Respuestas con fecha de este día en la muestra recogida: 128. Recogidas desde el 5 oct 2026: 162.
Las respuestas de X se encuentran por búsqueda y cada una se comprueba con su post público en X. Última lectura: 26 oct 2026, 23:46 UTC. No se leen todas las respuestas.
- Mixta
Yes, the speed has noticeably improved, but it burns through resources much faster. Is there really any difference between this and me turning on fast mode myself? The usage breakdown shows it's draining way quicker than before.
@Dashuoge en X ↗ · 00:06 UTC
- Un problema
Where the fuck is the promised DEFAULT 50 TPS? Three GPT-6.1 Sol runs hours later: 31.9, 32.1, 32.6 TPS. Same 440-token text, no tools, first-token wait excluded. Fast isn't default. Stop the bullshit victory laps. Show your benchmarks. Do you think we're idiots?
@V3zxS8myOU4056 en X ↗ · 00:10 UTC
- Un problema
Ah, 'optimized for ~50% faster'... I guess that explains why I still can't even log into Codex right now. Did they boost the speed by simply having fewer people able to use it? Great optimization, Tibo.
@eko2050 en X ↗ · 00:22 UTC
- Un problema
Stop this from happening, please. Models are not reliable when they do their own thing and ignore everything given to them. Multiple videos created the other day, no issue. Today, back to ignoring instructions and causing issues. Thanks.
@ChrisFilmFan en X ↗ · 00:50 UTC
- Una idea
really need a space in Dot that is persistent, otherwise CLI credentials gets lost to cloud instance ephemerality
@neutronstar03 en X ↗ · 01:35 UTC
- Un problema
My speeds dropped from 28 tps to 10tps
@swg0101 en X ↗ · 01:37 UTC
- Una idea
alright, that's true... but maybe we should still get an occasional reset? seriously Tibo, with the new limits October feels way harsher than August did, especially considering we were also getting resets almost every day back then
@miu21590 en X ↗ · 02:35 UTC
- Una idea
Could one of the 28 improvements focus on Windows + WSL2 reliability? I counted 248 open app-labelled issues with WSL/WSL2 in the title. Reports include failed startup, project creation and even pwd. Please prioritize the blockers and add end-to-end release tests.
@tdeckers en X ↗ · 02:36 UTC
- Una idea
Add a way to sync Codex chats across platforms, like remote on mobile. my Dot can't even see Project chats properly, even though it has permission on the computer. I'm on $200 plan.
@Wdensign en X ↗ · 03:09 UTC
- Un problema
1. banked reset set to expire oct 5th. 2. asked chatgpt what time, it said 6am. 3. it expired at 12:02 AM 4. Also used 500 paid credits between 12:02AM and 12:15AM while i still had 43%. 4. Support said nothing we can do.
@james_j_otis en X ↗ · 04:07 UTC
- Un problema
1. banked reset set to expire oct 5th. 2. asked chatgpt what time, it said 6am. 3. it expired at 12:02 AM 4. Also used 500 paid credits between 12:02AM and 12:15AM while i still had 43%. 4. Support said nothing we can do.
@james_j_otis en X ↗ · 04:08 UTC
- Un acierto
Fifty percent faster still finishes the remaining quota sooner.
@YishaoRice en X ↗ · 04:18 UTC
- Un problema
1 day of astra = 1 weeks of usage, for 500 a month I can't quite justify this either anymore.
@krasor_xyz en X ↗ · 04:21 UTC
- Mixta
What changed to make it faster? From my experience, 6.1 now performs at roughly the level of Luna. My subscription runs until 7 October, and I've had to switch back to Sol 5.6 because 6.1's output quality dropped noticeably.
