What was actually draining Codex usage limits
- Source
- x.com
- Date
Update on rate limits in Codex. We do see that for some users the cache hit rate has been worse this week than the stable state the weeks before. This could explain that usage is draining somewhat faster for those users as hitting the cache consistently is an important component of being efficient. We are investigating and will have an update tomorrow.
Update on rate limits in Codex. We’ve found (a) some inefficiencies when using images in long sessions with multiple compactions (b) high p95+ usage for Computer History (c) a feature that was meant to generate conversation titles that was draining a bit more usage than intended. And we have a tiger team combing through everything and shipping fixes tomorrow. We also found a novel approach to drive efficiency up significantly that is completely unrelated and we will be working on next week. As part of some of the fixes tomorrow, we will also do a full reset of the usage for all paid subscriptions. See you then.
Good Sunday. Reset has been propagated to accounts and we landed some fixes to usage for things mentioned yesterday as issues we found. You should feel a positive difference. More to come tomorrow and will keep communicating. https://t.co/6mgUetJnhg
Context
The August 22–24, 2026 Codex investigation identified ways supporting features could consume OpenAI subscribers' usage allowances. The thread first raised reduced cache efficiency as a possible explanation, then reported problems involving images in long conversations, unusually heavy Computer History usage and conversation-title generation.
Its final update said paid accounts had received a usage reset and some fixes had landed. The sequence matters: the initial explanation was tentative, while the later posts identified specific inefficiencies. It supports an account of partial diagnosis and repair, without establishing a single cause for every report of rapid usage depletion.
Checking sign-in…
Loading comments…