Posted by mudkipdev 3 hours ago
That's hilarious. Does OpenAI even know this doesn't work?
Or, people just stopped thinking about any sort of UX. These sort of mistakes are all over the place, on literally all web properties, some UX flows just ends with you at a page where nothing works sometimes. Everything is just perpetually "a bit broken" seemingly everywhere I go, not specific to OpenAI or even the internet.
OpenAI now has three price points: GPT 5.1, GPT 5.2 and now GPT 5.4. There version numbers jump across different model lines with codex at 5.3, what they now call instant also at 5.3.
Anthropic are really the only ones who managed to get this under control: Three models, priced at three different levels. New models are immediately available everywhere.
Google essentially only has Preview models! The last GA is 2.5. As a developer, I can either use an outdated model or have zero insurances that the model doesn't get discontinued within weeks.
What's funny is that there is this common meme at Google, you can either use the old, unmaintained tool that's used everywhere, or the new beta tools that doesn't quite do what you want.
Not quite the same, but it did remind me of it.
"Ok, here is the translation:"
'we don't want to offer support'Really, the economics makes no sense, but that's what they're doing. You can't have a consistent model because it'll pin their hardware & software, and that costs money.
Sure, makes total sense guys.
I don't know, this feels unnecessarily nitpicky to me
It isn't hard to understand that 5.4 > 5.2 > 5.1. It's not hard to understand that the dash-variants have unique properties that you want to look up before selecting.
Especially for a target audience of software engineers skipping a version number is a common occurrence and never questioned.
It's really nice to see Google get back to its roots by launching things only to "beta" and then leaving them there for years. Gmail was "beta" for at least five years, I think.
Why are you using the same model after a month? Every month a better model comes out. They are all accessible via the same API. You can pay per-token. This is the first time in, like, all of technology history, that a useful paid service is so interoperable between providers that switching is as easy as changing a URL.
I guess that's true, but geared towards API users.
Personally, since "Pro Mode" became available, I've been on the plan that enables that, and it's one price point and I get access to everything, including enough usage for codex that someone who spends a lot of time programming, never manage to hit any usage limits although I've gotten close once to the new (temporary) Spark limits.
naming things
cache invalidation
off by one errors
I also don't believe there is any value in trying to aggregate consumers or businesses just to clean up model makers names/release schedule. Consumers just use the default, and businesses need clarity on the underlying change (e.g. why is it acting different? Oh google released 3.6)
Also per pricing, GPT-5.4 ($2.50/M input, $15/M output) is much cheaper than Opus 4.6 ($5/M input, $25/M output) and Opus has a penalty for its beta >200k context window.
I am skeptical whether the 1M context window will provide material gains as current Codex/Opus show weaknesses as its context window is mostly full, but we'll see.
Per updated docs (https://developers.openai.com/api/docs/guides/latest-model), it supercedes GPT-5.3-Codex, which is an interesting move.
> For models with a 1.05M context window (GPT-5.4 and GPT-5.4 pro), prompts with >272K input tokens are priced at 2x input and 1.5x output for the full session for standard, batch, and flex.
Taken from https://developers.openai.com/api/docs/models/gpt-5.4
> GPT‑5.4 in Codex includes experimental support for the 1M context window. Developers can try this by configuring model_context_window and model_auto_compact_token_limit. Requests that exceed the standard 272K context window count against usage limits at 2x the normal rate.
I don't think that's a fair reading of the original post at all, obviously what they meant by "no cost" was "no increase in the cost".
Switch between Claude models. Applies to this session and future Claude Code sessions. For other/previous model names, specify with --model.
1. Default (recommended) Opus 4.6 · Most capable for complex work
2. Opus (1M context) Opus 4.6 with 1M context · Billed as extra usage · $10/$37.50 per Mtok
3. Sonnet Sonnet 4.6 · Best for everyday tasks
4. Sonnet (1M context) Sonnet 4.6 with 1M context · Billed as extra usage · $6/$22.50 per Mtok
5. Haiku Haiku 4.5 · Fastest for quick answersFor Codex, we're making 1M context experimentally available, but we're not making it the default experience for everyone, as from our testing we think that shorter context plus compaction works best for most people. If anyone here wants to try out 1M, you can do so by overriding `model_context_window` and `model_auto_compact_token_limit`.
