Replies dated this day in the collected sample: 91. Collected since 5 Oct 2026: 2305.
X replies are found by search and each one is checked against its public post on X. Last read: 26 Oct 2026, 23:46 UTC. Not every reply is read.
- A problem
Fix the fucking usage. its so much worse than last week...
@DanoManoNFT on X ↗ · 01:10 UTC
- A problem
Day 16/ we have nerfed everyone’s usage and on the grandathered pro 20x you get 7 hours of sol 6.1 usage before usage is depleted 🤣🤣🤣
@Sacredcorey on X ↗ · 01:10 UTC
- A problem
How... in 5 hours i lost all my usage to 6.1 sol... 100$ plan... what's happening ?
@anonymous10552 on X ↗ · 01:13 UTC
- A problem
That is a good thing. However, why does the Codex app lose its TTL (time-to-live) in just 30 minutes? Managing it is so difficult that many users end up shifting their usage to Claude instead.
@isajunsei on X ↗ · 01:17 UTC
- A problem
Why does the cache still break randomly in a single-threaded conversation? It literally still happens all the time.
@Acehan_ on X ↗ · 01:19 UTC
- A win
Ok now THIS is a seriously good improvement. Long running tasks that pause because I'm AFK shouldn't chew through usage now. Awesome 😎
@bobbostwick on X ↗ · 01:21 UTC
- Mixed
Thanks, I suspected the most drained usage is because of repeated missed cache hit rate.
@nathan_tulu on X ↗ · 01:22 UTC
- A problem
“Selected model is at capacity. Please try a different model” please stop reset, before solve this issue!
@cooljdan on X ↗ · 01:24 UTC
- A problem
Selected model is at capacity. Please try a different model. 上周还是正常的,本周开始出现问题。
@ghost_cxl on X ↗ · 01:25 UTC
- A problem
If cache is guaranteed when resuming Codex sessions, why does quota usage spike after a weekly reset? For the same Sol 6.1 High task, 425M tokens used approximately 23% of my weekly quota before the reset. However, immediately after the reset, 93M tokens used 11%, while another…
@SerefC on X ↗ · 01:27 UTC
- A problem
at capacity???again???
@yichuanapia on X ↗ · 01:28 UTC
- A problem
I've been complaining about Codex quotas for months. Running 3 Pro $200 accounts, back when that was the top plan, before the $500 tier, was never normal. Resets help, but they mask the problem. Tibo reset me <24h ago, one project only, and I'm already at 10% left with 6 days to…
@je4ndev on X ↗ · 01:28 UTC
- A problem
If this has already rolled out, it’s honestly terrible—it doesn’t seem to help at all. I’m a $200 Pro user, and one 20-minute Codex task with Astra Ultra used 15% of my weekly limit. If rollout is ongoing, please expedite it and reset all users’ weekly limits as compensation
@A_Damn_Love_520 on X ↗ · 01:29 UTC
- A problem
I didn’t see reset
@Afar_Ming on X ↗ · 01:29 UTC
- A problem
Selected model is at capacity is all i’m seeing 😢
@Trebor46393323 on X ↗ · 01:29 UTC
- A win
persistent agents are where this pays. my eval loops hold context for hours. resuming one used to mean rebuying the context i already built. 12h cache turns the afternoon rerun into a cheap sequel.
@lloydjf7 on X ↗ · 01:30 UTC
- A problem
I encountered this on the Windows ChatGPT/Codex desktop app and recovered my projects successfully. In my case, the project folders and database records still existed, but the local configuration was damaged. The logs showed: [global-state] Failed to load global state…
@EricStoneChina on X ↗ · 01:31 UTC
- A problem
honestly with these resets, its hard to plan on what to do with codex (PRO acc.). And I feel like not liking it. the resets came so suddenly when my usage was still at 95% (like 3-4 days until wkly reset) and extends my wait time for 7 days.
@ArnelCatangay on X ↗ · 01:39 UTC
- A problem
I have 18% weekly usage left on the $100 plan and I haven't used codex at all. These reset are a band aid on a gaping wound. Complete trash.
@PG1235620 on X ↗ · 01:51 UTC
- A problem
If you don't fix this 'Selected model is at capacity. Please try a different model' bullshit, then Codex is officially cooked. I haven’t touched it for ten days, came back today, and I'm still getting hit with the exact same error. Just absolutely brilliant.
@piyuyxttdfv3wh on X ↗ · 02:00 UTC
- A problem
Can you please look into an issue that I've been experiencing? I have been randomly losing access to my chats. I lost a whole project folder with 22 chats in it and now while I was working in a new chat that I created today, I lost access to that one too. Please help.
