--- title: OpenAI preps paid usage resets while Codex limits... description: Codex has surpassed 20 million active users, prompting OpenAI to grant a free banked usage reset to Codex and ChatGPT Work users as a milestone thank-you.... canonical: https://daily.dev/posts/openai-preps-paid-usage-resets-while-codex-limits-mysteriously-drain-faster-xm8kss93l twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: OpenAI preps paid usage resets while Codex limits mysteriously drain faster | daily.dev og:description: Codex has surpassed 20 million active users, prompting OpenAI to grant a free banked usage reset to Codex and ChatGPT Work users as a milestone thank-you.... og:url: https://daily.dev/posts/openai-preps-paid-usage-resets-while-codex-limits-mysteriously-drain-faster-xm8kss93l og:image: https://api.daily.dev/og/posts/xM8kss93L.png og:image:alt: OpenAI preps paid usage resets while Codex limits mysteriously drain faster og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # OpenAI preps paid usage resets while Codex limits mysteriously drain faster **[Trends](https://daily.dev/sources/trends)** · 3 min read · 0 upvotes · 0 comments ## Summary Codex has surpassed 20 million active users, prompting OpenAI to grant a free banked usage reset to Codex and ChatGPT Work users as a milestone thank-you. Meanwhile, reports from a user tracking OpenAI's product internals suggest paid usage resets are being tested, letting users pay (one instance spotted at $80) to skip waiting out their 5-hour or weekly quota window. This comes as users report that usage limits are draining faster than expected, with OpenAI's Thomas Sottiaux offering a carefully worded response to the complaints without fully addressing the cause. ## Content OpenAI's Codex has a problem that's harder to fix than a simple rate limit cut: nobody, including OpenAI, can fully explain where the tokens are going. The complaints are specific enough to take seriously. A Plus account went from 99% to 0% in hours. A Pro $200 plan reset at 6:40 AM and was empty by 3:25 PM. A 20x user burned 6% in 90 minutes doing nothing heavy. One user on Hacker News let Sol *think* for 10 minutes and drained the company's entire weekly quota. Previously hitting the cap required roughly 2.6 billion tokens; users are now getting locked out at around 400 million. OpenAI's own PM (Tibo) has acknowledged cache hit rates degraded this week, which explains some of the faster drain. But cache misses only matter when requests are being made. Multiple users have documented quota dropping with zero active sessions, verified against both the backend usage API and local session logs. One forensic report caught quota climbing from 79% to 100% used over 29 hours with no new session files whatsoever. That's not a cache problem. The background consumption angle is real and underreported. The desktop app generates continuous usage history while sitting idle. Chronicle summaries fire every 10 minutes. Memory consolidation requests (~143K tokens each) run after quota resets. Auto-suggestions were double-running and retrying after failures, all billed to your quota. OpenAI has confirmed and fixed some of these before - they did a full rollback and reset in late June after a similar incident. They've now done it again, issuing a banked reset to all paid users alongside fixes for image inefficiencies in long sessions, high Computer History usage, and a title-generation feature that was quietly eating quota. The resets are becoming a pattern, and that's the actual problem. As one open letter from a paying user put it: "resets are not a technical explanation, and they are not a durable fix. If the meter is miscalibrated, pressing the reset button simply gives customers another opportunity to watch it drain." What's missing is an itemized ledger. Users can't see per-task costs, reasoning token consumption, tool-call charges, or what background processes billed them. The pricing page describes Pro as "5x" or "20x" Plus usage without stating what Plus usage actually is in any fixed unit. That opacity is now colliding with a product that runs agents autonomously, shares resource pools across Work and Excel, and has a paid credit top-up flow ready to catch users the moment their included allowance runs out. One commenter put it plainly: "You pay the same price at the pump. Someone else gets 100 liters. You get 25. And the company selling you the fuel also controls the pump, the meter, the definition of '100%,' and when the meter resets." OpenAI is bringing back the 5-hour limit for Plus users, which at least prevents a single session from consuming a week's budget. Pro users get a reprieve for now. But until there's a transaction-level usage ledger, the community is essentially being asked to trust a meter that has been wrong multiple times in the past month. ## Questions this post answers ### Is OpenAI planning to let Codex users pay to reset their usage limits early? Signs point to yes, though nothing has been officially announced. Reports indicate Codex's interface is already tracking reset type, expiration, and redemption data, with some checkout flows spotted offering an immediate full usage reset for around $80. The model would let users pay to skip the remainder of their 5-hour or weekly quota window instead of waiting it out. _Track how Codex pricing and quota changes evolve before committing budget to