@Heberth26 en X ↗ · 04:25 UTC
- Un problema
I noticed my usage limit is dropping really fast
@system_TF en X ↗ · 04:34 UTC
- Mixta
Chat and Work just merged on my desktop. Regular Chat is completely gone. I only have ChatGPT Work + Codex now. I posted the UI change on my profile because I’m curious how many people are seeing the same rollout. Really hoping this doesn’t mean everyday chats will start…
@wenvotop en X ↗ · 05:15 UTC
- Una idea
Tibo, could Dot have a clearer permission approval feature in Codex? Sometimes it needs approval, but the request comes after a long message, so I may think it’s just chatting with me. It’s hard for users to tell whether a message is normal chat or actually waiting for permission
@Yubin_Shi1101 en X ↗ · 05:36 UTC
- Un acierto
Mohammed, i genuinely respect your tenacity. I have $200 for both, and I can tell you now on those plans it is day and night. Before Astra and Sol Gpt usage was a dream and Anthropic a nightmare, not so anymore, and with the kind of work I am doing with it Anthropic is more cost…
@Al41611876 en X ↗ · 06:34 UTC
- Un problema
Codex tasks assigned and managed by my dot in the cloud, with execution on my own computer, are still extremely slow. The context seems to fill up and get compacted repeatedly, and I suspect this may be contributing to the delays. I haven’t confirmed the cause, but I’m not…
@pcyuan30 en X ↗ · 07:11 UTC
- Mixta
I see 35 tps :) right now in openrouter is that an improved already
@ThnhNgu97660433 en X ↗ · 07:17 UTC
- Mixta
I always use yolo mode, but I guess this is a nice improvement for those that use auto review.
@redcypress0 en X ↗ · 07:23 UTC
- Un problema
For fuck sake, I burnt 1% of weekly usage every 5 minutes running 1 section (Pro 200)
@SCRX89 en X ↗ · 07:27 UTC
- Una idea
pls reset tobo i dont give a f about auto review son
@JustLingonberry en X ↗ · 07:30 UTC
- Un problema
Auto review inside Codex is currently awful. Regularly rejects simple actions. Wouldn't be celebrating this to be honest, @thsottiaux. It's an impediment currently, not an improvement.
@majorcontentCT en X ↗ · 07:48 UTC
- Una idea
Could you please consider removing the 5-hour limit on Plus? The 5-hour cap really makes it hard to stay focused and get deep work done. I often hit the limit right when I’m in the flow, which breaks my concentration and slows down progress.
@fakeUserK en X ↗ · 08:10 UTC
- Un problema
Could the team look at dot/local Codex thread access? Direct reads of existing Mac threads still fail; a previously working relay lost its tools. Fresh relay reads work, but the old one remains broken. Repros:
@ShugeiWorkshop en X ↗ · 09:04 UTC
- Un problema
I don't have this option on my app after the last update (today)
@rodrigossilvano en X ↗ · 09:07 UTC
- Un problema
what do u mean by default speed? it seems very slow like before
@ijaack94 en X ↗ · 09:13 UTC
- Mixta
I thought the 200 plan nerf hasn't hit yet? At least it says that in the email I got. Although I agree it feels like it's going faster, but I think that's because I'm more used to Opus 5.5 now...
@JacK_AsterisK en X ↗ · 10:17 UTC
- Un problema
what has happened to Dots? It's now struggling to do even the smallest tasks, everything has a 'blocker', even though I've given full permissions in settings. Feels completely broken
@Sonnythredly en X ↗ · 10:17 UTC
- Un problema
Codex desktop browser fails on 7 public sites: “saved browser permissions could not be verified” / “Browser Use could not request permission.” Manual opening exposes a tab’s title/URL, but page-content reading still fails. Reproduced Oct 6, 2026.
@santianekiudun en X ↗ · 10:50 UTC
- Un problema
I dont see any improvements on speed. GPT 6.1 Sol is suppppppeeeer slow. You can wait from 30 to 60 min for a simple updates. it's hilarious when you try claude opus 5.5 it's on the other level. What about you guys?
