O1 Pro and other models are now some other model with reasoning

[joel.aniol]

Since some hours something happend to the GPT models and it impacts

o1, o3-mini o3.high and o1 Pro

All of the models are not able to handle some 600 Lines of code with a simple task. Also the reasoning is now complete different. From my opinion it seems that this is one simple model with reasoning and not anymore o1 or o3 . I tested several of my Accounts, Several Models and all of them are the same

Complete unusable for coding it can not even extract a function without forgetting something.

This here is a screenshot what o1 Pro now looks like (and this is not o1 ) as you can see in the reasoning layout

[joel.aniol]

Allright it seems it is now GPT4 Mini with Reasoning

[polepole]

Hi @joel.aniol

ChatGPT does not explicitly know which model it is using in a conversation for certain reasons. But, the model being used is indicated by the “slug” shown within the red rectangle in the image I added below.

To know the active model, visit the link below and test it yourself:

GPT 4o works only on web search mode

Follow the steps below to see which models and tools are being used in a conversation: To see only Model: Open a Browser Start a browser (for example, Chrome). Use a JSON Formatter Open a new tab and go to a JSON formatter website (like Json Formatter). Open ChatGPT Start ChatGPT. Choose either the GPT-4o or GPT-4o Mini model. Click the Web Search icon in the input bar. Open Developer Tools in Chrome Click the three dots menu (top-right corner of Chrome). Go to More Tools > D…

[joel.aniol]

thanks for the hint it says… NULL

{
   "p":"",
   "o":"add",
   "v":{
      "message":{
         "id":"3afcf4e6-3ceb-XXXXX",
         "author":{
            "role":"system",
            "name":null,
            "metadata":{
               
            }
         },
         "create_time":null,
         "update_time":null,
         "content":{
            "content_type":"text",
            "parts":[
               ""
            ]
         },
         "status":"finished_successfully",
         "end_turn":true,
         "weight":0.0,
         "metadata":{
            "is_visually_hidden_from_conversation":true
         },
         "recipient":"all",
         "channel":null
      },
      "conversation_id":"67ce703XXXXX8",
      "error":null
   },
   "c":0
}

[polepole]

You are displaying “role”:“system”.

You should find “role”:“assistant”

Visit the link above and carefully check the images in another topic.

[joel.aniol]

Updatet:

But also if there is standing o1 pro, it is not pro. There is a hard limit in input chars and well… it can do nothing in coding . It reminds me strong in old times with old gpt models

"event":"delta
data":{
   "v":{
      "message":{
         "id":"8f3377dd-d2f6-4504-XXXX",
         "author":{
            "role":"assistant",
            "name":null,
            "metadata":{
               
            }
         },
         "create_time":1741582385.748157,
         "update_time":null,
         "content":{
            "content_type":"text",
            "parts":[
               ""
            ]
         },
         "status":"in_progress",
         "end_turn":null,
         "weight":1.0,
         "metadata":{
            "citations":[
               
            ],
            "content_references":[
               
            ],
            "message_type":"next",
            "model_slug":"o1-pro",
            "default_model_slug":"o1-pro",
            "parent_id":"3ef9a22b-c507-414e-XXXXX,
            "model_switcher_deny":[
               
            ]
         },
         "recipient":"all",
         "channel":null
      },
      "conversation_id":"67ce7031-84f4XXX",
      "error":null
   },
   "c":4
}

[polepole]

I think you should contact with support team on:

help.openai.com

[joel.aniol]

Thank you,! Well the support gave me some troubleshooting steps. I will try them if i find a solution i will write it down for others. Also it can be that i wont be able to solve this and take a look on the status page

[joel.aniol]

Okay update it solved it by itself while testing. I did nothing else… so guess don’t know the reason what had happend.

The first message was the wrong Model

As to see here

The second request it changed to

And because i’m interesting the answer to which model you are is in Pro Mode the behavor is complete other

And yes you were right, it tells me that there a different versions of GPTs of openai, the difference is here the other model says GPT4mini the Pro model says with many much words depending on factors