@NeroDylan on X ↗ · 02:25 UTC
- Mixed
i have a banked reset expire on 23th, i believe a lot of people have as well
@seanbne on X ↗ · 02:55 UTC
- A problem
Selected model is at capacity. Please try a different model. why??????
@rookiebaby188 on X ↗ · 03:05 UTC
- A problem
My situation is even worse. I’m on Plus, and Codex has been completely unusable for 3 days. Every model shows “Selected model is at capacity” across all platforms. My weekly usage is only 2%. I can’t even send “Hello” in a new session.
@Fsshfg on X ↗ · 03:25 UTC
- A problem
We keep getting burned by dynamic tool schemas busting our cache mid-run. Curious how strict the prefix matching is.
@techwithjoe_io on X ↗ · 03:46 UTC
- A problem
I just burned through 2x 200 weekly plans since your last reset. How can I see what chats caused the most token usage/hr?
@GlobalAITeacher on X ↗ · 03:57 UTC
- A problem
plz can you check this? Codex got very very slow. I use 6.1 sol and subagents, when start the thread is quite fast but its speed dramatically drops after around 30min executions
@IMEditor_ai on X ↗ · 04:23 UTC
- A problem
Tibo, it looks to me like my account wasn't included in the reset. Could you check it...?
@vismotive on X ↗ · 04:33 UTC
- A problem
Since yesterday’s “reset,” every model has displayed the same message: “The selected model has reached its capacity limit.” So, what is the point of this so-called “reset”? While it ostensibly refreshes the quota, in reality, it further restricts actual usage.
@HHH7vd on X ↗ · 04:41 UTC
- A problem
Same here. I pay $200/month. Have barely scratched my monthly usage limit but this has become unbearable. A dozen of 429 errors today for no reason at all.@OpenAI help pls
@dalijet on X ↗ · 04:47 UTC
- A problem
今天的额度格外不耐用,我已经蹬完了
@changguangyu on X ↗ · 04:57 UTC
- Mixed
Day16 reset Day17 reset Day18 reset …
@rerey0adel on X ↗ · 05:30 UTC
- A problem
I used up 30% usage in 10hr of coding with GPT 6.1 sol @xhigh
@pavankushnure on X ↗ · 05:38 UTC
- A problem
on a 100 plan. drained 99% on normal usage in less than half a day!? and i have not received any emails
@kingwaihenry on X ↗ · 06:15 UTC
- A problem
我是pro500 用户,最近经常 selected model is at capacity please try a different model
@luckylucas1976 on X ↗ · 06:17 UTC
- A problem
The usage limit needs to be adjusted; on Claude, it lasts almost a whole week, but here it barely lasts 2 days on the same plan.
@honziik_90 on X ↗ · 07:24 UTC
- A problem
After the update, the model is full again. Can't this problem be completely solved?
@RhlTrPE8jU2230 on X ↗ · 07:25 UTC
- A problem
■ Selected model is at capacity. Please try a different model. Still doing this, 2 nights in a row. I prefer banked resets.
@Rasterdeck25086 on X ↗ · 07:27 UTC
- A problem
I’ve encountered this issue as well. it only started happening today. Was there a release today that might have caused this?
@heart_ers on X ↗ · 07:32 UTC
- A problem
我从昨天重置到现在一共用了2%,根本没法用
@LeeLantao on X ↗ · 07:39 UTC
- A problem
Reset pls. 2 account resets all gone in a day?
@ChaoWangjer6 on X ↗ · 08:04 UTC
- A problem
所选模型已达容量上限。请尝试其他模型 一样的 真恶心
@youmila1688 on X ↗ · 08:17 UTC
- A problem
I think my quota is being used up too quickly, even faster than the past couple of days, and I also got this error: stream disconnected before completion: Our servers are currently overloaded. Please try again later.
@iuxue262905 on X ↗ · 08:28 UTC
- A problem
Tibo, how many times do Plus users have to tell you this? Codex keeps slamming the door in the middle of real work while our weekly quota sits unused. We paid to get shit done, not to babysit a fucking countdown. Why are you still defending this 5-hour nonsense?
@noonegjm on X ↗ · 08:45 UTC
- A problem
Hello, tibo, I seriously suspect that there is a serious shrinkage problem for the usage quota of plus users. Using gpt6-astra high to work for 5 hours, the 5-hour limit is only 15 minutes, and a sub-task is over before it is over.
@AliTangSanJiao on X ↗ · 08:47 UTC
- A win
I received my credits
@MartinMachava1 on X ↗ · 09:16 UTC
- A problem
Hello, my ChatGPT Work mode quota is being consumed abnormally fast. I suspect the usage counter has a metering error. Even new Work conversations cannot allocate sandbox after quota ran out. Please check my usage logs and help reset the Work usage limit
@miener7 on X ↗ · 09:51 UTC
- A problem
I burned through my $200 so quickly. why?
