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

GPT-4.1 in the API

openai.com

Post reply on HN
Print view View original

121–130 of 513 posts

Re: GPT-4.1 in the API

#121

Post by Philpax »

camdenreslink wrote: post #87
rvz wrote: post #21

Earlier quoted context omitted.

> Yesterday, using v0, I replicated a full nextjs UI copying a major saas player. No backend integration, but the design and UX were stunning, and better than I could do if I tried. Exactly. Those who do frontend or focus on pretty much anything Javascript are, how should I say it? Cooked? > Software will get automated The first to go are those that use JavaScript / TypeScript engineers have already been automated ou…

I have a medium-sized typescript personal project I work on. It probably has 20k LOC of well organized typescript (react frontend, express backend). I also have somewhat comprehensive docs and cursor project rules. In general I use Cursor in manual mode asking it to make very well scoped small changes (e.g. “write this function that does this in this exact spot”). Yesterday I needed to make a largely mechanical chang…

> Cursor agent mode puked all over itself using Gemini 2.5. It could summarize what changes would need to be made, but it was totally incapable of making the changes.

Gemini 2.5 is currently broken with the Cursor agent; it doesn't seem to be able to issue tool calls correctly. I've been using Gemini to write plans, which Claude then executes, and this seems to work well as a workaround. Still unfortunate that it's like this, though.

Re: GPT-4.1 in the API

#122

Post by greenavocado »

croes wrote: post #19
polytely wrote: post #18

It seems that OpenAI is really differentiating itself in the AI market by developing the most incomprehensible product names in the history of software.

They learned from the best: Microsoft

Microsoft Neural Language Processing Hyperscale Datacenter Enterprise Edition 4.1

A massive transformer-based language model requiring:

- 128 Xeon server-grade CPUs

- 25,000MB RAM minimum (40,000MB recommended)

- 80GB hard disk space for model weights

- Dedicated NVIDIA Quantum Accelerator Cards (minimum 8)

- Enterprise-grade cooling solution

- Dedicated 30-amp power circuit

- Windows NT Advanced Server with Parallel Processing Extensions

~

Features:

- Natural language understanding and generation

- Context window of 8,192 tokens

- Enterprise security compliance module

- Custom prompt engineering interface

- API gateway for third-party applications

*Includes 24/7 on-call Microsoft support team and requires dedicated server room with raised floor cooling

Re: GPT-4.1 in the API

#123

Post by nneonneo »

kmeisthax wrote: post #117
elashri wrote: post #75

Are there any benchmarks or someone who did tests of performance of using this long max token models in scenarios where you actually use more of this token limit? I found from my experience with Gemini models that after ~200k that the quality drops and that it basically doesn't keep track of things. But I don't have any numbers or systematic study of this behavior. I think all providers who announce increased max tok…

The problem is that while you can train a model with the hyperparameter of "context size" set to 1M, there's very little 1M data to train on. Most of your model's ability to follow long context comes from the fact that it's trained on lots of (stolen) books; in fact I believe OpenAI just outright said in court that they can't do long context without training on books. Novels are usually measured in terms of words; an…

I mean, can’t they just train on some huge codebases? There’s lots of 100KLOC codebases out there which would probably get close to 1M tokens.

Re: GPT-4.1 in the API

#124

Post by taikahessu »

> They feature a refreshed knowledge cutoff of June 2024.

As opposed to Gemini 2.5 Pro having cutoff of Jan 2025.

Honestly this feels underwhelming and surprising. Especially if you're coding with frameworks with breaking changes, this can hurt you.

Re: GPT-4.1 in the API

#126

Post by GaggiX »

asdev wrote: post #67

it's worse than 4.5 on nearly every benchmark. just an incremental improvement. AI is slowing down

It's better on AIME '24, Multilingual MMLU, SWE-bench, Aider’s polyglot, MMMU, ComplexFuncBench while being much much cheaper and smaller.

Re: GPT-4.1 in the API

#127

Post by tootyskooty »

asdev wrote: post #108

> We will also begin deprecating GPT‑4.5 Preview in the API, as GPT‑4.1 offers improved or similar performance on many key capabilities at much lower cost and latency. why would they deprecate when it's the better model? too expensive?

sits on too many GPUs, they mentioned it during the stream

I'm guessing the (API) demand isn't there to saturate them fully

Re: GPT-4.1 in the API

#128

Post by yberreby »

Tiberium wrote: post #116
yberreby wrote: post #68

Earlier quoted context omitted.

Good point, but how does that relate to, or explain, the decision not to release 4.1 in ChatGPT? If they have a nice post-training pipeline to make 4o "nicer" to talk to, why not use it to fine-tune the base 4.1 into e.g. chatgpt-4.1-latest?

Because chatgpt-4o-latest already has all of those improvements, the largest point of this release (IMO) is to offer developers a stable snapshot of something that compares to modern 4o latest. Altman said that they'd offer a stable snapshot of chatgpt 4o latest on the API, he perhaps did really mean GPT 4.1.

> Because chatgpt-4o-latest already has all of those improvements

Does it, though? They said that "many" have already been incorporated. I simply don't buy their vague statements there. These are different models. They may share some training/post-training recipe improvements, but they are still different.

Re: GPT-4.1 in the API

#129

Post by forbiddenvoid »

taikahessu wrote: post #124

> They feature a refreshed knowledge cutoff of June 2024. As opposed to Gemini 2.5 Pro having cutoff of Jan 2025. Honestly this feels underwhelming and surprising. Especially if you're coding with frameworks with breaking changes, this can hurt you.

It's definitely an issue. Even the simplest use case of "create React app with Vite and Tailwind" is broken with these models right now because they're not up to date.

Re: GPT-4.1 in the API

#130

Post by roflmaostc »

kmeisthax wrote: post #117
elashri wrote: post #75

Are there any benchmarks or someone who did tests of performance of using this long max token models in scenarios where you actually use more of this token limit? I found from my experience with Gemini models that after ~200k that the quality drops and that it basically doesn't keep track of things. But I don't have any numbers or systematic study of this behavior. I think all providers who announce increased max tok…

The problem is that while you can train a model with the hyperparameter of "context size" set to 1M, there's very little 1M data to train on. Most of your model's ability to follow long context comes from the fact that it's trained on lots of (stolen) books; in fact I believe OpenAI just outright said in court that they can't do long context without training on books. Novels are usually measured in terms of words; an…

What about old books? Wikipedia? Law texts? Programming languages documentations?

How many tokens is a 100 pages PDF? 10k to 100k?

Post reply on HN