Curious to hear if people have use cases where they find 1M works much better!
(I work at OpenAI.)
I too tried Codex and found it similarly hard to control over long contexts. It ended up coding an app that spit out millions of tiny files which were technically smaller than the original files it was supposed to optimize, except due to there being millions of them, actual hard drive usage was 18x larger. It seemed to work well until a certain point, and I suspect that point was context window overflow / compaction. Happy to provide you with the full session if it helps.
I’ll give Codex another shot with 1M. It just seemed like cperciva’s case and my own might be similar in that once the context window overflows (or refuses to fill) Codex seems to lose something essential, whereas Claude keeps it. What that thing is, I have no idea, but I’m hoping longer context will preserve it.
Feels like a losing battle, but hey, the audience is usually right.
https://apps.apple.com/us/app/clean-links-qr-code-reader/id6...
Anywhere I can toss a Tip for this free app?
Especially when it’s to the point of, you know, nagging/policing people to do it the way you’d prefer, when you could just redirect your router requests from x.com to xcancel.com
For me, I would say speed (not just time to first token, but a complete generation) is more important then going for a larger context size.
Reverse engineering [1]. When decompiling a bunch of code and tracing functionality, it's really easy to fill up the context window with irrelevant noise and compaction generally causes it to lose the plot entirely and have to start almost from scratch.
(Side note, are there any OpenAI programs to get free tokens/Max to test this kind of stuff?)
Sometimes I’m exploring some topic and that exploration is not useful but only the summary.
Also, you could use the best guess and cli could tell me that this is what it wants to compact and I can tweak its suggestion in natural language.
Context is going to be super important because it is the primary constraint. It would be nice to have serious granular support.
Like, I'd love an optional pre-compaction step, "I need to compact, here is a high level list of my context + size, what should I junk?" Or similar.
The agent could pre-select what it thinks is worth keeping, but you’d still have full control to override it. Each chunk could have three states: drop it, keep a summarized version, or keep the full history.
That way you stay in control of both the context budget and the level of detail the agent operates with.
I'm now more careful, using tracking files to try to keep it aligned, but more control over compaction regardless would be highly welcomed. You don't ALWAYS need that level of control, but when you do, you do.
I've also had it succeed in attempts to identify some non-trivial bugs that spanned multiple modules.
https://developers.openai.com/api/docs/pricing is what I always reference, and it explicitly shows that pricing ($2.50/M input, $15/M output) for tokens under 272k
It is nice that we get 70-72k more tokens before the price goes up (also what does it cost beyond 272k tokens??)
Even right now one page refers to prices for "context lengths under 270K" whereas another has pricing for "<272K context length"
For example on Artificial Analysis, the GPT-5.x models' cost to run the evals range from half of that of Claude Opus (at medium and high), to significantly more than the cost of Opus (at extra high reasoning). So on their cost graphs, GPT has a considerable distribution, and Opus sits right in the middle of that distribution.
The most striking graph to look at there is "Intelligence vs Output Tokens". When you account for that, I think the actual costs end up being quite similar.
According to the evals, at least, the GPT extra high matches Opus in intelligence, while costing more.
Of course, as always, benchmarks are mostly meaningless and you need to check Actual Real World Results For Your Specific Task!
For most of my tasks, the main thing a benchmark tells me is how overqualified the model is, i.e. how much I will be over-paying and over-waiting! (My classic example is, I gave the same task to Gemini 2.5 Flash and Gemini 2.5 Pro. Both did it to the same level of quality, but Gemini took 3x longer and cost 3x more!)
It was generally smarter than pre-5.2 so strategically better, and codex likewise wrote better database queries than non-codex, and as it needs to iteratively hunt down the answer, didn't run out the clock by drowning in reasoning.
Video: https://media.ccc.de/v/39c3-breaking-bots-cheating-at-blue-t...
We'll be updating numbers on 5.3 and claude, but basically same thing there. Early, but we were surprised to see codex outperform opus here.
I find both Codex and Claude Opus perform at a similar level, and in some ways I actually prefer Codex (I keep hitting quota limits in Opus and have to revert back to Sonnet).