@Lilt558 on X ↗ · 09:59 UTC
- A win
thank you @sama, my usage limit finally usable!
@Joyortoy7 on X ↗ · 10:17 UTC
- An idea
can you, ADD: Bulk delete / archive chats; FIX: Deleting chats - it is very slow, and often after deleting, chats stay there inside the App on Mac but when click on them, they just cannot be loaded because being deleted.
@nicrames on X ↗ · 11:00 UTC
- A problem
“Guaranteed to hit the cache” isn’t what my logs show. Codex Desktop 0.163.0-alpha.10.2, logged in, today: • 37 min idle → 89% cached • 56 min idle → 0 of 258k cached • <1 min gap → 0 of 315k cached The last can’t be TTL. Fully rolled out yet? Logs available
@Shermanaetor on X ↗ · 11:01 UTC
- A problem
能不能快点解决429的问题。。。。一直重置有什么用
@Faunavaloris on X ↗ · 11:24 UTC
- A problem
Suddenly codex shows "exceeded retry limit, last status: 429 Too Many Requests, request id: c761f1f0-a60b-4a66-a269-2ccfba718957", what happeded???
@HongliangSu on X ↗ · 12:32 UTC
- A problem
I am literally going through the 5 hour work usage limit in just a few minutes! It is getting extremely frustrating, I cannot proceed with work like this, get a tiny bit done then wait for 5 hours & so on, at least send a reset.
@rashidkhan on X ↗ · 13:44 UTC
- A problem
Why? I didn’t get the reset, this image was just captured.
@GEric2m10d on X ↗ · 13:58 UTC
- A problem
I redeemed a Codex reset credit on October 20, only for OpenAI to issue another automatic reset shortly afterward. My remaining restored usage was effectively wasted before I could fully use it. This is extremely frustrating and unfair to paying subscribers
@zihjianhong on X ↗ · 15:07 UTC
- A problem
Hey @tibo, Codex GPT-6 Astra keeps failing remote compaction with “stream disconnected / error decoding response body.” Reproduced on Desktop + CLI 0.163.0. Fresh chats work, but my original and migrated sessions both failed. Any known issue or fix?
@LoVe1eS3 on X ↗ · 15:29 UTC
- A problem
Was it a bankable reset, if so I didn't get it :(
@mfucek_ on X ↗ · 16:03 UTC
- A problem
I’m using 7 Codex accounts and this is what I get? I only switched to Codex 3 months ago, but I’m already thinking about going back to Claude. At least Claude doesn't have this stupid issue. I pay monthly just to get these error messages while my task is still pending?
@d_huy99 on X ↗ · 16:10 UTC
- A problem
Selected model is at capacity. Please try a different model. I kept getting this
@Sadanpan3 on X ↗ · 16:17 UTC
- A problem
I was timing my usage around the weekly reset, so the reset kind of threw me off 😅 My Pro quota is at 24%, and I thought the next reset was close, which made sense. Then I checked and it says 6 days left, because Tibo reset it on the 20th. I honestly don't get how I burned…
@rizon_cn on X ↗ · 16:41 UTC
- A problem
I don’t see the reset in my history tab on Codex for Oct 20. Can anyone explain why is that the case?
@dakshkhetan on X ↗ · 17:15 UTC
- A problem
Upgraded from $200 to $500 mid-task last Friday, expecting more than double the usage. Nope! Blew through the new limits that same day, doing less than I used to. Somehow it feels worse. I really don’t want to crawl back to Claude, but I might have no choice...
@TraderNomadX on X ↗ · 18:14 UTC
- A problem
I'm in token hell / token triage mode now more than ever...after I upgraded to $500. I mostly shifted back to Claude and just have subagent requests going to Codex agents...and I can still watch the Codex usage meter decrement in real time. Warning: the upgrade doesn't help!…
@TraderNomadX on X ↗ · 18:20 UTC
- A problem
One request used up my entire 5-hour GPT-6.1 Sol limit in 3 minutes, and Astra's in just 1 minute! What changed? Why am I paying if I can only send one request before waiting 5 hours? This is ridiculous and makes the service nearly unusable. I may really have to switch to Claude.
@PonceDeLeonnn on X ↗ · 18:29 UTC
- A problem
You misunderstood what I meant: the 5h limit on Codex is draining like crazy, compared to Claude. I should've been more clear.