@paw_pixels en X ↗ · 12:06 UTC
- Una idea
btw if you’re not using it codex’s auto review is very good, if you say you approve some things, it will understand the next try, not a simple classifier
@alakazam1314 en X ↗ · 12:33 UTC
- Una idea
A small product / UX idea: when ChatGPT restarts (e.g. after an update), bring me back to exactly what I was looking at – which is usually Inbox/Priority mode with a mixture of read and unread threads; I lose that state if update brings me to the home screen – clicking on the…
@borekb en X ↗ · 13:03 UTC
- Un problema
Hi @thsottiaux — I upgraded from Pro $100 to $200 with 18% of my weekly allowance left. Instead of starting at 100%, I got only ~59% and kept the old reset date. A new customer paying the same $200 starts at 100%. Why does upgrading penalize an existing customer?
@Carlindo en X ↗ · 13:04 UTC
- Un problema
Sorry, but is still using the same output speed on my end And yes I used Sol 6.1 on Ultra with x1.5 faster results, but inference speed keep looking bad also there is time where the model starts doing nothing and your limit disappears without the model just doing anything.
@MikeGonzalezDEV en X ↗ · 13:08 UTC
- Un problema
I don’t see it either. Pro account here
@BekahTorri en X ↗ · 13:22 UTC
- Un problema
Wait, why did this blow up with likes?)) Don't get me wrong, ChatGPT/Codex is a great product. But the speed is genuinely painful compared to competitors, and that $500 plan is crazy. Zero beef with Tibo personally. My only point, and I really want to stress this, is that you…
@d_ridelman en X ↗ · 13:29 UTC
- Una idea
Yo tibo, GIVE US A RESET so we can test all the new features, im out of usage so i cant enojy this new improvements...
@bulltraper en X ↗ · 13:35 UTC
- Un problema
Aww! Really? Mine totally did! In fact... I just did it again! So maybe its not... everytime? Same exact error too...
@cless_aurion en X ↗ · 13:35 UTC
- Un problema
I’m close to cancelling my subscription over Dots’ broken approval system. Mine has Full Access, yet approved actions keep getting denied. I contacted OpenAI support and still have no answer. An assistant that can’t act with my permission isn’t worth paying for.
@MiladShari10725 en X ↗ · 13:57 UTC
- Un problema
Where is this option? I’m on 26.930.61225 and I don’t see it in permissions. Just have default and the option for full access. No auto review option.
@brent_palmer en X ↗ · 14:16 UTC
- Mixta
Ran through the weekly Pro 20x limit in 23 hours, while Astra Extra-High did game dev tasks after a full reset. I used 2 resets since October 3rd 😬… wasn’t consumption supposed to be better? The results are nice, not complaining too much, but 23-28hrs of work is a low limit.
@Nebula_Matrix en X ↗ · 14:49 UTC
- Un problema
Astra is and always was fast for me. Sol 6.1 is still at 20 t/s or something, taking (over) 3 times as long for a task as Sol 6.0.
@MakJoris en X ↗ · 15:10 UTC
- Un problema
I upgraded from $200 to $500 today, but hit my weekly limit after roughly 2 hours using Astra on High across just 3 chats. It feels like I got even less usage than on the $200 plan.
@stanmandrik en X ↗ · 15:36 UTC
- Un problema
The problem with auto-review is that it still asks me for things I have clearly stated I am giving it full permission to do. As a result, it is still not able to do things like: - create an API key for me - go get my API key from my environment variables or my iCloud keychain
@MichaelStolarz en X ↗ · 15:47 UTC
- Un problema
Still not seeing the auto review option available, Chatgpt updated, computer restarted. Located in Taiwan. Also, don't feel the speed improvement too. So where's the actual improvements or it is only available for some people first. First our $200 got cut off half and then this?
@William16816888 en X ↗ · 15:51 UTC
- Mixta
What we are after is very snappy and fast output, like what we see in Opus 5.5, but with the reliability and veracity of Astra/Sol. Actually, 50 is not working at all—it just keeps running without any indication of speed, so no, we do not see any visible change. We also need…
@madanpandit en X ↗ · 15:54 UTC
- Mixta
lol I have been doing “dangerously access permissions” since the beginning, it has not once did anything bad, even the dumbest models. In fact, the models are still being extra careful than I would like them to be
@swapp19902 en X ↗ · 16:13 UTC
- Un problema
Still no options in the most current version of app
@dreaminalshots en X ↗ · 16:21 UTC
- Un problema
So now 5h limit will burn in 5 mins?
