OpenAI said it would reimpose the 5-hour usage limit on Codex and ChatGPT Work for ChatGPT Plus subscribers starting August 25 CITE:E1. Thibault Sottiaux, OpenAI's engineering lead for Codex and ChatGPT, confirmed the change on X on the evening of August 24, stating the limit had been planned earlier but temporarily postponed CITE:E3. His post read: "Tomorrow we will bring back the 5h limit for Plus accounts across ChatGPT Work and Codex. I had mentioned this a while ago, but then postponed it." CITE:E4 Sottiaux's August 24 post specified that the 5-hour rolling limit would formally take effect August 25 for Plus users on both products CITE:E8.
Plus users return to a "5-hour plus weekly" dual-limit structure after a period governed only by the weekly cap CITE:E2. The 5-hour limit caps how much Codex and ChatGPT Work capacity a Plus user can consume within any rolling 5-hour window, while a separate weekly total governs overall consumption — the two operate as distinct layers, with the short-cycle limit controlling usage intensity and the weekly limit managing cumulative volume CITE:E11.
Usage on Codex and ChatGPT Work surged within 48 hours of GPT-5.6 Sol going live, and because the two products draw from a single shared quota pool, the spike in one directly squeezed capacity available to the other CITE:E10. Sottiaux said reinstating the limit was meant to "smooth out compute load" while preserving the generosity of the Plus plan's weekly quota CITE:E12.
OpenAI suspended the 5-hour rolling cap for Plus, Pro, and Business plans on July 12, resetting all users' usage counters at the same time CITE:E9. The suspension coincided with OpenAI marking a milestone of 1 million combined active users across its two merged platforms, during which it temporarily lifted the 5-hour limit and, on several occasions, reset users' weekly quotas early CITE:E7. That pause ended on August 25, when the 5-hour limit returned for Plus accounts CITE:E1.
The August 25 restoration applies to ChatGPT Plus accounts, while Pro, Enterprise, and Edu plans are not affected for now CITE:E5CITE:E6.
| Tier | Monthly Price | 5-Hour Limit Status | Key Date |
|---|---|---|---|
| ChatGPT Plus | — | Restored (5-hour + weekly) | Aug 25, 2026 |
| Pro | $100 | Exempt for the coming months | Launched Apr 2026 |
| Pro | $200 | Exempt for the coming months | — |
| Enterprise / Edu | — | No change (point-based quota) | — |
Both the $100 and $200 Pro tiers will remain free of the 5-hour limit for the coming months CITE:E5CITE:E13. Sottiaux did not mention any changes for Enterprise and Edu accounts, which continue to run on a separate point-based quota system CITE:E14. OpenAI launched the $100-per-month Pro plan in April, aimed at developers CITE:E15.
Taken together, the sequence shows OpenAI toggling the 5-hour cap in step with real-time load rather than removing it permanently: it lifted the limit on July 12 to mark the 1-million-active-user milestone CITE:E7CITE:E9, then reinstated it on August 25 after usage spiked within 48 hours of GPT-5.6 Sol's launch strained the shared Codex-Work quota pool CITE:E10. The restored cap falls on Plus accounts while the $100 and $200 Pro tiers — the latter aimed at developers since its April debut — stay exempt for now, alongside Enterprise and Edu's unchanged point-based system CITE:E5CITE:E13CITE:E14CITE:E15.
OpenAI said it would reimpose the 5-hour usage limit on Codex and ChatGPT Work for ChatGPT Plus subscribers starting August 25 CITE:E1.
Plus users return to a "5-hour plus weekly" dual-limit structure after a period governed only by the weekly cap CITE:E2.
Usage on Codex and ChatGPT Work surged within 48 hours of GPT-5.
OpenAI suspended the 5-hour rolling cap for Plus, Pro, and Business plans on July 12, resetting all users' usage counters at the same time CITE:E9.
The real constraint here isn't the 5-hour cap itself but the shared quota pool between Codex and ChatGPT Work — a spike in one product directly squeezes capacity in the other, which is exactly what happened when GPT-5.6 Sol's launch drove a usage surge within 48 hours. Reinstating the rolling limit is OpenAI's lever for smoothing that shared load rather than provisioning the two products separately. The more telling signal is which tiers stay exempt: the $100 and $200 Pro plans keep running without the cap while Plus absorbs it again, a gap that has held since Pro's April debut and now doubles as the pressure valve for developers who hit Codex limits. Worth watching next: whether OpenAI eventually decouples the Codex and Work quota pools outright, rather than continuing to manage shared-capacity spikes through tier-based exemptions.
Google is rolling out Motion Assist, a Play Services feature for Android 17 that places moving dots along the screen edges to track a vehicle's motion using the phone's accelerometer and gyroscope, aiming to reduce or eliminate motion sickness — two years after Apple shipped a similar Motion Cues feature in iOS 18.
More than 100 million Americans lack a primary doctor. Atlantic Health and K Health's AI platform, synced to Epic, undercuts $12k-$24k concierge fees.
D-Wave, Rigetti, Quantum Computing Inc. and IonQ all gained between roughly 8% and 11.5% on August 21, lifted by quarterly results showing steep revenue growth from near-zero prior-year bases, a fresh $35 BMO Capital Markets price target on D-Wave, and NVIDIA's expanding quantum-GPU integration stack.