If your question is related to morality (the thing about US politics, DoD contract and so on)... I am not from the US, and I don't care about its internal politics. I also think both OpenAI and Anthropic are evil, and the world would be better if neither existed.
Exact same situation here. I've been using both extensively for the last month or so, but still don't really feel either of them is much better or worse. But I have not done large complex features with it yet, mostly just iterative work or small features.
I also feel I am probably being very (overly?) specific in my prompts compared to how other people around me use these agents, so maybe that 'masks' things
I can tell claude to spawn a new coding agent, and it will understand what that is, what it should be told, and what it can approximately do.
Codex on the other hand will spawn an agent and then tell it to continue with the work. It knows a coding agent can do work, but doesn't know how you'd use it - or that it won't magically know a plan.
You could add more scaffolding to fix this, but Claude proves you shouldn't have to.
I suspect this is a deeper model "intelligence" difference between the two, but I hope 5.4 will surprise me.
That's not the experience I have. I had it do more complex changes spawning multiple files and it performed well.
I don't like using multiple agents though. I don't vibe code, I actually review every change it makes. The bottleneck is my review bandwidth, more agents producing more code will not speed me up (in fact it will slow me down, as I'll need to context switch more often).
There's no mention of pricing, quotas and so on. Perhaps Codex will still be preferable for coding tasks as it is tailored for it? Maybe it is faster to respond?
Just speculation on my part. If it becomes redundant to 5.4, I presume it will be sunset. Or maybe they eventually release a Codex 5.4?
My own tooling throws off requests to multiple agents at the same time, then I compare which one is best, and continue from there. Most of the time Codex ends up with the best end results though, but my hunch is that at one point that'll change, hence I continue using multiple at the same time.
Like, if you really don’t want to spend any effort trimming it down, sure use 1m.
Otherwise, 1m is an anti pattern.
It might be my AGENTS.md requiring clearer, simpler language, but at least 5.4's doing a good job of following the guidelines. 5.3-Codex wasn't so great at simple, clear writing.
If you gave the exact same markdown file to me and I posted ed the exact same prompts as you, would I get the same results?
>Today, we’re releasing GPT‑5.4 in ChatGPT (as GPT‑5.4 Thinking),
>Note that there is not a model named GPT‑5.3 Thinking
They held out for eight months without a confusing numbering scheme :)
We got:
- GPT-5.1
- GPT-5.2 Thinking
- GPT-5.3 (codex)
- GPT-5.3 Instant
- GPT-5.4 Thinking
- GPT-5.4 Pro
Who’s to blame for this ridiculous path they are taking? I’m so glad I am not a Chat user, because this adds so much unnecessary cognitive load.
The good news here is the support for 1M context window, finally it has caught up to Gemini.
What the hell is a "safety score for violence"?
A “safety score for violence” is usually a risk rating used by platforms, AI systems, or moderation tools to estimate how likely a piece of content is to involve or promote violence. It’s not a universal standard—different companies use their own versions—but the idea is similar everywhere.
What it measures
A safety score typically evaluates whether text, images, or videos contain things like:
Threats of violence (“I’m going to hurt someone.”) Instructions for harming people Glorifying violent acts Descriptions of physical harm or abuse Planning or encouraging attacks
{ tools: [ { name: "nuke", description: "Use when sure.", ... { lat: number, long: number } } ] }It's very similar to "Battle Brothers", and the fact that RPG games require art assets, AI for enemy moves, and a host of other logical systems makes it all the more impressive.
> we’re also releasing an experimental Codex skill called “Playwright (Interactive) (opens in a new window)”. This allows Codex to visually debug web and Electron apps; it can even be used to test an app it’s building, as it’s building it.
This is on the edge of what the frontier models can do. For 5.4, the result is better than 5.3-Codex and Opus 4.6. (Edit: nowhere near the RPG game from their blog post, which was presumably much more specced out and used better engineering setup).
I also tested it with a non-trivial task I had to do on an existing legacy codebase, and it breezed through a task that Claude Code with Opus 4.6 was struggling with.
I don't know when Anthropic will fire back with their own update, but until then I'll spend a bit more time with Codex CLI and GPT 5.4.
