Daniel Shapiro

The PA rn (shorts safu)

Blockworks Research team showing out at @blockworksDAS Check out @0xcarlosg, @0xMether, and @shaundadevens. Going to be absolutely electric

- with @ i want to be able to tag folders instead of just files. a way to filter for ONLY folders would be even better - hotkeys for switching between threads. for instance cmd + 1,2,3,4... for the top thread in your folders. Also cmd + for going to most recent completed thread ty Tibo 🙌

What are we consistently getting wrong with codex that you wish we would improve / fix?

BWR subs saw it first

@iannuttall id xhigh stable now?

i'm addicted to gpt-5.4 xhigh fast with subagents

DAS VIP completely sold out Limited availability left and GA prices rising soon Grab one soon or miss out on the best insitutional conference of the year

BREAKING: @saylor to speak at DAS NYC March 26th. Be there.

Have you been running it on both inside and outside the app? Do you actually feel like you are getting 2x limits in the app? Especially last few days where limits are getting reset daily so i can compare usage directly, I personally haven't seen a difference. Been trying to measure this bc I prefer cmux > app

@thsottiaux @gabrielchua you probably know, but this reset gimmick is buying you a lot of goodwill. now the final way to ascend would be to announce on april 2nd that 2x limits are the new normal limits.

OK, Codex is back and stable and we should be good for a while. Reset button pressed, should see it in a bit

NEW: @BNYglobal to speak at DAS NYC Hear from Robin Vince on how he’s driving digital transformation at one of America’s most storied financial institutions

The CEO of one of the worlds largest banks will be speaking at DAS in two weeks Best institutional conference in crypto, hands down

@zeddotdev When drag folder into ghostty terminal and the folder path is automatically pasted 😭😭

Coming Zednesday...

My thoughts (pro sub): -worse than 5.3 codex at coding and planning. It doesn't "just get it", was failing at simple redactors -uses up limits nearly twice as fast (haven't been using fast mode) -tried it in the codex app, limit burn the same. There's absolutely no way you get 2x limits in the app -extremely slow. I honestly can't tell a difference in fast mode Very disappointed. Just using 5.3-codex now

GPT 5.4 has been out for a few days now. So, what's the verdict? How are the vibe benchmarks going? I've been traveling and havent had the chance to play with it. Just a wee bit jealous. Honest opinions? 👇

@LucasVHoutven @thsottiaux I'm not using fast mode, not using 1MM context. just gpt-5.4 xhigh in the codex app.

@thsottiaux @thsottiaux 5.4 even on medium is munching away my usage even on non-fast mode and no 1M context. I’m on pro. The difference with 5.3 codex is night and day. Have y guys noticed this? I though my it’s more efficient?

Codex at 2M+ active users up 25% week over week... and that was before we launched the app on Windows and GPT-5.4!

@LucasVHoutven @thsottiaux my 7d limit got destroyed today, and i am completely in the codex app. no way this is giving 2x usage...

@thsottiaux hey Tibo - would love the codex app to have sub-terminals instead of threads like cmux. a lot easier to navigate many sessions that digging through the individaul folder threads

GPT-5.4, WebSockets, /fast mode, oh my Codex Update now for the latest goodies openai.com/index/introduc…

@LLMJunky what's your intuition here? Do you think it's worth it? I feel like auto-compaction is so good, would rather use the extra limits on speed as opposed to 1M window. curious to hear your thoughts once you start testing

GPT 5.4 has an experimental 1M context window you can configure inside of Codex. And unlike the flicker company, it works on your ChatGPT plan instead of requiring API rates. It does consume 2x more usage, but that's still notably cheaper than paying $22.50/mtok. To enable, add this to the top of your config file: model = "gpt-5.4" model_context_window = 1000000 model_auto_compact_token_limit = 900000

We’re excited to be working with HyperLend as it continues to grow lending adoption in the HyperEVM ecosystem. 🧵