Wudlig

@MilksandMatcha found you through your cerebras/exa collab on building ai research agents. i build many myself! would love this.

Giving away 5 Codex Pro plans Each person will get 3 months of free Codex Pro (highest tier). Winners will be selected from comments in 48 hours, comment below why you want it.

@Yuchenj_UW Maybe I'm just an imbecile and a 0.1x dev but I truly don't understand why anyone could possibly need more than a ChatGPT Pro and Claude Max subscription unless their job is infinite, unchecked slop generation.

If you had two software engineering offers: > One pays you $500k/year salary, but covers zero LLM tokens. > One pays you $400k/year salary, but gives you $500/day free LLM tokens. Which one are you taking?

@CaoYu25060 @OpenAIDevs @openingai_com @NotionHQ What's 40

@OpenAIDevs @openingai_com @NotionHQ You should shift your attention from codex to 4o

Codex 🤝 @NotionHQ Meet us in NYC on March 17 for a night packed with: Codex demos. Practical workflows. Builders to meet and learn from. luma.com/52o30i5i

Why does Linus know anything about Halal mortgages @linusgsebastian

@XProfessah @wgussml Wtf are you building that you're okay with a model running for 11.4hrs without you verifying its work

i actually prefer 5.3 codex to 5.4 it just doesn't stop until the job is done

@Ominousind They're routing your rich man traffic to us plus subs. The proletariat will rise

Has anyone else seen ChatGPT doing this? I have a pro sub and am getting this limit after sending ~5 messages to 5.4 thinking.

@sandersted @dylhunn @diegocabezas01 i'm pretty sure the website already said 196k for ChatGPT plus at some point. Why does it say 32k again (I say again but I might be misremembering) now?

@dylhunn @diegocabezas01 bumped context windows to 256k for thinking on Feb 20: help.openai.com/en/articles/68… and note to readers: context window is not the max input limit - that’s 128k for ChatGPT

Did you know GPT-5.4 Thinking has a 1M token context window in the API, but only 32K in ChatGPT Plus ($20/month) and 128K in Pro ($200/month)?

Very strange. It's known that they bumped it up to 196k and I'm pretty sure the website was also already updated to say 196k long ago. But now it says 128k again. Either their employees are confused or they're trying to confuse us. Reminds me of when Logan Kilpatrick (love him but it's just an interesting story), replied to me correcting me that plus only gets 8k context for GPT-4T, even though it had already long been upgraded to 32k. He deleted his comment after I corrected him.

@diegocabezas01 They're all using 192k (at least they should if they haven't changed it since GPT-5)

I don't know if it's just me but so far gpt 5.4 has solved 0 coding problems for me and I've ended up reverting and getting Opus 4.6 to do it every time

@sharghzadeh sorry unrelated but just realized you're @lunch_enjoyer 's twin may i please introduce you

We are living in a clown world.

@Andr3jH i declare mujtaba khamenei supreme leader of your girlfriend

I have accepted the role as Alexandra Daddario's boyfriend

@sec0ndstate @SIGKITTEN @thsottiaux I'm not sure the exact amount of tokens I used, but I used it a decent bit, often with xhigh, for a session and barely used up 4% of my weekly limit. I wasn't using /fast though.

@SIGKITTEN @thsottiaux how, lmao? I ran 160k tokens on xhigh 5.4 fast and it blew 12% of my weekly, then took me less than an hour to get down to 20% remaining, even after turning off /fast, it’s cooked.

We are investigating reports of higher usage drain than expected for Codex when WebSockets are enabled, the team is investigating and we will provide updates as we go

@SIGKITTEN @thsottiaux was about to tag you

@thsottiaux shit i barely used like 20%

@kitlangton Aren't you supposed to be fighting the army of the dead or sth rn

I've been using 5.4 for a few weeks now and here are some loose thoughts: I had 5.4 early and you didn't ha ha ha ha h ah ah ah ah haaaaaaa ha h aha ha haaaaaaa

@FriesIlover49 yees so true they should reset our limits once i run out aha

Codex usage must be bugged theres no way im running prompts that before barely used my limits to now being down by fricking 15% on the weekly usage which would be 30% if we didnt have 2x rate limits, thats worse than claude pro using sonnet. Whats going on lol

@_aidan_clark_ i think you meant 4o-mini because o4 was never released

Wait that doesn't exist anymore, I mean o4.

What are people smoking, I've had GPT-4.5 for months!

@SIGKITTEN banger

im just gonna smash /fast gpt-5.4 xhigh and hope they reset limits because of some edge case issue in a couple of days

okay i guess old habits die hard

@OpenAI okay so what's gonna happen to 5.3 instant? @aidan_mclau did you work on this the same way?

GPT-5.4 Thinking and GPT-5.4 Pro are rolling out now in ChatGPT. GPT-5.4 is also now available in the API and Codex. GPT-5.4 brings our advances in reasoning, coding, and agentic workflows into one frontier model.

@FriesIlover49 I'm a big arc-agi hater anyway so I'm not indexing on it much

@wudlig tbf thats jhon benchmaxxing himself, thats difficult to beat

Mogged again by Gemini