They show an example of 5.4 clicking around in Gmail to send an email.
I still think this is the wrong interface to be interacting with the internet. Why not use Gmail APIs? No need to do any screenshot interpretation or coordinate-based clicking.
Screenshots on the other hand are documentation, API, and discovery all in one. And you’d be surprised how little context/tokens screenshots consumer compared to all the back and forth verbose json payloads of APIs
I think an important thing here is that a lot of websites/platforms don't want AIs to have direct API access, because they are afraid that AIs would take the customer "away" from the website/platform, making the consumer a customer of the AI rather than a customer of the website/platform. Therefore for AIs to be able to do what customers want them to do, they need their browsing to look just like the customer's browsing/browser.
Of course APIs and CLIs also exist, but they don't necessarily have feature parity, so more development would be needed. Maybe that's the future though since code generation is so good - use AI to build scaffolding for agent interaction into every product.
If an API is exposed you can just have the LLM write something against that.
Optimizations are secondary to convenience
But people are intimidated by the complexity of writing web crawlers because management has been so traumatized by the cost of making GUI applications that they couldn’t believe how cheap it is to write crawlers and scrapers…. Until LLMs came along, and changed the perceived economics and created a permission structure. [1]
AI is a threat to the “enshittification economy” because it lets us route around it.
[1] that high cost of GUI development is one reason why scrapers are cheap… there is a good chance that the scraper you wrote 8 years ago still works because (a) they can’t afford to change their site and (b) if they could afford to change their site changing anything substantial about it is likely to unrecoverably tank their Google rankings so they won’t. A.I. might change the mechanics of that now that you Google traffic is likely to go to zero no matter what you do.
Plenty of companies make the same choice about their API, they provide it for a specific purpose but they have good business reasons they want you using the website. Plenty of people write webcrawlers and it's been a cat and mouse game for decades for websites to block them.
This will just be one more step in that cat and mouse game, and if the AI really gets good enough to become a complete intermediary between you and the website? The website will just shutdown. We saw it happen before with the open web. These websites aren't here for some heroic purpose, if you screw their business model they will just go out of business. You won't be able to use their website because it won't exist and the website that do exist will either (a) be made by the same guys writing your agent, and (b) be highly highly optimized to get your agent to screw you.
This is prescient -- I wonder if the Big Tech entities see it this way. Maybe, even if they do, they're 100% committed to speedrunning the current late-stage-cap wave, and therefore unable to do anything about it.
Google has a good model in the form of Gemini and they might figure they can win the AI race and if the web dies, the web dies. YouTube will still stick around.
Facebook is not going to win the AI race with low I.Q. Llama but Zuck believed their business was cooked around the time it became a real business because their users would eventually age out and get tired of it. If I was him I'd be investing in anything that isn't cybernetic let it be gold bars or MMA studios.
Microsoft? They bought Activision for $69 billion. I just can't explain their behavior rationally but they could do worse than their strategy of "put ChatGPT in front of laggards and hope that some of them rise to the challenge and become slop producers."
Amazon is really a bricks-and-mortar play which has the freedom to invest in bricks-and-mortar because investors don't think they are a bricks-and-mortar play.
Netflix? They're cooked as is all of Hollywood. Hollywood's gatekeeping-industrial strategy of producing as few franchise as possible will crack someday and our media market may wind up looking more like Japan, where somebody can write a low-rent light novel like
https://en.wikipedia.org/wiki/Backstabbed_in_a_Backwater_Dun...
and J.C. Staff makes a terrible anime that convinces 20k Otaku to drop $150 on the light novels and another $150 on the manga (sorry, no way you can make a balanced game based on that premise!) and the cost structure is such that it is profitable.
I am not sure about that. We techies avoid enshittification because we recognize shit. Normies will just get their syncopatic enshittified AI that will tell them to continue buying into walled gardens.
some sites try to block programmatic use
UI use can be recorded and audited by a non-technical person
You can also test this yourself easily, fire up two agents, ask one to use PL meant for humans, and one to write straight up machine code (or assembly even), and see which results you like best.
Then go ahead and make an argument. "Why not do X?" is not an argument, it's a suggestion.