OpenAI has reinstated a five-hour rolling usage limit on ChatGPT Plus Codex and Work environments, ending six weeks of unlimited access that began with the GPT-5.6 Sol rollout on July 13, 2026. The reversal follows two earlier quota resets in August and bug fixes for abuse-detection systems.
TL;DR
- OpenAI has brought back the five-hour usage cap for ChatGPT Plus Codex and Work after six weeks of unlimited access.
- The change comes after two earlier quota resets and fixes to abuse-detection systems that were draining user allowances.
- This development is significant for developers and users who have been taking advantage of the extended access.
What happened
OpenAI's decision to reinstate the five-hour rolling usage limit on August 25, 2026, marks the end of a six-week period of unlimited access for ChatGPT Plus users. This period began with the rollout of GPT-5.6 Sol on July 13, 2026.
The reversal follows two earlier quota resets in August and the implementation of bug fixes for abuse-detection systems. These systems were previously causing user allowances to deplete faster than intended.
Why it matters
For developers and users who have been utilizing the extended access, this change will impact their workflow and usage patterns. The five-hour cap means they will need to manage their usage more carefully.
This development highlights OpenAI's ongoing efforts to balance user access with the prevention of abuse. It also underscores the challenges of managing large-scale AI systems and ensuring fair usage among users.
Key facts
- OpenAI reinstated the five-hour rolling usage limit on August 25, 2026.
- The unlimited access period began on July 13, 2026, with the rollout of GPT-5.6 Sol.
- Two earlier quota resets occurred in August 2026.
- Bug fixes were implemented to address issues with abuse-detection systems.
- The five-hour cap applies to ChatGPT Plus Codex and Work environments.
- The change follows six weeks of unlimited access for ChatGPT Plus users.
Context
OpenAI's decision to reinstate the usage cap is part of a broader trend in the AI industry to manage resource allocation and prevent abuse. As AI models become more powerful and widely used, companies are grappling with how to ensure fair access while maintaining the integrity of their systems.
This development also reflects the ongoing evolution of AI models and the challenges of scaling them effectively. The rollout of GPT-5.6 Sol and the subsequent adjustments highlight the iterative nature of AI development and the need for continuous monitoring and improvement.
