ChatGPT model should be renamed to GPT 5 Chat

theDoctor

OpenAI has a model called 'ChatGPT' which still points to the version of 4o that is used in ChatGPT. This is their only model that references 'ChatGPT' by name - no 5.x endpoints use that nomenclature.
https://platform.openai.com/docs/models/chatgpt-4o-latest

Meanwhile, Kagi's 'ChatGPT' model is pointing to the 5.x Chat endpoint.

As a developer who works with these models directly, when I saw the 'ChatGPT' model, I assumed it was this older 4o-based endpoint.

Reasons to change this nomenclature:

Specificity/Clarity - The default model in ChatGPT is actually the auto-router. If Kagi's 'ChatGPT' is mirroring the 'Chat' endpoint, then it should say so. This makes it clear and accurate as to which model you are actually using.
Uniformity - Claude, GLM, and Qwen all have non-reasoning models in Kagi Assistant and they are labeled directly as such. Meanwhile, 5.x Chat is using an entirely different name.

Also, I assume this is a relic from when Kagi's 'ChatGPT' model was actually the 4o-based model, but if this is just the 5.x Chat model, the Last updated field should no longer be listed as 'Rolling', as the 'Chat' endpoints are not rolling like the 'ChatGPT' endpoint was.

The 'ChatGPT' model in the assistant should be listed as 'GPT 5.x Chat' or 'GPT 5.x Instant' with appropriate help text similar to how Claude models are listed.

yiwei-1

Hi, just to clarify:

GPT-5.x Chat is a rolling-update model, so there is no fixed release date. The official model name in the API is gpt-5.2-chat-latest, which does not include a timestamp.
We initially named it something like ChatGPT-4o at launch, but we received feedback from users who were confused about which one was the same model they used in the ChatGPT app. As a result, we renamed it to simply ChatGPT. Perhaps we could consider adding a note in the model description indicating that it currently points to the GPT-5.x Chat model?

theDoctor

At a minimum, a clarification in the description would be nice.

The model "used in the ChatGPT app" could be one of many at this point. I understand that the premise is that 5.x is the default, but even then, the default is the auto router, which may be -chat or may be -thinking. There is also a Claude app, but Kagi does not list any models as just "Claude".

I'm not trying to be pedantic, I promise, but clarity and uniformity would be nice. At least a clarification in the hover text, if not just listing a 5.x-Instant / 5.x-Thinking.

Personally, I think that (-Instant vs -Thinking) would be best - it's expressly clear and follows the format of other models. It could also be clarified in the hover text that "This is the default model that ChatGPT uses".