Hacker News Reader

Comment by yberreby | Hacker News Reader

> Note that GPT‑4.1 will only be available via the API. In ChatGPT, many of the improvements in instruction following, coding, and intelligence have been gradually incorporated into the latest version (opens in a new window) of GPT‑4o, and we will continue to incorporate more with future releases.

The lack of availability in ChatGPT is disappointing, and they're playing on ambiguity here. They are framing this as if it were unnecessary to release 4.1 on ChatGPT, since 4o is apparently great, while simultaneously showing how much better 4.1 is relative to GPT-4o.

One wager is that the inference cost is significantly higher for 4.1 than for 4o, and that they expect most ChatGPT users not to notice a marginal difference in output quality. API users, however, will notice. Alternatively, 4o might have been aggressively tuned to be conversational while 4.1 is more "neutral"? I wonder.

View on HN

There's a HUGE difference that you are not mentioning: there are "gpt-4o" and "chatgpt-4o-latest" on the API. The former is the stable version (there are a few snapshot but the newest snapshot has been there for a while), and the latter is the fine-tuned version that they often update on ChatGPT. All those benchmarks were done for the API stable version of GPT-4o, since that's what businesses rely on, not on "chatgpt-4o-latest".

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.

I disagree. From the average user perspective, it's quite confusing to see half a dozen models to choose from in the UI. In an ideal world, ChatGPT would just abstract away the decision. So I don't need to be an expert in the relatively minor differences between each model to have a good experience.

Vs in the API, I want to have very strict versioning of the models I'm using. And so letting me run by own evals and pick the model that works best.

> it's quite confusing to see half a dozen models to choose from in the UI. In an ideal world, ChatGPT would just abstract away the decision

Supposedly that’s coming with GPT 5.

I agree on both naming on stability. However, this wasn't my point.

They still have a mess of models in ChatGPT for now, and it doesn't look like this is going to get better immediately (even though for GPT-5, they ostensibly want to unify them). You have to choose among all of them anyway.

I'd like to be able to choose 4.1.

Reply on news.ycombinator.com