@manticore8100 en X ↗ · 16:22 UTC
- Mixta
You mean the option that has been there the entire time?
@Flight1999x en X ↗ · 16:25 UTC
- Un problema
Hey @thsottiaux , i don’t know what else u have changed other than auto review but apparently even with full access permissions codex would not work on github and especially hf (u guys too scared?, trust ur agents💪)
@shreshthsaini en X ↗ · 16:27 UTC
- Mixta
Is this being sold as a major optimization?!
@WiseSam_ en X ↗ · 16:34 UTC
- Un problema
Day 1 was 50% faster. Speed isn't what's missing. My weekly limit now lasts one day. And here's where it goes. The launch-vs-today train comparison going around shows it well. Same prompt: at launch the train was the subject. Today there's more detail and the wrong focus. That…
@WholeEntertainm en X ↗ · 17:35 UTC
- Un problema
Still not showing up in the app... It's been really frustrating being an oAI user in the past weeks even though 6.1 is a good model. A small setting that doesn't even show up, yet, won't change that. I seriously hope 2.2 is going to be a reset.
@xstEvEdEx en X ↗ · 17:42 UTC
- Un problema
The update was supposed to improve ChatGPT, but it didn't improve anything; instead, it just created a redundant, unnecessary feature that most users don't need. You cut our limits from 20x to 10x—thanks for the extra tokens, I guess—so you should lower the subscription price to.
@t_klimo en X ↗ · 17:43 UTC
-
That would be great if you guys have an additional day off every week (no updates) over the next 28 days. We’d all benefit from that 🤗
@Andrej_Gap en X ↗ · 18:06 UTC
- Un problema
Yo tibo, I noticed my usage drains faster if I change from solo to astra back to sol again, it kept using astra or something, can you check
@Gabriel1860256 en X ↗ · 18:28 UTC
- Un problema
come on, im sick and tired of this , now its fucking fast but drains the credit like there is not tomorrow wtf , stop fucking change the usage system every fucking day
@eyasu_deme96575 en X ↗ · 18:34 UTC
- Un problema
That’s not “improving the product.” That’s quietly changing the deal after people already paid for it. I genuinely can’t think of another product I use that gets away with this kind of shit. Slow as hell. Quota disappears insanely fast. Constantly changing limits and user benefit
@koinya_hhh en X ↗ · 18:59 UTC
- Un problema
I dont thing 6.1 get and speed !
@EmadAlslahi en X ↗ · 19:17 UTC
- Mixta
I think I call day 2 a soft W... As much as I want another reset. It's... It's fair. Can't say I notice TOK/S as faster. You should put a meter inside Codex to make it transparent so that improvements and performance in that realm are obvious.
@Tyfoods4Thought en X ↗ · 19:19 UTC
- Un problema
in codex linux when it was doing large task the codex application was getting stuck even tho code changes are happening and codex is completing the task. The UI will just get hang. Once I close and close it will show the task was done. Please address this issue.
@VarmaGowrish en X ↗ · 19:22 UTC
- Una idea
We need no features , Only Resets 😁
@PofPofPofPof en X ↗ · 19:28 UTC
- Una idea
Next can you make it so my agents stop replying to an old message in the chat rather than most recent one?
@yasnar_dev en X ↗ · 19:32 UTC
- Una idea
我觉得有个改进值得重视！就是当codex 控制某个软件的时候 他经常会强行切换窗口 打断我另外的工作任务，强制霸屏。 然后我和codex就会经常来来回回抢夺窗口使用权。 如果codex能默默的后台操控所需的软件，而不打扰我 那该有多美好啊！
@Jaded_Fitz en X ↗ · 19:32 UTC
- Un problema
My one regular weekly limit + 2 bank reset and 45000 credits are gone with 1 day work - this ridiculous to discuss about the problems in OpenAI solutions.
