Again, ChatGpt switched from 4.0 to 3.5. Need help

ChatGPT
1

I have a big problem. In the middle of my work, version 4.0 has told me that I have reached the limit and I have to wait until a certain time to be able to continue working with model 4.0. After this information, the model automatically switched to 3.5.
The specified time has long since been reached but ChatGpt no longer reverts to version 4.0

Can anyone help me?
I have already read a lot about this topic in the forum here but I don’t understand what I should or can do.

Is there an easy way to get this problem under control?

Best regards and many thanks for your help
Sarah

PS: Unfortunately, support can’t help me or hasn’t responded and I’ve been waiting for an answer for 20 hours.

2 Likes
2

Welcome to the dev forum @WEARESILK

There are hourly limits on the ChatGPT Plus Plan for GPT-4 IIRC.

The support is the only place to best convey the dissatisfaction as feedback. They tend to pay attention to that.

The forum isn’t frequented by support staff.

If you want higher limits I think the Teams plan could do the job, or if your org uses ChatGPT and is large enough, they can get that sweet ChatGPT Enterprise.

1 Like
3

Thank you for your answer.
I am indeed alone and a larger plan would not be worthwhile for me.
I can live with the limit, the only problem is that I have waited the hours during which I can no longer use version 4.0. However, it does not revert back to version 4.0 but remains at version 3.5.
So the limit, i.e. the hours are long gone but unfortunately nothing happens. Do you understand what I mean? It’s difficult for me to write this because I’m not that good technically.

Kind regards

4

I have faced that same thing as well.

If I understand correctly, when you are on 3.5, you aren’t able to switch but have to start a new chat to do that.

5

Hi!
There should be a little star icon below the model reply where you can choose the model variant and even switch during a conversation.
This feature is available when using chat.openai.com

Screenshot_20240425_114857_Opera

1 Like
6

Mh, thats crazy.
I dont have this star icon :frowning:

1 Like
7

Sorry to hear that.
I am also not in the loop which features are currently tested and which ones have been rolled out to all users.
From my earlier experience it was a deliberate choice to proceed with the conversation using 3.5 when hitting the message cap per three hours. In this case I would treat it as a (disliked) hard stop to my current work and proceed later.

One trick you can try is to send a single message to the model before you actually start working. Then you have the rest of the messages available when starting to work and a reset to the message cap when crossing the three hour mark based on the first message you send earlier. This gives you a artificial message cap of 79 for a short period of time.

Hope this helps, somewhat.

1 Like
8

Their support is terrible so I would not look for much help there

9

Fiz a assinatura hoje e estava usando para estudos.

Vou providenciar o cancelamento da assinatura agora mesmo, péssimo serviço, pagamos o GPT plus e temos limite de uso?

Péssimo.

10

yes, you are right. support simply doesn’t respond.
It’s just a shame that I can no longer keep days of work consistent because I have no chance of continuing to write this work in 4.0 instead of 3.5. I always ask myself why I pay for a service if I can’t use it. It’s just really annoying that no one in support can help and you’re on your own with the question or spend hours searching the internet for an answer.

11

You’re gonna find this hilarious, but I frequently experience this large message cap, and other times experience the LOW cap.
The problem is, I don’t know what I did, how my interaction initiated the higher cap.

Will you please be more specific, if you have time? Specifically, I’m only working in a private custom GPT inside My GPT. I’ve found it to be highly convenient, when I have a proper message cap.

Thank you.

12

This feature has just been added to the new version, so updates may affect everyone differently.

13

I’m not entirely sure if I have the same issue or another variation of it.

I am on the Plus plan and occasionally exhaust the usage limit (messages per unit of time, I suppose).

When this happens, the UI looks like this:

image
image811×150 4.06 KB

Notice that there is no indication as to when my usage limit resets. Up until fairly recently (the last few days), the time was displayed, so I could at least go do something else until the stated time.

Now, it’s just a guessing-game. Am I overlooking something that should be obvious here?

Also, there is no star/sparkle icon below each response, as mentioned earlier in the thread.

Furthermore, the message which states, New responses will use GPT-3.5 until your GPT-4 limit resets., is entirely inconsistent with the UI’s behavior.

Not only are the message input field and send button disabled completely, contrary to the message’s implication that I can just keep chatting using 3.5 if I am so inclined, but if I try to switch models “manually” up at the top of the interface, I’m forced to create a new chat.

In summary, despite the glitchy behavior here, all I really care about is knowing when my usage limit resets.

Has anybody else noticed that the explicitly stated time is now gone from the UI?

Thanks for any assistance!