OpenAI has announced the reinstatement of a five-hour usage limit for specific subscription tiers, including ChatGPT Work and Codex Platform Plus accounts, effective August 26, 2026. Tibo Sottiaux, the head of OpenAI Codex, confirmed the move via the X platform, noting that the restriction was previously postponed but is now necessary to optimize infrastructure performance. The decision aims to balance the computational demands of advanced artificial intelligence models with the need for stable service delivery across the global user base.
Optimizing Computational Load and User Experience
The resumption of the five-hour window is primarily driven by the need to smooth computational load across OpenAI's distributed server networks. By implementing these intervals, the platform can maintain more generous weekly usage quotas without risking system instability during peak hours. Computational resources for LLMs (Large Language Models) remain a significant overhead for providers as demand for real-time AI processing grows.
Data indicated that Plus users, who are often characterized as relatively light or new users, frequently exhausted their weekly quotas inadvertently. This led to a degraded user experience where individuals were unable to access services for several days. The five-hour limit acts as a pacing mechanism to prevent accidental depletion of resources.
- Resumption Date: August 26, 2026.
- Affected Tiers: ChatGPT Work and Codex Platform Plus.
- Primary Goal: Load balancing and preservation of weekly quotas.
- Exemptions: Pro subscriptions and high-tier enterprise accounts.
Impact on Pro Subscriptions and Developers
While the Plus and Work tiers face these new restrictions, OpenAI clarified that Pro subscriptions will remain unaffected by this specific limit for the foreseeable future. This distinction is crucial for developers and blockchain engineers who utilize the Codex engine for writing smart contracts or auditing decentralized applications (dApps).
"The 5-hour limit helps to smooth computational load and maintain generous weekly usage quotas", stated Tibo Sottiaux, emphasizing that Pro tiers will not see these changes in the coming months.
This strategic separation ensures that power users and professional developers operating on high-frequency schedules have the necessary uptime and throughput to continue their operations without interruption, while more casual users are guided toward sustainable consumption patterns.
The reintroduction of these limits highlights the ongoing challenges AI firms face in scaling their hardware capabilities to meet surging global demand. As the integration of AI within the cryptocurrency and tech sectors deepens—ranging from automated trading bots to AI-driven security analysis—the management of limited GPU resources becomes a critical factor for service providers. Users on the affected tiers are advised to monitor their usage statistics starting tomorrow to ensure continuous access to the platform's features.
Frequently Asked Questions
Quick answers to the most common questions about this topic.
For educational purposes only. Nothing here constitutes financial or investment advice. Crypto markets are highly volatile — always DYOR before making any decisions.
Fact-checked per our editorial policy. We maintain strict independence from advertisers. Spot an error? Let us know.