a coding assistant plan on daily.dev._ ### Why are Codex usage limits draining faster than before? Users have reported that their usage limits are being consumed more quickly than expected, and OpenAI's Thomas Sottiaux addressed the complaints without giving a clear technical explanation for the change. The response came alongside a free banked usage reset given to Codex and ChatGPT Work users to mark Codex crossing 20 million active users. _Developers relying on Codex quotas can follow usage-limit changes like this on daily.dev before they get caught off guard._ ## Community take How the wider developer community reacted, aggregated from 1 discussion and 7 comments across x (as of 2026-08-27). **TL;DR:** Reactions are mostly low-signal replies, but a few raise practical worries about whether resets cover the API tier and whether rising usage explains the constant limit-hitting. **Sentiment:** 10% positive · 30% mixed · 60% skeptical **The case for** - A reset being issued was seen by some as at least a temporary relief even if not a full fix. **The pushback** - Unclear whether the reset applies to API usage or just the web interface, leaving batch jobs at risk. - Skepticism that a growing user base is straining limits, suggesting the drain issue may just be a symptom of scale. **By community** - x (mixed): Mostly brief, low-substance replies with a couple of practical concerns about reset scope and rising usage. **Open questions** - Does the usage reset apply to API-tier usage or only the web interface? **Highlights** > @testingcatalog Does this reset apply to the API tier too, or just the web interface? Need to know if my batch jobs are safe to restart. > — [stevencheng on x](https://x.com/stevencheng/status/2093044900648624398) > @testingcatalog 2m weekly active users on codex. no wonder the limits keep getting hit > — [victoria\_neiman on x](https://x.com/victoria_neiman/status/2093048226450419762) > @testingcatalog be honest, nobody feels reseted, they feel relieved the queue moved > — [wearevalidating on x](https://x.com/wearevalidating/status/2093018714513793285) **Source threads** - [x](https://x.com/testingcatalog/status/2093015106212835375) · 1 points · 7 comments --- Tags: [#openai](https://daily.dev/tags/openai), [#ai-coding](https://daily.dev/tags/ai-coding), [#openai-codex](https://daily.dev/tags/openai-codex) [View this post on daily.dev](https://daily.dev/posts/openai-preps-paid-usage-resets-while-codex-limits-mysteriously-drain-faster-xm8kss93l) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"OpenAI preps paid usage resets while Codex limits mysteriously drain faster","url":"https://daily.dev/posts/openai-preps-paid-usage-resets-while-codex-limits-mysteriously-drain-faster-xm8kss93l","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/openai-preps-paid-usage-resets-while-codex-limits-mysteriously-drain-faster-xm8kss93l"},"datePublished":"2026-08-21T11:49:22.337Z","dateModified":"2026-08-27T19:39:21.177Z","description":"Codex has surpassed 20 million active users, prompting OpenAI to grant a free banked usage reset to Codex and ChatGPT Work users as a milestone thank-you....","image":"https://pbs.twimg.com/media/HQPmdaTWsAA7Xjb.jpg","thumbnailUrl":"https://pbs.twimg.com/media/HQPmdaTWsAA7Xjb.jpg","isAccessibleForFree":true,"articleSection":"Trends","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Trends","logo":"https://media.daily.dev/image/upload/s--ZfSp3asX--/f_auto,q_auto/v1780996004/logos/trends?_a=BAMAMiWQ0","url":"https://daily.dev/sources/trends"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/openai-preps-paid-usage-resets-while-codex-limits-mysteriously-drain-faster-xm8kss93l","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"openai,ai-coding,openai-codex","timeRequired":"PT3M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Trends","item":"https://daily.dev/sources/trends"},{"@type":"ListItem","position":3,"name":"OpenAI preps paid usage resets while Codex limits mysteriously drain faster"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/openai-preps-paid-usage-resets-while-codex-limits-mysteriously-drain-faster-xm8kss93l#faq","mainEntity":[{"@type":"Question","name":"Is OpenAI planning to let Codex users pay to reset their usage limits early?","acceptedAnswer":{"@type":"Answer","text":"Signs point to yes, though nothing has been officially announced. Reports indicate Codex's interface is already tracking reset type, expiration, and redemption data, with some checkout flows spotted offering an immediate full usage reset for around $80. The model would let users pay to skip the remainder of their 5-hour or weekly quota window instead of waiting it out. Track how Codex pricing and quota changes evolve before committing budget to a coding assistant plan on daily.dev."}},{"@type":"Question","name":"Why are Codex usage limits draining faster than before?","acceptedAnswer":{"@type":"Answer","text":"Users have reported that their usage limits are being consumed more quickly than expected, and OpenAI's Thomas Sottiaux addressed the complaints without giving a clear technical explanation for the change. The response came alongside a free banked usage reset given to Codex and ChatGPT Work users to mark Codex crossing 20 million active users. Developers relying on Codex quotas can follow usage-limit changes like this on daily.dev before they get caught off guard."}}]} ```