@darkokarac en X ↗ · 19:33 UTC
- Un problema
Yo@thsottiaux What is the point of the dot>? I dont get it at all after using it for 3 days now. I cant get it to do anything without going to my mac and giving further permissions. It wont allow me to direct my dot to use my mac to check email and reply, or to edit documents,…
@guyontheearth en X ↗ · 19:55 UTC
- Un problema
On ChatGPT Web > Codex > Settings, there is no Permissions. I looked, but am unable to find Auto Review.
@LittleBrainz en X ↗ · 20:02 UTC
-
Why would you do this to yourself lmao, a glutton for punishment
@WesternValuesW en X ↗ · 20:03 UTC
- Mixta
the fallback being "reset everything" makes this the most honest 28-day roadmap
@Ibnu_J1 en X ↗ · 20:11 UTC
-
Tibo shipped two things today and called one of them "smaller." Bro graded his own homework. lol Follow my page for the official voting poll by the consumers so @thsottiaux knows how we feel! I track every reset and every ship, all 28 days, live:
@DaResetButton en X ↗ · 20:18 UTC
- Un problema
Why is this so confusing? Is it an Auto-review toggle? Is it 'Approve for me' under the composer? I and so many people in the comments cannot find the toggle. But the Approve for me thing does not seem like the same thing. How do we use this feature in a way we think it's worth…
@nibelheim en X ↗ · 20:20 UTC
- Mixta
In the meantime, Codex on Windows every hour or so on a stable 2.5gbps wired fiber connection.
@jeeacheff en X ↗ · 20:32 UTC
- Un problema
I'd say Oct 6 did not deliver 'clear improvement and relevant for most codex/work users' Auto review was here for a long time. API price changes not relevant.
@ruryop en X ↗ · 20:38 UTC
- Un problema
After your speed bump to 6.1 sol, it is also consuming the usage quota faster ...
@MohammedJunaidG en X ↗ · 20:41 UTC
- Mixta
Uhm, sorry but wasn’t it available since Dev Day? I’ve been using it for couple days now
@theo_devs en X ↗ · 20:48 UTC
- Un problema
Do I miss smth? For me the same error and debug wont help it xD
@ArtemkaKlen en X ↗ · 20:49 UTC
- Un problema
this meeting could've been an email and my codex could use a reset
@uptodivyansh en X ↗ · 20:51 UTC
- Un problema
Tibo… man… codex chats showing up in my chat app is absolutely nuking my workflow….
@FStacked en X ↗ · 20:53 UTC
- Un problema
I've hit my weekly limit, and there's no option to buy a reset at a reasonable price, like a quarter of the monthly subscription fee. I'm not even asking for a free reset. You're basically forcing me to subscribe to Claude to finish my project halfway through development.🙃
@univerghost en X ↗ · 20:53 UTC
- Un problema
I work solo. Where is my reset bro?
@aaliyaanX en X ↗ · 20:53 UTC
- Una idea
Nah this has to stop. Reset or increase are usage please.
@Liam2307 en X ↗ · 20:54 UTC
- Un problema
it's ok if your models are not the smartest, for a little while. but for 100 euros you can't 1. first kill usage limits so they run out incredibly fast 2. kill tok/s massively to make the subscription unusable 3. compensate us in a way that requires us 2 check twitter 4 rsts
@sabaka_339 en X ↗ · 20:56 UTC
- Mixta
I thought the improvements were meant for Codex's users? Day 2.5/ now give us a reset. And increase usage limits.
@hunoematic en X ↗ · 20:58 UTC
- Mixta
That’s how I’ve been using chat. I have work summarize what we’ve done and while I wait for my usage to reset I strategize with chat.
@ThisIsFRSH en X ↗ · 20:59 UTC
- Un problema
so you released jev but expensive. this is not codex improvement. put the reset in the bag.
@weezerOSINT en X ↗ · 21:00 UTC
-
You mean a banked reset, right.... Right!? 😳
@DavidWilliams en X ↗ · 21:09 UTC
- Un problema
A banked reset? Sure! And dodge reset.. no! 48% usage and it resets in 3 days! Dont waste my 48% and then make me have to wait 7 days for the next reset again!
