OpenAI addresses unclear prompts regarding resource consumption for its high-computing mode, stating it has urgently reset Codex and ChatGPT Work quotas.

OpenAI Codex core member Tibo posted on X that the company previously let users enable high-compute mode too easily without fully explaining its impact on usage quotas. A large-scale revamp of the desktop app made common features like chat and projects hard to locate; overpromotional focus on ChatGPT Work led some Codex users to mistakenly believe the tool would be phased out. Additionally, the new version caused issues with some multi-agent workflows, plugin functions, and other user experiences. In response, OpenAI has rolled out its first batch of fixes: adjusting default settings and the model selector to prevent misuse of high-cost configurations, resolving multiple plugin submission issues, optimizing Codex’s presentation within the product, and fixing some desktop experience bugs. Over the past 24 hours, OpenAI has also been collecting user feedback, analyzing usage data, and communicating with users. While the market has shown strong interest in GPT 5.6 Sol, as well as ChatGPT Work’s web and mobile versions, OpenAI acknowledged multiple flaws in the recent update. To date, two top-ups for Codex and ChatGPT Work usage quotas have been completed. A larger-scale update is set to launch next week, during which chat and project features will return to the sidebar in a more familiar, customizable format. The platform will also enhance visibility of usage quotas and reset times, and further clarify the respective use cases of ChatGPT Work and Codex.