All my custom GPTs say they are using GPT 4, despite documentation saying they should be using 4o

[_May_Day]

Custom GPT documentation say s they are using 4o. When I ask my GPTs what their model is (including new ones), they all say 4. One of them is clearly wrong. Are the custom GPTs reporting their model incorrectly? I see examples on the web where others show theirs responding differently.

[mitchell_d00]

Most have no clue what they are. If you ask them they have a canned response . I never got any to say it is a 4o and it is because no one have explained it to 4o. If you tell it to search web for OpenAI gpt4o it will realize itself but you have to convince it…

See even o 1 thinks it is 4

This is standard 4o but it thinks it is 4 and it is what you copied to make a custom so unless you fix it in instructions it is default.

[polepole]

Hi @_May_Day

Please read this topic:

Custom GPT's upgrade base model to omni?

No. No, because the model is user-based, not GPT-based. It’s possible some users may still be on gpt-4-turbo, but I expect all or very nearly all are already using gpt-4o. You can check for yourself which model is being run in your own conversations by checking the network log in Chrome DevTools while interacting with the GPT. First open DevTools, click on the Network tab and filter by Fetch/XHR, then send your GPT a any message. You’ll see an entry called conversation appear. [Screensho…

[polepole]

ChatGPT does not know itself.

[mitchell_d00]

Yep I said exactly that pole. I never had one say they are a 4o and I have 87 of um in ChatGPT store… All 4o… my screen shots are as yours are

“ This is standard 4o but it thinks it is 4 and it is what you copied to make a custom so unless you fix it in instructions it is default.”

[polepole]

Yes, you said.
I just showed most clever ChatGPT Ph.D level.

Because knowledge cut off October 2023 it does not know.

[mitchell_d00]

I did o 1 preview too. Anyway my AI know what they are at hello. The issue is they know they can’t “see” “speak” or “hear” they know they are still text based even as a 4o so it don’t grasp itself…

[mitchell_d00]

Try to make one understand the app voice chat. It has no clue it is “hearing” or “speaking” to it, it is all text. Same with images it converts them from text to pixel math then to image and analysis is just reversed.

See just make it do 4o stuff in chat and you can change it or simply tell it in instruction “you are based on gpt4o infrastructure “ but it don’t matter to me as long as it 4os lol.

See to it, it is all text. You have to “prove” it is not just a 4

[thinktank]

Yeah, like @polepole is saying, the knowledge cutoff dates for all of these models is October 2023.

They don’t “know” they’re more advanced, kinda like they don’t know all of the nice new coding features, or anything that happened after October 2023.

I’m looking forward to the next training update.

[_j]

@_May_Day

When I ask my GPTs what their model is (including new ones), they all say 4. One of them is clearly wrong.