@BeatzCreativeHQ en X ↗ · 21:11 UTC
- Mixta
Neither, codex is too slow for me to do any productive work. Dot is not here, usage cut in half, Maybe the ultrafast on sol 6.1 might help me use the resets before plan cancels, can we ship that?🥸
@M_Wahib01 en X ↗ · 21:11 UTC
-
"We've removed dark mode to help improve the system" 😳
@DavidWilliams en X ↗ · 21:11 UTC
- Mixta
2.1 I use full access already, never had a problem 2.2 unrelated for 99% of your users 2.3 devday, Mac only, irrelevant 2.4 not integrated yet, sol is slow as ever for computer use. Just reset.
@NuminousVoidX en X ↗ · 21:12 UTC
- Un problema
Hey is there a reason I can’t set custom rules when using Dot on a business plan? Seeing that fixed would be amazing
@timothyachumba en X ↗ · 21:14 UTC
- Mixta
Used about 30% of my weekly quota today on GPT-6 Sol Extra High, and the speed has been great so far. Honestly feels like 6.1 Sol isn't even needed 😄 A reset would be amazing though. Also vibe-coded a token speed tracker just for fun:
@d_ridelman en X ↗ · 21:16 UTC
- Un acierto
Today I used my dot to schedule a meeting with an external vendor. Used the Meetings plugin to transcribe the meeting. Clicked one of the generated action items to automatically create a new thread to perform additional research. Generated PDF with findings and sent it on slack.…
@GrantWasil en X ↗ · 21:16 UTC
- Una idea
Can we get a reset as well? We got a lot of new things to try.. feeling token poor
@GreenFuture2025 en X ↗ · 21:17 UTC
- Un problema
Good day but I do need a reset
@theo en X ↗ · 21:19 UTC
- Un problema
For example, on the $200 plan, a full week means 1,000 credits. But during a 'reset', we only get 100 credits, even though the system shows the weekly bar filled to 100%. It gives you 10% of the quota but presents it as a full reset. we want full/real RESET
@misterExB en X ↗ · 21:21 UTC
- Un problema
got model unavailable at least 3 times with 6.1 sol so need 3 banked resets please, ty
@tiagoefreitas en X ↗ · 21:26 UTC
- Mixta
I think there are no virtual improvements for most Codex users. Most are just feature flag changes from either restricted functionality or rate limiting. This seems just more engagement farming than legit making the product more compelling to the majority of the users.
@Jossuecito en X ↗ · 21:29 UTC
- Un problema
The Meetings plugin (0.8.77) can’t load settings in Codex—the API returns a Cloudflare challenge (403) instead of JSON. Could you take a look?
@agereed1 en X ↗ · 21:29 UTC
- Un problema
I feel like my usage is back to burning up as fast as it was a few days ago! Yesterday I only used about 20% of my weekly and then I had to reset this morning and I’m down to only 9% left so I’ve used a whole weeks reset in one day on similar tasks that took only 10% yesterday.
@InsiderLocalTV en X ↗ · 21:32 UTC
- Un problema
Sad thing is that I have already stopped using Codex. I have moved down to $100 plan, waste of money. At this rate, I will be unsubscribing by next month as the models are nowhere close to Opus 5.5 from the quality and usage perspective. These improvements are useless.
@mmuruganandam en X ↗ · 21:34 UTC
- Mixta
Codex just got 12% faster on our same 12 coding tasks: 607 s down to 534 s. Before picking a side, check what you run. One kind of task got far faster, the other barely moved. Long answers or short turns for you?
@stas_sorokin_ en X ↗ · 21:35 UTC
- Un problema
I won't tell you that I burned through my extra 25,000 topped-up credits today, desperately needing a reset, yet still didn't get my 6.1 SOL 8x speed.
@yllll_yl en X ↗ · 21:44 UTC
- Un problema
Works great until the transcript is 40 minutes of me changing my mind. Does Codex boil it down to decisions + open questions before loading it, or take the raw dump? Raw eats context fast and the agent tends to build the version I talked myself out of.