@nomadhedge on X ↗ · 18:47 UTC
- A problem
my usage today drop from 50% to 0% in no time, with a pro 500 and I was using with dots and 3 agent in parallel with sol xhigh... something is happening above usage limits, and I cannot see the usage today to see where went the tokens.
@JCM_Chile on X ↗ · 18:49 UTC
- A problem
All models at capacity now.
@sebasmediaprod on X ↗ · 19:09 UTC
- A problem
Ran through the Pro 20x weekly limit in less than 36hrs while using Sol 6.1 on High 🤷♂️ Seems like it doesn’t make a difference using Astra or Sol, both are consuming tokens too fast which results in weekly limits being used in a 1-2days.
@Nebula_Matrix on X ↗ · 19:25 UTC
- A problem
Hey Tibo, what’s going on with Codex? All I’m getting is “Bad Request”
@FelixWmrv0 on X ↗ · 19:37 UTC
- A problem
My usage is still cooked
@neyis_g on X ↗ · 20:04 UTC
- A problem
haven't received my reset
@davidarduch on X ↗ · 20:18 UTC
- A problem
Since I don't have a mobile phone number that can be used to receive the verification code, I can't continue to use the client. Now it's been a day since I sent the email, and I haven't received a reply yet. I don't know when I can use codex?
@wzd166196 on X ↗ · 20:22 UTC
- A problem
Unfortunately, I don't think it's ended yet. It's hard to believe I've used a week's worth of 500 in a single day.
@gregoryt9009 on X ↗ · 21:24 UTC
- A problem
I used astra light for 2 hrs and my 80% usage is gone today too, finally better cancel codex i have renewal in 3 days from loving codex to finally giving up 🆙 best wishes tibo
@riwajrise on X ↗ · 21:24 UTC
- A problem
Hey Tibo. I experienced two crashes in Codex today, and my laptop shut itself down. Problems :/ @thsottiaux
@vibestressing on X ↗ · 21:35 UTC
- A problem
Hi Tibo, I’m on the 200 plan, but I used up the ASTRO model in only 6 hours. I strongly suspect that I wasn’t actually given the 200 plan.
@MgqZqsjvjBMUb3g on X ↗ · 21:45 UTC
- A problem
im disappointed. nothing for me as a normal user changed. one High Astra Coding Instance burns through my weekly limit within less then 10 hours.
@LeoGamrs on X ↗ · 21:56 UTC
- A problem
Max user here, drained weekly usage in one day in 3 sessions. Not normal.
@ThoughtsOfDevo on X ↗ · 22:13 UTC
- An idea
I couldn't agree more and this is an issue I've persistently been calling for to be solved. 👇 Two temporary solutions I've found are: 1. Branch Chat & Continue - > If you reached the limit in the middle of getting an output, take a screenshot of your…
@iamelijahkhan on X ↗ · 22:31 UTC
- A problem
im not sure if somethign is going on but i havent recieved the last two resets i actually have two account pro max and pro lite. currently cannot work because the last weekly usage for me on monday got drained all the way down to zero for 1 task can you guys check
@Kironkeyz on X ↗ · 22:35 UTC
- Mixed
Stupid idea cuz i can use opus for a solid hour, doing better work than astra which hits 5hr limit in like 20 mins.
@jessehill12333 on X ↗ · 22:39 UTC
- A problem
it seems like since I upgraded to $200 plan 2 nights ago my weekly limit just crashed in 2 conversations!! something is really wronnggg as it seems upgrading midsession does not update the actual weekly quota to the new subscription... PLEASEEEE fixx thisss..or issue a refund..
@amogh2026 on X ↗ · 22:46 UTC
- Mixed
i experienced the same quick drain. i'd like to be included in the added credits. Also, chatgpt desktop codex appears to be loading a model (was working on sol 6.1 when it went stagnant) for about an hour now and i haven't been able to work ever since. Can we get a fix for this?
@domsonline on X ↗ · 23:13 UTC
- A problem
I used astra for literally 1 hrs and my usage is gone again fk ChatGPT
@riwajrise on X ↗ · 23:27 UTC
- A problem
Couple of questions, praying I can get a response here because I have not been able to get through to support for days and it feels like I'm being ignored. - How do we make contact regarding this issue? Pro 5x plan drained 2 resets + you reset me after I wasted a reset
@CodyB5lv6 on X ↗ · 23:58 UTC
- A problem
Pro user here: since Oct 21, ChatGPT Work mode seems to have lost per-message time context, disrupting ongoing chats. Could this be related to the caching change? Support says it can’t escalate or open a ticket. Please investigate and connect me with human support.
@arirro400 on X ↗ · 23:59 UTC
4 X replies from this day are not quoted: Rantr does not have their words or held them back after automated screening. They stay on X.