming' strategy.

This policy shift has ignited a wave of frustration on developer forums, with some threatening to migrate to competing platforms like Anthropic's Claude or open-source models. The core complaint is that OpenAI is prioritizing user growth for its free consumer product over supporting the paid ecosystem that builds upon its technology. This tension exposes the challenging economics of sustaining massive AI inference costs and raises questions about how OpenAI will balance its dual roles as a consumer-facing app company and an infrastructure provider for developers.