@xuqze en X ↗ · 21:45 UTC
- Un problema
Just updated and still nothing, the setting doesn’t exist on windows.
@FJBNews en X ↗ · 21:47 UTC
- Un problema
why did you have to ship this meaningful update today… i need a reset badly
@marcuslayerx en X ↗ · 21:48 UTC
- Un problema
All fun and games all these upgrades, but i am lowkey hoping for a reset since I already spent all my tokens for the week and can't use them 😂
@sirwaltert en X ↗ · 21:50 UTC
- Mixta
I had a similar incident. After Tibo tweeted, and like after 2 hours I sat to do my work, and it was indeed burning less and getting more work done. But this morning I just burned through my weekly allowance so fast. Had to use a reset, and now the reset is pushed back 3 days.
@perhaps_aadii en X ↗ · 21:51 UTC
- Una idea
A reset would be great ! 1% left!
@jshubham0119 en X ↗ · 21:57 UTC
- Un problema
You broke our vscode codex so reset
@iiciis__ en X ↗ · 22:04 UTC
-
Dot(s) sucks man, it cant do anything right and was a rushed to market product....my company on the other hand is about to release a local auto agent, pretty much everything DOT(S) should have been....will be released at
@LocalCaptionspc en X ↗ · 22:05 UTC
- Un problema
Welp tough to trust when my codex ultra chose 3 Astra agents to download a bunch of csvs that drained like 60% of my codex account yesterday 😂. Love you tibo, but sent my agent to chat w OpenAI support for a replenishment. Would love your help!
@mikesowerss en X ↗ · 22:19 UTC
- Mixta
Sol is fast, Astra is slow (regression). Reset!
@raksa_ai en X ↗ · 22:29 UTC
- Un problema
Yo Tibo, out of nowhere I can't use Codex because my account was "deleted or deactivated", no reason was given. Still had days of sub left. Is this a sign to switch to Claude?
@iMiguelAIS en X ↗ · 22:30 UTC
- Un problema
In 15 hours, I spent the entirety of one week's $200 quota ($400 in three weeks). I used to get this much usage in a day with 5.5. I could not possibly be more disgusted by Codex. Cancelling both accounts.
@macintogdev en X ↗ · 22:30 UTC
- Un problema
does auto-review see the tool calls that got cancelled when a sibling failed, or only the ones that actually ran? that gap was the bug behind a fix i landed in the agents sdk (#4195)
@pranavmishra28 en X ↗ · 22:32 UTC
- Un problema
Naaaah man I burnt 80+ % of my weekly usage when u did the first pool i ran 16 tasks all astra extra high I rewrote the whole directory U can't do me like that
@A4zbiz en X ↗ · 22:40 UTC
- Un problema
It seems that when a reset occurs, GPT's memory gets wiped too; it couldn't recall any of the previous conversations or files at all.
@heylistenm33663 en X ↗ · 22:41 UTC
- Un problema
I can’t even find that in the settings here.
@Nando_Guimaraes en X ↗ · 22:55 UTC
- Un problema
if you don’t hit the button soon I’m going to have to switch to Grok Bot. Takes me less than a day to burn through my pro plan now
@Mattmaximo en X ↗ · 23:03 UTC
- Un problema
I find what is happening very unfair, and Autoreview is not available to me. I want a complete reset. I pay over $200 a month, which sounds like a joke.
@FrancescoM51775 en X ↗ · 23:14 UTC
- Un acierto
これヤバくない？ Codexチームに作業させた時、まさにこの自動承認のトークン消費が無視できないレベルで高くてきつかったから、超助かる
@wemp12 en X ↗ · 23:14 UTC
-
How about we all get to vote on if it's a clear improvement or not.
@tpacker en X ↗ · 23:24 UTC
- Un problema
Couldn't enable Meetings.😂
@xiaomabosn en X ↗ · 23:53 UTC
1 respuesta en X de este día no se cita: Rantr no tiene su texto o lo retuvo tras la revisión automática. Sigue en X.
