Mateusz Zelent

Codex is REALLY SLOW recently. Even on fast mode. Anyone else experiencing this?

There's something seriously WRONG with Codex the past few days Even on GPT 5.5 xhigh Fast model it's painfully slow Is it just me or you also facing the same issue?

With our Codex Usage limits consistently draining faster by the week. Heres a few tips to help reduce the quota decline : •Always start a new thread after 2-3 compactions. •Configure explorer agents, make them mandatory for broad code scans ( hooks help with this ), low reasoning does just fine. • /goals is incredibly powerful, but it drains quota 2/3x faster. So, use parallel orchestration. -Workers - 5.3 codex xhigh has been doing great recently. • Cache begins to build exponentially, so setup an automation, or run manually, each day to clear most of it. Codex knows exactly what to do. • Aim to keep your Agents.md under 300 lines. Im aware thats not enough room, however setup a .projectmemory to overload most of that context. Keeping Agents.md lean. The guiding doctrine. Context engineering now matters more than ever, its not about giving the model more context, but the right context. Hope this helps

Have the codex limits been fixed? I haven't seen any improvements...

This bug in codex /goal just torched my tokens and nobody's talking about it if your agent requires an input and you clicked resume as goal it gets stuck in an infinite loop and obliterates your balance Be aware everyone @thsottiaux

it's been less than 24 hours since my codex limit reset and i've already put a 40% dent in my weekly limit. i can't tell if limits were nerfed or if i'm ripping goal mode too hard. what is happening?

Something is up with codex usage limits, been able to burn through 2 codex acc before 1 Claude ran out which is just crazy

are you still noticing problems with Codex limits?

GPT 5.5 /goal is completely BROKEN It 's draining my usage limit 2-3x faster At this rate, Codex is about to compete with Claude in hitting token limit

What did OpenAI do to the GPT-5.5 deployment now? People are experiencing some major issues with it. I might have to swap over to just all Composer 2.5 for everything now where I can.

Something is WRONG with Codex GPT 5.5 now Token per speed in chat is ridiculously low I guess OpenAI is losing computation power

You guys ever noticed Codex usage limits sometimes randomly drops a few %, before you even used it on the day?

请教一下，为什么什么都没做，每隔一段时间打开 Codex，就会发现使用额度少了 5%？

Are you guys also getting slammed by codex rare limits or am I just too compute poor?

@FCB_Cartel @sama @thsottiaux I have the same observations.

Codex in the last 3 days has been a nightmare. Hallucination after hallucination, writing so much unwanted code, creating things that are not at all intended, taking too much time and burning tokens unnecessarily. Especially /goal. @sama something is wrong. It burned 50% of my Pro plan tokens for the week in 1 night creating a PR of size 104k lines and when asked about the PR : it said it’s useless and cleared the whole PR and closed it.

@jturntdev I have the same fillings

Is it just me or ChatGPT 5.5 nerfed? This isn’t the model i used on release. Claude’s somehow seeing cracks that 5.5 is completely missing.

Codex Usage Limits was fine for two days, now cooked. Either， /goals is completely broken and drains usage 2-3x faster. Or their caching has destabilised. Im on 20x and it feels like im on 5x

@Oliwier1PL @dawid_dawkak o widzisz ;)

@dawid_dawkak Przeniosłem się na Codexa i jestem zadowolony póki co

Google na I/O pokazało Gemini 3.5 Flash. Nie chodzi tylko o kolejny model. Chodzi o szybki model pod agentów, coding i codzienną pracę z AI. 3.5 Flash jest już dostępny, a Gemini 3.5 Pro ma pojawić się w przyszłym miesiącu. Google najpierw stawia na szybkość i użyteczność, potem dorzuci pełną moc w Pro.