Very important note:

>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

If anyone here doesn't know, OpenAI does offer the ChatGPT model version in the API as chatgpt-4o-latest, but it's bad because they continuously update it so businesses can't reliably rely on it being stable, that's why OpenAI made GPT 4.1.

OpenAI (and most LLM providers) allow model version pinning for exactly this reason, e.g. in the case of GPT-4o you can specify gpt-4o-2024-05-13, gpt-4o-2024-08-06, or gpt-4o-2024-11-20.

https://platform.openai.com/docs/models/gpt-4o

> chatgpt-4o-latest, but it's bad because they continuously update it

Version explicitly marked as "latest" being continuously updated it? Crazy.

So you're saying that "ChatGPT-4o-latest (2025-03-26)" in LMarena is 4.1?

No, that is saying that some of the improvements that went into 4.1 have also gone into ChatGPT, including chatgpt-4o-latest (2025-03-26).

Who has the numbers vs. Claude Sonnet 4.6 and Gemini 2.5 Pro? And is it available in Cursor yet?

GPT-4.1 probably is a distilled version of GPT-4.5

I dont understand the constant complaining about naming conventions. The number system differentiates the models based on capability, any other method would not do that. After ten models with random names like "gemini", "nebula" you would have no idea which is which. Its a low IQ take. You dont name new versions of software as completely different software

Also, 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. I have 15 years of backend experience at FAANG. Software will get automated, and it already is, people just havent figured it out yet

> The number system differentiates the models based on capability, any other method would not do that.

Please rank GPT-4, GPT-4 Turbo, GPT-4o, GPT-4.1, GPT-4.5, o1-mini, o1, o1 pro, o3-mini, o3-mini-high, o3, and o4-mini in terms of capability without consulting any documentation.

Btw, as someone who agrees with your point, what’s the actual answer to this?

There's no single ordering -- it really depends on what you're trying to do, how long you're willing to wait, and what kinds of modalities you're interested in.

Really? Is o3-mini-high better than o1-pro?

> You dont name new versions of software as completely different software

macOS releases would like a word with you.

https://en.wikipedia.org/wiki/MacOS#Timeline_of_releases

Technically they still have numbers, but Apple hides them in marketing copy.

https://www.apple.com/macos/macos-sequoia/

Though they still have “macOS” in the name.

Feel free to lay the naming convention rules out for us man.

Just add SemVer with an extra tag:

4.0.5.worsethan4point5

> 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 out of a job. It is all over for them.

Yeah its over for them. Complicated business logic and sprawling systems are what are keeping backend safe for now. But the big front end code bases where individual files (like react components) are largely decoupled from the rest of the code base is why front end is completely cooked

> I don't understand the constant complaining about naming conventions.

Oh man. Unfolding my lawn chair and grabbing a bucket of popcorn for this discussion.

> using v0, I replicated a full nextjs UI copying a major saas player. No backend integration, but the design and UX were stunning

AI is amazing, now all you need to create a stunning UI is for someone else to make it first so your AI can rip it off. Not beating the "plagiarism machine" allegations here.

Heres a secret: Most of the highest funded VC backed software companies are just copying a competitor with a slight product spin/different pricing model

Exactly, they like to call it “bringing new energy to an old industry”.

Got any examples?

No benchmark comparisons to other models, especially Gemini 2.5 Pro, is telling.

Gemini 2.5 Pro gets 64% on SWE-bench verified. Sonnet 3.7 gets 70%

They are reporting that GPT-4.1 gets 55%.

Are those with «thinking» or without?

Big focus on coding. It feels like a defensive move against Claude (and more recently, Gemini Pro) which became very popular in that regime. I guess they recently figured out some ways to train the model for these "agentic" coding through RL or something - and the finding is too new to apply 4.5 on time.

ChatGPT currently recommends I use o3-mini-high ("great at coding and logic") when I start a code conversation with 4o.

I don't understand why the comparison in the announcement talks so much about comparing with 4o's coding abilities to 4.1. Wouldn't the relevant comparison be to o3-mini-high?

4.1 costs a lot more than o3-mini-high, so this seems like a pertinent thing for them to have addressed here. Maybe I am misunderstanding the relationship between the models?

It's not the point of the announcement, but I do like the use of the (abs) subscript to demonstrate the improvement in LLM performance since in these types of benchmark descriptions I never can tell if the percentage increase is absolute or relative.

pretty wild versioning that GPT 4.1 is newer and better in many regards than GPT 4.5.

I think they're doing it deliberately at this point

Tomorrow they are releasing the open source GPT-1.4 model :P

Testing against unspecified other "leading" models allows for shenanigangs:

> Qodo tested GPT‑4.1 head-to-head against other leading models [...] they found that GPT‑4.1 produced the better suggestion in 55% of cases

The linked blog post goes 404: https://www.qodo.ai/blog/benchmarked-gpt-4-1/

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".

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.

For conversational AI, the most significant part is GPT-4.1 being 2x faster than GPT-4o at basically the same reasoning capabilities.

Does someone have the benchmarks compared to other models?

claude 3.7 no thinking (diff) - 60.4%

claude 3.7 32k thinking tokens (diff) - 64.9%

GPT-4.1 (diff) - 52.9% (stat is from the blog post)

https://aider.chat/docs/leaderboards/

GPT-4.1 Pricing (per 1M tokens):

gpt-4.1

- Input: $2.00

- Cached Input: $0.50

- Output: $8.00

gpt-4.1-mini

- Input: $0.40

- Cached Input: $0.10

- Output: $1.60

gpt-4.1-nano

- Input: $0.10

- Cached Input: $0.025

- Output: $0.40

The cached input price is notable here: previously with GPT-4o it was 1/2 the cost of raw input, now it's 1/4th.

It's still not as notable as Claude's 1/10th the cost of raw input, but it shows OpenAI's making improvements in this area.

The big change about this announcement is the 1M context window on all models.

But the price is what matters.

Nothing compared to Llama 4's 7M. What matters is how well it performs with such long context, not what the technical maximum is.

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

GpTeams Classic

GPT 4 Workgroups

I wonder how they decide whether the o or the digit needs to come first. (eg. o3 vs 4o)

I need an AI to understand the naming conventions that OpenAI is using.

so we're going back... .4 of a gpt? make it make sense openai..