Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

GPT-5.4

openai.com

Post reply on HN
Print view View original

141–150 of 868 posts

Re: GPT-5.4

#141

Post by npilk »

mattas wrote: post #5

"GPT‑5.4 interprets screenshots of a browser interface and interacts with UI elements through coordinate-based clicking to send emails and schedule a calendar event." 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.

It feels like building humanoid robots so they can use tools built for human hands. Not clear if it will pay off, but if it does then you get a bunch of flexibility across any task "for free".

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.

Re: GPT-5.4

#142

Post by zzleeper »

dandiep wrote: post #129

Anyone know why OpenAI hasn't released a new model for fine tuning since 4.1? It'll be a year next month since their last model update for fine tuning.

For me the issue is why there's not a new mini since 5-mini in August.

I have now switched web-related and data-related queries to Gemini, coding to Claude, and will probably try QWEN for less critical data queries. So where does OpenAI fits now?

Re: GPT-5.4

#145

Post by joshuacc »

simianwords wrote: post #94
catketch wrote: post #83

Earlier quoted context omitted.

-codex variant models in earlier version were just fine tuned for coding work, and had a little better performance for related tool calling and maybe instruction calling. in 5.4 it looks like the just collapsed that capability into the single frontier family model

Yes so I’m even more confused. Why would I use codex?

Presumably you don’t anymore if you have 5.4.

Re: GPT-5.4

#147

Post by nsingh2 »

oersted wrote: post #76
ZeroCool2u wrote: post #24

Bit concerning that we see in some cases significantly worse results when enabling thinking. Especially for Math, but also in the browser agent benchmark. Not sure if this is more concerning for the test time compute paradigm or the underlying model itself. Maybe I'm misunderstanding something though? I'm assuming 5.4 and 5.4 Thinking are the same underlying model and that's not just marketing.

I believe you are looking at GPT 5.4 Pro. It's confusing in the context of subscription plan names, Gemini naming and such. But they've had the Pro version of the GPT 5 models (and I believe o3 and o1 too) for a while. It's the one you have access to with the top ~$200 subscription and it's available through the API for a MUCH higher price ($2.5/$15 vs $30/$180 for 5.4 per 1M tokens), but the performance improvement…

From what I've read online it's not necessarily a unquantized version, it seems to go through longer reasoning traces and runs multiple reasoning traces at once. Probably overkill for most tasks.

Re: GPT-5.4

#149

Post by OutOfHere »

paxys wrote: post #136

"Here's a brand new state-of-the-art model. It costs 10x more than the previous one because it's just so good . But don't worry, if you don't want all this power you can continue to use the older one." A couple months later: "We are deprecating the older model."

That's a misrepresentation of the cost. It is simply false. The cost is noted here: https://news.ycombinator.com/item?id=47265144

Re: GPT-5.4

#150

Post by throwup238 »

bazmattaz wrote: post #124

Anyone else feel that it’s exhausting keeping up with the pace of new model releases. I swear every other week there’s a new release!

Yes, that's a common feeling. 5.3-Codex was released a month ago on Feb 5 so we're not even getting a full month within a single brand, let alone between competitors.
Post reply on HN