Do you guys even test these things before you push them to the store?
The most recent Android app update with the UI redesign is utterly broken, even when any other model is selected it ALWAYS defaults to o1.
I can even see at the top it says 4o but every response is o1, I change it and regenerate, which for one response uses the correct model, and then it reverts for the next response!!!
Honestly this is totally rookie.
Android Build 1.2025.028
Screenshots attached of me starting a new conversation saying hi and showing the o1 reasoning prompt despite selected 4o model
I’m having the exact same issue. Does anyone know a solution?
Use the website mobile UI or activate the voice mode (Standard works, I’ve. not tried Advanced) and it forces it into 4o.
This is happening to me as well and several reports on Reddit of the same.
This worked, but I interact with ChatGPT mostly through prompts in the app, so hopefully they fix this
Same issue here. I tried fixing it by uninstalling, cleared cache, reinstall. And it worked fine the 1 try (using 4o). Tried it again, and it is defaulting to o1 again and displaying limit issue.
Running version ChatGPT for Android 1.2025.035
On my end, I did exactly the same thing and got the same results. Additionally, the first chat has access to the chat history, but any new chats after that don’t.
Same for me. Uninstalled, cleared all data. Even went to regular ChatGPT instead of beta. Works for the first prompt, then defaults to o1. Also using version 1.2025.035 for Android.
Looks like they fixed it. Mine is now working as expected.
