@iyzebhel To get these responses they generate these screenshots in gpt-image-2 for engagement
English
Alex Berger
23 posts
What do people do to get these responses? lol
I love this side of him.
Shub@shub0414
Bruh, who tf Claude think he is
English
We need to talk about token spend creep… $1.3 million per month cannot be efficient.
The future of software should not add a $20,000/month fee to your business just for tokens to run a single feature, especially as you add hundreds of features…
Peter Steinberger 🦞@steipete
The latest CodexBar update renders API costs wayyyy nicer. codex.bar
English
@robinebers You say image generation workflow is a dead end, but, respectfully, it doesn’t make sense. We as humans can imagine things, not only write them down. There is a reason why you had UX/graphic designers alongside frontend devs to get the best results. gpt-i-2 is slow as hell though
English
goodbye codex
unsubscribed from the $200/mo plan for now
→ love the team
→ love the app
→ love the intelligence of the model
but my business isn't just about backend code
it's a lot of landing pages. a lot of emails. a lot of content, intros, scripts, tweets, and instagram carousels.
none of which codex is good at
English
Hiring a dev team would be cheaper than this 😭
Peter Steinberger 🦞@steipete
The latest CodexBar update renders API costs wayyyy nicer. codex.bar
English
In hindsight Shl0ms really should've disclosed upfront that the painting was done by Claude
𒐪@SHL0MS
i just generated an image in the style of a Monet painting using AI please describe, in as much detail as possible, what makes this inferior to a real Monet painting
English
@rawespresso Seems more like your definition of comfortable doubled in 15 years.
English
The salary you actually need to feel comfortable in the UK in 2026 has more than doubled in the last 15 years.
In 2010, £35,000 was a solid graduate salary. You could rent decently, eat out, take a holiday, save a bit.
In 2026, the equivalent number — same lifestyle, same level of comfort — is closer to £75,000.
That's the difference between 'affording your life' and 'getting by.' £35K used to put you on the comfortable side of that line. £75K is now the entry point.
Most people earning £45-£60K in 2026 are quietly running paycheck to paycheck, feeling vaguely confused about why a salary that sounds substantial doesn't actually go very far.
English
@mattpocockuk I wonder if they truly believe they gifted us a present, or if this is British Airways-level of hypocrisy from multiple Anthropic accounts.
English
This is the clarity we've been crying out for.
But it's a poisoned chalice. This is a 10X cut to claude -p disguised as a monthly bonus.
Anthropic is discouraging any kind of programmatic usage. And that's fine - no subsidy lasts forever.
But it's time to try Codex.
ClaudeDevs@ClaudeDevs
Starting June 15, paid Claude plans can claim a dedicated monthly credit for programmatic usage. The credit covers usage of: - Claude Agent SDK - claude -p - Claude Code GitHub Actions - Third-party apps built on the Agent SDK
English
@weswinder Most of my claude -p usage is cron jobs. I guess I will just migrate them to routines then. Those Mac Minis will finally be very useful soon 😆
English
anthropic just downgraded claude code subs
programmatic usage previously used your full rate limits
which was something like a 5-10X discount vs api costs
now ALL programmatic usage is capped at $20/$100/$200 in api costs (a single opus session can easily burn $20)
not good
ClaudeDevs@ClaudeDevs
Starting June 15, paid Claude plans can claim a dedicated monthly credit for programmatic usage. The credit covers usage of: - Claude Agent SDK - claude -p - Claude Code GitHub Actions - Third-party apps built on the Agent SDK
English
gpt-image-2 is incredible, but also incredibly slow. Is there no way to speed things up inside Codex?
English
@theo @ClaudeDevs What did they change exactly in this case? Any claude -p will drain these new credits?
English
@ClaudeDevs Framing this as a free credit instead of a regression for users is wild.
Now I have to make the Claude Code experience on T3 Code significantly worse in order to keep using the 25x subsidization provided on a Claude Code subscription.
x.com/theo/status/20…
Theo - t3.gg@theo
Solution incoming…
English
Starting June 15, paid Claude plans can claim a dedicated monthly credit for programmatic usage.
The credit covers usage of:
- Claude Agent SDK
- claude -p
- Claude Code GitHub Actions
- Third-party apps built on the Agent SDK
English
@levelsio The craziest thing is that OpenAI still doesn’t send invoices by email I think and you have to go into the UI to download it from Stripe 😆 When invoices come by email I just auto forward it and they get semi-automatically reconciled with Dext and Xero, no trouble.
English
Tax and accountancy software and bookkeeping industry have perverse incentives to keep the tax code and bookkeeping hard so you have to keep paying a lot for their software and to hire them
If they let AI just do it via their APIs, their value woudl go to $0 fast!
Ingmar Bruinsma@Kingmar1991
@levelsio Should not be that hard for companies to provide a pull function. Just that. So you can pull stuff after you've gotten a token from them. I don't see how that is negatively impacting anyone. If I hack your password for your login to download your invoices, that's the same thing.
English
@StatisticUrban And that’s why London is such a great place to live. With spaces like this easily accessible by tube/Overground.
English
🚨 ChatGPT is the best frontend model of the world
They pretty much solved it this generation , we need one extra step
1. generate a image of the ui
2. clone pixel to pixel
3. best frontend of the world 😼
Angel 🌼@Angaisb_
I hope OpenAI finally solves front-end with GPT-5.6, GPT-6 is still too far away To be honest I thought GPT-5.5 would've already solved it
English
@steipete Now that Codex has image-2 inside, have you tried asking it to mockup first, review/improve the mockup, then code vs go straight into coding? Output feels much more "designed". One shot example (obv needs iterating)
English
@catalinmpit I have Codex in ChatGPT app, but it is only for Cloud agents, which I don't find really useful.
So I am forced to remote desktop from iOS into my Macbook fml
English
Whatever Anthropic's reasons are not to create an image model, I think it will hold them back vs OpenAI.
English
What seems to work well: Telling Codex to generate a UI mock up, then review and critique it, then iterate with image generation. Once it is good after several turns, give it the mockup as a goal to implement in code.
English
Claude beats Codex at frontend? Not anymore.
Having gpt-image-2 in Codex feels like a game changer. Going between image generation for mockups and code goes much harder than Claude alone.
English